Bechtel Telecommunications

Technical Journal
An Independent Analysis of Current

Telecommunications Operational Issues January 2006

Global Headquarters
Bechtel Telecommunications Authors Contents
5295 Westview Drive
Frederick, Maryland 21703-8306

Bechtel Telecommunications Technical Journal

Tel: +1 301-228-7500 (Main) or v Foreword
Tel: +1 800-946-3632 (Toll Free)
vii Editorial
Fax: +1 301-694-6971
Timothy D. Statton and Jake MacLeod 1 Telecommunications Research and Development in the United States:
London A State of Crisis?
Bechtel Telecommunications
S. Rasoul Safavian, PhD 13 IP Multimedia Subsystems (IMS): A Standardized Approach to
245 Hammersmith Road
All-IP Converged Networks
Hammersmith
London W6 8DP Brian Coombe 37 A Survey of MEMS-Enabled Optical Devices – Applications and
United Kingdom Drivers for Deployment
Tel: +44 208-846-5093
Fax: +44 208-846-4938
Jungnam Yun, PhD 45 PHY/MAC Cross-Layer Issues in Mobile WiMAX
(POSDATA America R&D Center) and (Invited Paper)
Prof. Mohsen Kavehrad, PhD
(The Pennsylvania State University [CITCTR])
www.bechteltelecoms.com
Peter Moskal and Krishnamurthy Raghu 57 ANSI/TIA Standard 222 – Structural Standard for
Antenna Supporting Structures and Antennas:
A Comparison of Revisions F and G

Brian Perkins 65 Reducing the Amount of Fiber in Fiber-to-the-Home Networks

Esmael Dinan, PhD, and Aleksey A. Kurochkin 73 The Impacts of Antenna Azimuth and Tilt Installation Accuracy
on UMTS Network Performance

Glenn A. Torshizi 81 2.4 GHz Wi-FiTM Phased Array Antenna Evaluation

Volume 4, No. 1
2006

T ELECOMMUNICATIONS Volume 4 No. 1

T ELECOMMUNICATIONS
7211
 Bechtel Offices

Abu Dhabi Houston Moscow Shanghai

Tel: (971-2) 631-8999 Tel: (1-713) 235-2000 Tel: (7-095) 961-3020 Tel: (86-21) 3305-4567
Fax: (971-2) 631-0091 Fax: (1-713) 960-9031 Fax: (7-095) 961-3015 Fax: (86-21) 6391-2999

Al Khobar, Saudi Arabia* Istanbul New Delhi Singapore

Tel: (966-3) 882-5288 Tel: (90-212) 278-2900 Tel: (94-124) 509-8000 Tel: (65) 6332-1220
Fax: (966-3) 882-5306 Fax: (90-212) 278-2913 Fax: (94-124) 234-3110 Fax: (65) 6332-1335

TURNKEY Bangkok Jakarta New York Taipei*

DEPLOYMENT Tel: (66-2) 545-7100 Tel: (62-21) 5291-6000 Tel: (1-212) 689-0100 Tel: (886-2) 2703-2880
SERVICES Fax: (66-2) 545-7777 Fax: (62-21) 5296-2278 Fax: (1-212) 689-4829 Fax: (886-2) 2703-2885

Network Planning Beijing Kuala Lumpur Oak Ridge, Tennessee Tokyo

and RF Tel: (86-10) 8453-9998 Tel: (603) 2078-2050 Tel: (1-865) 220-2000 Tel: (81-3) 3214-4481
Fax: (86-10) 8453-8303 Fax: (603) 2078-1950 Fax: (1-865) 220-2100 Fax: (81-3) 3214-2596
Engineering
Brisbane Lima Paris Toronto
Tel: (61-7) 3295-1400 Tel: (51-1) 445-8746 Tel: (33-1) 4720-5304 Tel: (1-905) 362-6600
Site Acquisition and
Fax: (61-7) 3295-1450 Fax: (51-1) 446-1255 Fax: (33-1) 4720-5506 Fax: (1-905) 362-6767
Right of Way

Buenos Aires London (Hammersmith Rd.) Rio de Janeiro Tysons Corner, Virginia
Procurement
Tel: (54-11) 4315-8000 Tel: (44-208) 846-5111 Tel: (55-21) 2579-2400 Tel: (1-703) 748-9300
and Contracting Fax: (54-11) 4315-0081 Fax: (44-208) 846-6940 Fax: (55-21) 2579-2413 Fax: (1-703) 748-9442

Construction and Cairo London (Pilgrim St.) Riyadh Warrington, England

Equipment Installation Tel: (20-2) 754-6763 Tel: (44-207) 651-7777 Tel: (966-1) 482-9051 Tel: (44-1925) 857-000
Fax: (20-2) 754-6804 Fax: (44-207) 651-7972 Fax: (966-1) 482-9563 Fax: (44-1925) 857-001
Integration and
Optimization Calgary Louisville San Diego Washington, D.C.
Tel: (1-403) 290-5000 Tel: (1-502) 244-2574 Tel: (1-619) 687-8700 Tel: (1-202) 828-5200
Fax: (1-403) 290-5050 Fax: (1-502) 244-2211 Fax: (1-619) 687-8787 Fax: (1-202) 785-2645

Frederick, Maryland Manila San Francisco *Joint-venture office

Tel: (1-301) 228-6000 Tel: (63-2) 751-1000 Tel: (1-415) 768-1234

Fax: (1-301) 228-2200 Fax: (63-2) 751-1005 Fax: (1-415) 768-9038

Glendale, Arizona Mexico City Santiago

Tel: (1-602) 368-1500 Tel: (52-55) 5201-3500 Tel: (56-2) 675-1000
Fax: (1-602) 368-1615 Fax: (52-55) 5202-1785 Fax: (56-2) 232-5208

Hong Kong Montreal Seoul

Tel: (852) 2970-7000 Tel: (1-514) 871-1711 Tel: (82-2) 512-2921
Fax: (852) 2840-1272 Fax: (1-514) 871-1392 Fax: (82-2) 512-2931
 Bechtel
Telecommunications
Technical Journal
January 2006
 Bechtel Telecommunications
Technical Journal
Volume 4, Number 1
Contents

ADVISORY BOARD Foreword v

Jake MacLeod, Principal Vice President and
Chief Technology Officer Editorial vii
S. Rasoul Safavian, PhD, Vice President of Technology,
Americas Regional Business Unit Telecommunications Research and Development 1
Aleksey A. Kurochkin, Deputy Manager, in the United States: A State of Crisis?
Site Development and Engineering
Timothy D. Statton and Jake MacLeod
Brian Coombe, Systems Engineer

IP Multimedia Subsystems (IMS): A Standardized Approach 13

EDITORIAL BOARD
S. Rasoul Safavian, PhD, Editor-in-Chief
Richard Peters, Senior Technical Editor
Teresa Baines, Senior Technical Editor
Barbara Oldroyd, Technical Editor
Ruthanne Evans, Technical Editor
GRAPHICS/DESIGN
Keith Schools, Art Director
Daniel Kemp, Senior Graphic Designer
John Cangemi, Graphic Designer
David Williams, Graphic Designer
Susan Gann, Graphic Designer
Diane Cole, Desktop Publishing
The BTTJ is also available on the Web at
www.bechteltelecoms.com/jsp/labs/pubs.jsp
to All-IP Converged Networks
S. Rasoul Safavian, PhD
A Survey of MEMS-Enabled Optical Devices – 37
Applications and Drivers for Deployment
Brian Coombe
PHY/MAC Cross-Layer Issues in Mobile WiMAX 45
(Invited Paper)
Jungnam Yun, PhD (POSDATA America R&D Center) and
Prof. Mohsen Kavehrad, PhD (The Pennsylvania State University [CITCTR])
ANSI/TIA Standard 222 – Structural Standard for 57
Antenna Supporting Structures and Antennas:
A Comparison of Revisions F and G
Peter Moskal and Krishnamurthy Raghu

Reducing the Amount of Fiber in Fiber-to-the-Home Networks 65

TRADEMARK ACKNOWLEDGMENTS Brian Perkins
All product or service names mentioned in this Bechtel
Telecommunications Technical Journal are trademarks of The Impacts of Antenna Azimuth and Tilt Installation Accuracy 73
their respective companies. Specifically:
on UMTS Network Performance
3GPP is a trademark of the European Telecommunications Esmael Dinan, PhD, and Aleksey A. Kurochkin
Standards Institute (ETSI) in France and other
jurisdictions.
2.4 GHz Wi-FiTM Phased Array Antenna Evaluation 81
AirMagnet is a trademark of AirMagnet, Inc.
Glenn A. Torshizi
Aironet is a trademark of Aironet Wireless Communications,
Inc.
Antenex is a registered trademark of Antenex, Inc.
cdma2000 is a registered trademark of the
Telecommunications Industry Association (TIA-USA).
PLS-POLE is a registered trademark of the Power Line
Systems, Inc.
Telcordia is a registered trademark of Telcordia
Technologies, Inc., in the United States, other countries,
or both.
© 2006 Bechtel Corporation. All rights reserved.
TOWER is a registered trademark of the Power Line
Systems, Inc. Bechtel Telecommunications is a business unit of the Bechtel group of companies, including Bechtel Corporation in
the United States.
UNIX is a registered trademark of The Open Group.
Wi-Fi is a trademark of the Wireless Ethernet Compatibility Bechtel welcomes inquiries concerning the BTTJ. For further information or for permission to reproduce any paper
included in this publication in whole or in part, please contact Bechtel Telecommunications, 5275 Westview Drive,
Alliance, Inc. Frederick, MD 21703, telephone 301-228-7500 or toll-free 800-946-3232, and ask for the Chief Technology Officer.

Although reasonable efforts have been made to check the papers included in the BTTJ, this publication should not
be interpreted as a representation or warranty by Bechtel Corporation of the accuracy of the information contained
in any paper, and readers should not rely on any paper for any particular application of any technology without
professional consultation as to the circumstances of that application. Similarly, the authors and Bechtel Corporation
T ELECOMMUNICATIONS disclaim any intent to endorse or disparage any particular vendors of any technology.

January 2006 • Volume 4, Number 1 iii

 Foreword

I
t is with great pleasure that we bring you this latest issue of the Bechtel Telecommunications
Technical Journal (BTTJ). The BTTJ is a compilation of expert commentaries on global operational
issues important to wireless and wireline telecommunications systems operators.
This issue leads off with an interesting historical analysis of the impact in the United States of the Bell
Laboratories divestiture in the mid-1980s. Entitled “Telecommunications Research and Development
in the United States: A State of Crisis?,” the paper is not so much operationally focused as it is a
discussion of the importance of maintaining substantial emphasis on long-term telecommunications
research for the benefit of the global society. You’ll also find a paper by Peter Moskal and
Krishnamurthy Raghu entitled “ANSI/TIA Standard 222 – Structural Standard for Antenna
Supporting Structures and Antennas: A Comparison of Revisions F and G.” This paper is a significant
technical discussion of the importance of structural and mechanical standards for antenna
support structures.
This is the second issue in which we have invited guest authors to contribute. Dr. Jungnam Yun of
the POSDATA America R&D Center in Santa Clara, California, and Professor Mohsen Kavehrad,
PhD, of The Pennsylvania State University Center for Information and Communications Technology
Research in University Park, Pennsylvania, have collaborated on a very insightful technical
discussion of “PHY/MAC Cross-Layer Issues in Mobile WiMAX.” The paper provides an overview
of mobile WiMAX, especially on OFDMA/TDD systems, and addresses issues that need to be
resolved to increase throughput, cell coverage, and spectral efficiency.
My congratulations and sincere appreciation are extended to this issue’s authors and contributors.
A special level of gratitude goes to Dr. Rasoul Safavian, vice president of Technology, Americas
Regional Business Unit, for accepting the responsibility of editor-in-chief. Dr. Safavian brings
extensive technical knowledge and industry expertise to the BTTJ.
Thank you for your interest and your continued support. We encourage and welcome your
comments regarding the topics chosen. You can submit comments and suggestions for improving
the BTTJ by visiting the Bechtel Telecommunications Web site and clicking on the BTTJ
“Contact Us” section. Past issues can be downloaded from our technology Web site at
www.bechteltelecoms.com/jsp/labs/pubs.jsp.
May your new year be safe, productive, and prosperous.

Sincerely,

Jake MacLeod
Principal Vice President and Chief Technology Officer
Bechtel Telecommunications

January 2006 • Volume 4, Number 1 v

 Editorial

W
elcome to the latest issue of the BTTJ! We focused our journal’s spotlight on the truly
exciting times in which we are living. Technical advances are occurring at a phenomenal
pace. The dot.com collapse and the telecommunications slowdown of the early 2000s have
been replaced by a burgeoning array of new offerings to put the latest news, sports, music, and games
right in the consumers’ hands.
In the arena of wireless telecommunications, 3G network deployments and R&D on even more
powerful technologies are both well underway. With the December 2005 ratification of the WiMAX
mobile standard (802.16e or 2005), vendors are energetically engaged in developing and testing new
WiMAX mobile infrastructures. The IMS promises to further revolutionize the mobile landscape by
merging the power of the Internet with the convenience of wireless services. And FMC is also picking
up significant momentum.
Significant wireline developments are also occurring. The MEMS-based optical components now
being deployed increase speed while reducing costs, power consumption, and size. The prospect of
reduced capital costs is spurring a closer look at ways to decrease FTTH waste by reconfiguring
serving areas and hub locations. Even antennas and their supporting structures are in the spotlight as
they are about to benefit from state-of-the-art practices incorporated into the latest building codes
and standards.
At the same time, I would like the readers to pay particular attention to this issue’s insightful lead
paper by Tim Statton and Jake MacLeod, which provides a thorough analysis of the state of
Telecommunications R&D in the U.S. The article puts U.S. efforts in perspective with global
R&D activities, and highlights the essential long-term R&D areas required for the benefit of the
global community.
I also encourage you to read the timely and perceptive paper contributed by our invited guest authors,
Dr. Yun and Professor Kavehrad, who put the spotlight on an overview of mobile WiMAX, especially
OFDMA/TDD systems and some of the PHY/MAC cross-layer issues, and possible resolutions
through radio resource management.
I hope you find this new issue of the BTTJ informative and useful. As always, we look forward to your
comments and contributions.

Happy reading!

Dr. S. Rasoul Safavian

Editor-in-Chief

January 2006 • Volume 4, Number 1 vii

TELECOMMUNICATIONS RESEARCH AND
DEVELOPMENT IN THE UNITED STATES:
A STATE OF CRISIS?

Issue Date: January 2006

Abstract—The United States of America has held a leadership position in telecommunications R&D for the
past several decades and has significantly contributed to the global community’s deployment of high quality
communications systems. As a result of the divestiture of the Bell System in the 1980s, Bell Laboratories—the
primary telecommunications R&D facility in the US—was divided and refocused to concentrate on producing
near-term profits for the Regional Bell Operating Companies. This paper examines the impact of the lack of a
centrally funded mechanism for telecommunications R&D in the US and the declining position of the US as a
global contributor to the telecommunications industry. Suggestions for resolving this matter are offered for
consideration to the telecommunications industry and to the federal government.

BACKGROUND Universal Service

Timothy D. Statton
tstatton@bechtel.com
Jake MacLeod
jmacleod@bechtel.com
n the 1907 American Telephone and Telegraph
I Corporation (AT&T) annual report, AT&T
president Theodore Vail promoted a position
that, given the direction of the industry and the
technology, the business of telephony would be
much more efficiently and uniformly deployed to
US citizens if telecommunications was treated as
a legally sanctioned monopoly. Vail suggested
that regulation, “provided it is independent,
intelligent, considerate, thorough and just,”
would be the most effective approach versus a
business driven by market competition [1]. In
1913, the US government accepted the
proposition and formed an agreement, entitled
the Kingsbury Commitment, wherein the Bell
System was allowed to evolve as a natural
monopoly whose focus was to deploy a uniform
telecommunications system to provide universal
service to US citizens. Strict adherence to the
technical uniformity of the network design
specifications provided a foundation for reliable
communications that was unprecedented in the
global theater. The resulting telecommunications
network contributed significantly to the economic
and social successes of the US through periods
of peacetime and periods of war and crisis.
The telecommunications network provided a
platform that was an essential element for
developing an efficient and effective commercial
and governmental environment. It formed
the basis of what is known today as the
information highway.
One objective of the development of the
telephone network in the early days was
“universal service.” Essentially, the Bell System
was charged with providing affordable, high
quality, reliable telephone service to business and
residential customers in cities and rural areas.
Universal service was partially defined in the US
Communications Act of 1934: “For the purpose of
regulating interstate and foreign commerce in
communications by wire and radio so as to make
available, so far as possible, to all of the people of
the United States, without discrimination on the
basis of race, color, religion, national origin or sex,
a rapid, efficient, nation-wide and world-wide
wire and radio communication service with
adequate facilities at reasonable charges for the
purpose of the national defense, for the purpose
of promoting safety of life and property through
the use of wire and radio communication….” [2]
The natural tendency of the corporate focus is to
serve the more profitable market sectors and
avoid the less profitable. Obviously, the more
lucrative telephone customers are located in
central business districts. However, to encourage
extension of telecommunications into rural and
residential areas, the US government allowed the
Bell System to subsidize rural and residential
sectors with the profits realized from the
business sector. The Bell System was also
provided a guaranteed rate of return on
investment on all capital expenditures involved

© 2006 Bechtel Corporation. All rights reserved. 1

 ABBREVIATIONS, ACRONYMS, AND TERMS
2.5G “second-and-a-half” generation, IPv4 Internet Protocol Version 4
a marketing designation for an
IPv6 Internet Protocol Version 6
intermediate level of digital
mobile phone service responding ITR information technology research
to demands for greater bandwidth
ITU International Telecommunication
3G third generation enhanced digital Union
mobile phone service at
laser light amplification by stimulated
broadband speeds enabling both
emission of radiation
voice and nonvoice data transfer
LSN large scale networking program
3GPP™ Third Generation Partnership
Project MIMO multiple input, multiple output
4G fourth generation, enhanced digital NIAP National Information Assurance
mobile phone service boosting data Partnership
transfer rates to 20–40 Mbps
NIST National Institute of Standards
AMPS advanced mobile phone service and Technology
ARIB Association of Radio Industries NSF National Science Foundation
and Businesses
NTIS National Technical Information
AT&T American Telephone and Service
Telegraph Corporation
OECD Organisation for Economic
ATIS Alliance for Telecommunications Co-operation and Development
Industry Solutions
OTP Office of Technology Policy
CCSA China Communications Standards
QoS quality of service
Association
R&D research and development
CD compact disk
radar radio detection and ranging
CEPT Conference of European Posts and
Telegraphs RBOC Regional Bell Operating
Company
CTO chief technology officer
RTD research and technological
CTR communications technology
development
research
SHRCWC Shanghai Research Centre for
DARPA Defense Advanced Research
Wireless Communications
Projects Agency
sonar sound navigation and ranging
DoD Department of Defense
SME subject matter expert
DSP digital signal processor
TA Technology Administration
EDGE enhanced data rates for GSM
(US Department of Commerce)
evolution
TDD time division duplex
ETSI European Telecommunications
Standards Institute TIA Telecommunications Industry
Association
EU European Union
TTA Telecommunications Technology
FDD frequency division duplex
Association
FP Framework Programme
TTC Telecommunications Technology
GPRS general packet radio service Committee
GSM global system for mobile U.S. United States Alliance for
communications A-TEAM Technology and Engineering for
Automotive Manufacturing
HSDPA high speed downlink packet access
UMTS universal mobile telephone
HSUPA high speed uplink packet access
service
IEEE Institute of Electrical and
USCAR United States Council for
Electronics Engineers
Automotive Research
IETF Internet Engineering Task Force
UTRA universal terrestrial radio access
IMT-2000 International Mobile
WCDMA wideband code division multiple
Telecommunications-2000
access

2 Bechtel Telecommunications Technical Journal

in deploying telephony networks. The more
capital equipment the Bell System purchased and
commissioned into service, the more profit the
Bell System would realize because the capital
expenditures were added to the rate base,
thereby increasing the volume of the return. The
formula worked quite well, and the telephone
network expanded rapidly and uniformly across
the nation, in cities and rural areas
simultaneously. The US led the world in
telecommunications technology as well as in
percentage of population penetration.
It was recognized in 1925 that the development of
telecommunications network technology and
equipment relied heavily on scientific research
and development [3]. Therefore, AT&T
established Bell Laboratories for the sole purpose
of advancing telecommunications industry
technologies. Bell Labs was funded with a few
cents from every telephone bill. This centrally
funded research and development (R&D)
organization was stable with regard to the source
of funding and secure in the sense that the
scientists could depend on being employed by
Bell Labs for life. Bell Labs became a magnet for
the world’s leading scientists and engineers.
The network’s continued success and
modernization were ensured by the
telecommunications innovations that emanated
from Bell Labs. Bell Labs was a brain trust, a
national treasure, wherein the nation’s best and
brightest telecommunications research scientists
explored new avenues of communications
technology. Bell Labs was intentionally isolated
from the pressure for economic performance, to
stimulate the creativity of the scientists. As a
result, Bell Labs produced an impressive list of
innovative products and discoveries:
• Transistor—1947—Revolutionized not only
telephony, but the world
• Laser (light amplification by stimulated
emission of radiation)—1958—Enabled new
multibillion-dollar industries within the
fields of medicine, communications, and
consumer electronics
• Optical communications—Advances in
photonics extending from enhancement of
optical lasers to significant improvements in
the quality and capacity of optical fiber
• Data networking—1925: first fax; late 1940s:
first remote operation of a computer—
Continued significant developments in data
transport technology
January 2006 • Volume 4, Number 1
• Digital transmission and switching—
1962—First digitally multiplexed
transmission of voice signals
• Cellular telephone technology—1945:
cellular telephone concept developed;
January 1979: Bell System Technical Journal
issued commercial design guidelines for
advanced mobile phone service (AMPS)
• Communications satellites—1962—First
orbiting communications satellite, Telestar I,
built and successfully launched by Bell Labs
• Digital signal processors (DSPs)—1979— Bell Labs was a
First single-chip digital signal processor built
brain trust, a
by Bell Labs, launching numerous
multibillion-dollar businesses. DSPs are used national treasure,
in almost every aspect of our current daily wherein the nation’s
lives, from compact disk (CD) players to cars best and brightest
and trucks.
telecommunications
• Touch tone telephone—1963—Replacement
research scientists
for the rotary dial telephone
explored new
• UNIX® Operating System and C
Language—1969 and 1972, respectively— avenues of
Made large-scale networking of diverse communications
computing systems and the Internet practical technology.
In addition, the following technologies were
either developed or significantly enhanced at
Bell Labs:
• Radar (radio detection and ranging)—
Invented in Europe but further developed
and miniaturized for use in aircraft
during WWII
• Sonar (sound navigation and ranging)—
Enhanced by Bell Labs with data
compression and noise analysis techniques
• Hearing aid—Made possible in its modern
version by use of the junction transistor
invented at Bell Labs
• Artificial larynx—Invented by Bell Labs
in 1928
• Talking movies
• Two-way television
Bell Labs has produced some of the world’s
leading scientists in several disciplines and
whose global recognition is evidenced by:
• 6 Nobel Prizes in physics shared by
11 scientists
• 9 US Medals of Science
• 7 US Medals of Technology
• 1 Draper Prize
• 6 Marconi International Fellowship Awards
3
 • 7 Computer & Communication Prizes shared
by 12 scientists and engineers
• 28 IEEE Medal of Honor winners
• 28,000 patents since 1925
Divestiture
On January 1, 1884, it was deemed necessary by
the US government, via a decision by Federal
Judge Harold Greene, that the Bell System should
be split apart and its monopolistic hold on the
telecommunications industry be released. The
Centralized federal resulting seven Regional Bell Operating
funding of focused Companies (RBOCs) had specific market
mandates for service, competition, and
telecommunications
profitability, but the role of the centralized Bell
research has not
Labs was uncertain. Bell Labs was not
been supported by accustomed to the market pressures focused on
the federal budget quarterly returns and profitability rather than
providing a platform for visionaries developing
in past years.
products for the future use of society. The RBOCs
attempted to maintain the brain trust at Bell Labs
by establishing a governing board that would
jointly direct the efforts of the scientists, but
eventually each RBOC had different business
objectives for the common pool of scientists. After
much effort and wrangling, each RBOC
withdrew its proportional funding of Bell Labs
and the scientists were integrated into the various
RBOCs to establish decentralized, RBOC-focused
laboratory facilities. This disaggregation of the
brain trust marked the end of Bell Labs as we
knew it. The core teams were required to morph
into several smaller profit-generating machines.
THE SITUATION TODAY
oday the US depends on corporate budget
T allocations as its primary source of funding
for telecommunications R&D. The problem is
that the US financial market demands near-term
performance to the detriment of long-term
Table 1. Agency Detail of Selected Interagency R&D Efforts ($ million) [6]
Networking and Information
2003 Actual
Technology R&D
National Science Foundation 743
Health and Human Services 376
Energy 308
National Aeronautics and Space
213
Administration
Defense 296
Commerce 26
Environmental Protection Agency 2 4
Total 1,964
4 Bechtel Telecommunications Technical Journal
research; therefore, the corporate funds that are
reluctantly allocated to the R&D equation are
heavily weighted to the development (zero to
18 months) versus the research (24 to 60 months)
portion of the equation. Although this formula
may be effective to produce short-term
performance, when the momentum of the
previous research efforts reaches a zero state, the
results will unfortunately be very detrimental to
the US economy.
We are at a point in history when the US is no
longer driving the technology and setting the
direction of telecommunications. If the trend
continues and no immediate action is taken,
the US will simply remain a passenger on the
bus of technology, dependent on others (notably,
Europe and the Asia Pacific countries) to take us
to destinations that may not be in the nation’s
best interest. The most significant recent telecom-
munications advancements have not originated
from US R&D laboratories, but rather from
excellent, state-of-the-art laboratories abroad. No
longer are US laboratories the primary magnet
for exceptional scientific talent; in their place, the
secure, coordinated R&D focus of technical
communities abroad is attracting the top talent.
Federal Funding in the US
Centralized federal funding of focused
telecommunications research has not been
supported by the federal budget in past years. As
can be seen in Table 1, the US government
budgeted approximately $2 billion for net-
working and information technology R&D for
fiscal years 2004 and 2005 [4]. The funding was
spread across several agencies, so the focus of the
funded research was immediately diluted. The
$2 billion represents less than 1.6 percent of the
$132 billion actually requested for R&D by all
agencies for 2005. Further, only $330 million of
the $2 billion was requested for the large-scale
Dollar Change: Percent Change:
2004 Estimate 2005 Proposed
2004 to 2005 2004 to 2005
754 761 7 1
368 371 3 1
344 354 10 3
275 259 –16 –6
252 226 –26 –10
26 33 7 27
4 0 0
2,023 2,008 -15 -1
networking (LSN) program, which includes Table 2. Budgets for the EU Framework Programmes [8]
both computing and noncomputing (telecom-
Framework Time Funding
munications) research. Less than half of the LSN
Programme Frame (€ billion)
allocation was dedicated to telecommunications
R&D. Hence, less than one-tenth of total federal First 1984 to 1987 3.75
spending on networking and information
technology R&D was allocated for telecom- Second 1987 to 1991 5.36

munications research [5].

Although the US has verbally recognized for
several years the need (e.g., September 11, 2001;
Hurricane Katrina) for reliable, secure,
interoperable, feature-rich communications, little
or no focused strategic funding mechanism has
materialized. Other countries have made
strategically focused, concerted efforts to enhance
telecommunications R&D by providing central
repositories to fund long-term initiatives.
The European Centrally Funded Framework
Model
The European Union (EU) has established a
coordinated multi-annual, multinational funding
mechanism, the Framework Programme, that
helps to organize and financially support
cooperation among universities, research centers,
and industries—including small and medium-
sized enterprises. Oversight of this program is the
responsibility of the EU’s Research Directorate
General.
The Research Directorate General’s mission is
summarized as follows:
• “to develop the European Union’s policy
in the field of research and technological
development and thereby contribute to
the international competitiveness of
European industry;
• to coordinate European research activities
with those carried out at the level of the
Member States;
• to support the Union’s policies in other fields
such as environment, health, energy, regional
development etc;
• to promote a better understanding of the role
of science in modern societies and stimulate a
public debate about research-related issues at
European level.” [7]
As shown in Table 2, the First Framework
Programme for research and technological
development (RTD) was established in 1984 with
a budget of €3.750 billion, while the current Sixth
Framework Programme covering the period
2002–2006 has a budget of €17.5 billion.
Third 1990 to 1994 6.60
Fourth 1994 to 1998 13.12
Fifth 1998 to 2002 14.96
It is clearly evident
Sixth 2002 to 2006 17.50
that the EU
Seventh 2007 to 2013 72,726 recognizes the
criticality of
For the upcoming Seventh Framework Program communications to
(FP7), entitled “ManuFUTURE Vision for 2020,” developing and
the period of funding has been extended from the
sustaining society
normal 4 years to 7. FP7’s concentrated focus over
this period will be to develop innovations that and economic
will integrate digital technologies into the stability.
manufacturing processes. Competitive advantage
in the global market place is the objective. Of the
€72,726 billion that will be allocated to FP7,
€12,670 billion (17.4 percent) is focused on
Information Society research (see Table 3).
It is clearly evident that the EU recognizes the
criticality of communications to developing and
sustaining society and economic stability. The
apparent lapse, on the other hand, in the US’s
recognition that all manner of commerce and
society rides on the back of communications
technology, has caused us to lose our
commanding position in the telecommunications
research community and will continue to
diminish our global status if we do not establish a
strategic and focused mechanism to encourage
primary research.
A partial listing of significant telecom-
munications developments that have had their
genesis in Europe or that have been heavily
influenced by the excellent technologists in the
EU includes the following:
• GSM—In 1982, the Conference of European
Posts and Telegraphs (CEPT) formed a study
group called the Groupe Spécial Mobile
(GSM) to study and develop a pan-European
public land mobile system that had to meet
certain criteria. In 1989, the responsibility for
developing what is now known as the global
system for mobile communications (which
retains the GSM acronym) was transferred
to the European Telecommunications

January 2006 • Volume 4, Number 1 5

 Table 3. Budget Breakdown for the Seventh Framework Programme 2007–2013 [9]

Amount % %
Category Element
(€ million) Category Total
Health 8,317 18.72

Food, Agriculture, and Biotechnology 2,455 5.53

Information and Communications

Technologies 12,670 28.52
Nanosciences, Nanotechnologies, Materials and
new Production Technologies 4,832 10.88
Cooperation
Energy 2,931 6.60
Environment (including Climate Change) 2,535 5.71
Transport (including Aeronautics) 5,940 13.37
Socio-economic Sciences and the Humanities 792 1.78
Security and Space 3,960 8.91
Total Cooperation 44,432 61.10
Ideas European Research Council 11,862 100 16.31
People Marie Curie Actions 7,129 100 9.80
Research Infrastructures 3,961 53
Research for the Benefit of SMEs 1,901 25
Regions of Knowledge 158 2
Capacities
Research Potential 554 7
Science in Society 554 7
Activities of International Cooperation 358 5
Total Capacities 7,486 100 10.29
Nonnuclear Actions of the Joint Research Center 1,817 2.50
Total Budget 72,726

Standards Institute (ETSI). Phase I of the
GSM specifications was published in 1990,
and commercial service was started in 1991.
Today, GSM accounts for more than
75 percent of the world’s mobile telephone
market (1.2 billion subscribers) and is used in
more than 210 countries and territories. [10]
• GPRS—General packet radio service is a
GSM-based mobile data service widely used
throughout the world. It is commonly
referred to as 2.5G technology and provides
a transmission data speed of 171 kbps
(54 kbps effective).
• 3GPP™—Third Generation Partnership
Project is a collaboration agreement that was
established in December 1998. The agreement
brings together a number of globally located
telecommunications standards bodies. These
“Organizational Partners” are the
Association of Radio Industries and
Businesses (ARIB), China Communications
Standards Association (CCSA), ETSI,
Alliance for Telecommunications Industry
Solutions (ATIS), Telecommunications
Technology Association (TTA), and Telecom-
munications Technology Committee (TTC).
The original scope of 3GPP was to produce
globally applicable technical specifications
and technical reports for a 3G mobile system
based on evolved GSM core networks and
the radio access technologies that they
support (i.e., universal terrestrial radio access
[UTRA], both frequency division duplex
[FDD] and time division duplex [TDD]
modes). The scope was subsequently
amended to include the maintenance and
development of the GSM technical
specifications and technical reports,
including evolved radio access technologies
(e.g., GPRS and enhanced data rates for
GSM evolution [EDGE]).
• UMTS—Universal mobile telephone service,
commonly referred to as 3G mobile
technology, is an advanced mobile
communications technology that provides
enhanced features and functions at a reduced
network cost. UMTS was derived out of the
International Mobile Telecommunications-
2000 (IMT-2000) and 3GPP effort at the
International Telecommunication Union
(ITU). At the end of 2004, there were in excess
of 16 million 3G UMTS subscribers in

6 Bechtel Telecommunications Technical Journal

 60 networks in 25 countries. The growth
rate of UMTS is exceeding the GSM
growth curves at the comparable stage of
market introduction.
• HSDPA—High speed downlink packet
access is a packet-based data service in the
wideband code division multiple access
(WCDMA) downlink with data transmission
speeds up to 14 Mbps (20 Mbps with multiple
input, multiple output [MIMO]) over 5 MHz
channel bandwidth. HSDPA enhances the
data transmission capabilities of UMTS and
is expected to be widely deployed in 2006.
• HSUPA—High speed uplink packet access,
like HSDPA, is an enhancement to the UMTS
communications systems and will provide
data rates that will enable full-rate video
conferencing from mobile telephones. It is a
part of the 3GPP Release 6 standards.
• IPv6—Internet Protocol Version 6 is the next-
generation IP designed by the Internet
Engineering Task Force (IETF) to replace
IPv4. IPv6 will enhance IP features and
functionality, will provide a tremendous
increase in the number of available
addresses, and will allow network
auto-configuration and quality-of-service
(QoS) levels.
Economic and Social Development, examined
and approved by the Fifth Plenary Session of the
15th Central Committee of the Communist Party
of China, states, “We shall support, encourage
and guide the healthy development of private
enterprises, especially small and medium-sized
technology-intensive companies.” The proposal
also says, “The means of planning, finance and
banking must be used comprehensively. The
leverage functions of price, tax, interest rates and
exchange rates must be allowed to play their part,
so as to guide and promote economic
restructuring and safeguard the stable growth of But in 2004,
the economy.” Additionally, the proposal noted the top 10 list,
that promoting the application of information
which included four
technology is a trend in current world economic
and social development and the key link to US companies,
upgrading China’s industrial structure and five Japanese
realizing industrialization and modernization. companies, and
China’s Ministry of Information Industry has one South Korean
started compiling the 11th Five-Year Plan for company,
2006–2010. According to the ministry, the
was marked by
following projects are to be included in the plan:
the absence of
“information technology development in the next
five to 10 years, domestic and overseas market, a domestic
third generation mobile communications, digital telecommunications
television, new display devices, software and
company.
Table 4. Top Patenting Corporations [11, 12]

Company Patents
Patents
A statistical indication highlighting the erosion of 1977–1996
thought leadership in the US is the number of General Electric Corp. 16,206
patents filed in the US Patent and Trademark
International Business Machines Corp. 15,205
Office by US companies versus foreign
Hitachi Ltd. 14,500
companies. In the past, the statistics were heavily
weighted toward US companies, including Canon Kabushiki Kaisha 13,797
telecommunications companies like AT&T and Toshiba Corp. 13,413
Motorola (see Table 4). But in 2004, the top 10 list, Mitsubishi Denki Kabushiki Kaisha 10,192
which included four US companies, five Japanese
U.S. Philips Corp. 9,943
companies, and one South Korean company,
Eastman Kodak Co. 9,729
was marked by the absence of a domestic
telecommunications company. AT&T Corp. 9,380

Motorola Inc. 9,143

China’s Long-Term, Strategic Planning 2004

China is emerging as a significant world power in International Business Machines Corp. 3,248
the realm of telecommunications R&D. Like the
Matsushita Electric Industrial Co. Ltd. 1,934
EU, China has a coordinated, long-term, strategic
Canon Kabushiki Kaisha 1,805
plan with specific, focused objectives that
are centrally funded. According to recent Hewlett-Packard Development Company, L.P. 1,775

Organisation for Economic Co-operation and Micron Technology, Inc. 1,760

Development (OECD) reports, China’s total R&D Samsung Electronics Co. Ltd. 1,604
investments lag only those of the US and Japan in
Intel Corporation 1,601
absolute terms.
Hitachi Ltd. 1,514
The Proposal of the Central Committee of the Toshiba Corporation 1,310
Chinese Communist Party for Formulating the
Sony Corporation 1,305
10th Five-Year Plan (2001–2005) for National

January 2006 • Volume 4, Number 1 7

 integrated circuits, automotive electronics,
communications laws and regulations, study of
telecom competitiveness, telecom network
interoperation, value-added service and other
issues concerning the telecom trade, network
and information security system, government
online, e-commerce application and regional
information industry development.” If sustained,
this intense focus on telecommunications
and information technologies will definitely
propel China into a leadership position in the
global telecommunications marketplace in the
China is solidly near term.
focused on the Evidence of the strength of China’s strategy is seen
long-term benefits in Ericsson’s recent press announcement by CEO
of investing in Carl-Henric Svanberg announcing that, in a
strategic move in the field of R&D cooperation,
telecommunications
Ericsson has signed the first research agreement
R&D. with Shanghai Research Centre for Wireless
Communications (SHRCWC). Under the
agreement, Ericsson will collaborate with
SHRCWC in undertaking research projects on
future telecommunications technology such as
“Super 3G” and 4G. The agreement further
demonstrates Ericsson’s commitment to drive the
development of the telecoms industry in China.
During the Summit, Mats H. Olsson, president of
Ericsson Greater China, confirmed Ericsson’s
long-term commitment to the country,
announcing a US$1 billion investment for the next
5 years [13].
China is solidly focused on the long-term benefits
of investing in telecommunications R&D. China is
not hindered by the unbalanced competitive
market pressures that unreasonably emphasize
near-term performance to the detriment of long-
term investment. In forward-focused societies,
enterprises are severely penalized if they do not
have a long-term perspective to balance short-term
performance objectives.
SUGGESTED PATH FORWARD
he Telecommunications Industry Association
T (TIA), in coordination with its Chief
Technology Officer (CTO) Council, has carefully
considered several potential options to resolve the
US’s R&D crisis and to re-establish the leadership
position of the US telecommunications R&D
community. The TIA has distilled the numerous
options down to the following four mechanisms to
address the funding problem and then has
suggested five critical areas that require technical
focus [14].
8 Bechtel Telecommunications Technical Journal
Potential Funding Mechanisms
1. Establish a National Technology Council,
whose charter would be to define and guide
strategic areas in communications that
require further research critical to the future
growth of the US economy. Such a Council
should include representation from different
sectors, such as government, academia, and
industry.
a. To use scarce financial resources effectively,
representatives from government, aca-
demia, and industry should be sought to
establish long-term priorities. Additional
research would help identify the
technologies likely to be most relevant to
US economic growth and competitiveness.
b. This Council should be modeled after the
EU’s Sixth Framework Programme
initiative, wherein the Council receives
proposals from industry consortia
regarding specific areas of focused research
and development and has available
substantial funding from the government to
help fund those proposals.
c. This Council should also borrow from the
United States Alliance for Technology and
Engineering for Automotive Manufacturing
(U.S. A-TEAM), a partnership created
between the US Department of Commerce’s
Technology Administration (TA)
(consisting of the Office of Technology
Policy [OTP], the National Institute of
Standards and Technology [NIST], and the
National Technical Information Service
[NTIS]) and the United States Council for
Automotive Research (USCAR). The U.S. A-
TEAM brings together engineers from
government and industry bodies that are
parties to the agreement to facilitate
technological research and technology
policy analysis focused on improving the
manufacturing competitiveness of the US
automotive industry.
d. The Council, in cooperation with industry,
would determine the priority of the specific
research initiatives of national concern.
2. Prioritize communications research funding
within Department of Defense (DoD) 6.1
Basic Research Programs.
a. In the 1990s, the DoD and the Defense
Advanced Research Projects Agency
(DARPA) began to rely heavily on dual use
and industry research funding. Thus,
DoD funding became unavailable for
 commercially available technologies. As a
result, DoD restricted its research funding
to military-unique needs, which was
acceptable at the time because private-
sector-led research was driving high end
research.
b. With the communications downturn, how-
ever, the commercial sector has ceased to be
the major driver of high-end, long-term
research. As a result, DoD and DARPA
need to increase their focus on and
investment in dual-use technologies.
3. Prioritize communications research funding
within the National Institute of Standards
and Technology.
a. Miniaturization of electronic components in
communications devices continues,
resulting in faster, more powerful, more
reliable products. Yet, the continued
shrinking of component parts at the
nanoscale is hindered by metrology and
manufacturing challenges. NIST programs
address some of these key issues and
should be adequately funded.
b. Additionally, the TIA supports the
continuation of the National Information
Assurance Partnership (NIAP), a collab-
oration between NIST and the National
Security Agency. The long-term goal of
NIAP is to help increase the level of trust
consumers have in their information
systems and networks through the use of
cost-effective security testing, evaluation,
and validation programs.
4. Prioritize communications research funding
within National Science Foundation (NSF)
research programs.
a. Federal funding for physical sciences
research, the foundation of our nation’s
economic competitiveness, has
dramatically decreased. Technological
advances driving the economy require the
reversal of this trend.
b. The National Science Foundation Autho-
rization Act of 2002 called for doubling the
NSF budget over 6 years; fulfillment of that
goal is lagging.
c. In conjunction with increasing NSF’s
budget, the TIA advocates for the creation
of an NSF communications technology
research (CTR) program similar to the
information technology research (ITR)
program that recently concluded. Such a
program would greatly benefit the
January 2006 • Volume 4, Number 1
communications sector by creating
opportunities at the frontiers of
communications research and education.
Recommended Areas of Research Focus
The following TIA-recommended areas of focused
research closely parallel the technology research
budget recommendations by the Office of the
President of the United States [15]:
1. Universal Broadband—Affordable broad-
band access and connectivity, using all
available media (copper, coax, fiber,
spectrum, etc.), carrying all services (voice,
data, video) to all customers everywhere
(urban, suburban, rural, mobile) to enable a
greatly upgraded “superhighway.”
a. Broadband Internet access is critical to
support technology convergence and
advanced communications. A forward-
looking US government should support
universal access for broadband Internet, as
well as policies that promote widespread
connectivity. Infrastructure upgrades create
increasing returns to our economy and
encourage the development of busi-
nesses, entertainment, education, and
e-government solutions and capabilities.
b. Additional federally funded research in this
field is needed, particularly because special
technologies will be needed for rural access,
and corporate and venture capital financing
for research has dropped significantly over
the last several years. Extremely significant
cost reductions are necessary to meet the
technology needs of rural areas.
Additionally, the provision of broadband
access in rural areas is costly due to
challenges associated with terrain, low
population density, etc.
2. Security—New authentication, encryption,
and monitoring capabilities for all public
broadband networks to protect communi-
cations assets from attack.
a. The US is a post-industrial information
society and, as such, has a cyber-
infrastructure that is vulnerable to attack.
b. Continued research is needed to prevent
systemic attacks to infrastructure and may
provide an opportunity for university-
based “centers of excellence.”
3. Interoperable Mobility—The ability to access
commercial mobile services and emergency
services over any mobile network from any
mobile instrument.
9
 a. Interoperable mobility enables public safety
and law enforcement officials to use the
various public safety and cellular mobile
networks while avoiding the necessity of
carrying multiple mobile devices. It also
promotes coordinated communications
among various public service agencies and
allows higher priority use of scarce
spectrum resources for emergency use.
b. Federally funded research is necessary
because the emergency services market is
critical for the common good. Also, bringing
It is time for the US commercial technologies and emergency
to stop eating the services technologies closer together will
seed corn and to result in lower costs and more advanced
features for critical emergency services.
begin to plant for
future harvest. 4. Communications Research for Homeland
Security, including interoperability, security,
survivability, and encryption.
a. Homeland Security is a superset of several
other visions. Security technologies can
help protect public networks and other
public infrastructure from malicious
attacks. A large amount of economic
activity today depends on the continued
availability of public broadband networks
and infrastructure. Successful attacks can
slow national economic activity
significantly and can have other disastrous
consequences (e.g., in case of identity theft).
b. Research is needed in all areas
(interoperability, security, survivability,
and encryption) because the needs of first
responders and critical infrastructure
protection far exceed the needs of “typical”
commercial applications. Further research
is also needed because new worms and
viruses are constantly being invented, and
new techniques are needed to prevent
attacks before there is significant resulting
damage.
c. The country needs a broad program to
address our vulnerabilities and ensure the
integrity of first responder systems. The
government should support these “extreme
case” applications, since they are unlikely to
be sufficiently developed in normal
commercial systems.
5. Nanotechnology
a. Many communications advances have been
driven by fundamental scientific
discoveries of nanoscale materials.
b. Examples of important research areas
include sensors, displays, power systems,
radio frequency, and nanomicrophones.
10
c. Advances will reduce cost; increase
mobility; decrease power consumption; and
improve healthcare, homeland security,
and public safety.
CONCLUSIONS
he US has recently experienced the migration
T of manufacturing services to regions of the
world where total costs of manufacturing are
much reduced. Now the US is witnessing
a similar migration of the communications
research initiatives to Europe and Asia Pacific.
Pre-competitive communications research is
migrating to laboratories abroad for three
basic reasons:
• The total cost of conducting communications
research is lower in emerging countries.
• The financial pressure on public companies
to generate near-term profit severely
limits the allocation of budget to pre-
competitive research.
• There is a lack of stable, centrally funded and
focused communications research programs
in the US.
The US began to lose its research leadership
position in communications with the dissolution
of Bell Laboratories in the mid-1980s when market
forces and the pressure on delivery of short-term
earnings replaced “cost-of-service” rate-based
monopolies. The erosion of the research leadership
position has continued and will continue at an
accelerated rate unless immediate and decisive
action is taken. In the near to medium term, the US
will likely remain the dominant consumer of
externally developed and manufactured goods,
but the voice of the US will dwindle to a whisper
regarding its influence on the product
development and evolution. The US must choose
its areas of emphasis and importance. Consumer-
focused product development and manufacture is
an area unlikely to hold promise for revitalization
in the US. The US should, however, consider
funding pre-competitive research on commu-
nications products and technologies that involve
national security (including cyber security),
personal health, and safety and disaster recovery.
The suggestions offered by the TIA CTO Council
address exactly those issues.
The aforementioned suggestions are offered for
serious consideration and discussion . . . and
action. It is time for the US to stop eating the seed
corn and to begin to plant for future harvest.

Bechtel Telecommunications Technical Journal
TRADEMARKS
3GPP is a trademark of the European
Telecommunications Standards Institute (ETSI) in
France and other jurisdictions.
UNIX is a registered trademark of The Open
Group.
REFERENCES
[1] A Brief History: The Bell System
(http://www.att.com/history/history3. html).
[2] Communications Act of 1934, Title 1, Section 1
(http://www.fcc.gov/reports/1934new.pdf).
[3] Capsule History of the Bell System
(http://www.bellsystemmemorial.com/
capsule_bell_system.html).
[4] Office of Science and Technology Policy,
Executive Office of the President, 2005 Budget
Summary (http://www.ostp.gov/html/
budget/2005/ap05.pdf).
[5] National Coordination Office for Networking
and Information Technology Research and
Development (http://www.itrd.gov/pubs/
2006supplement/2006supplement.pdf).
[6] Office of Science and Technology Policy,
Executive Office of the President, 2005 Budget
Summary, page 62 (http://www.ostp.gov/
html/budget/2005/ap05.pdf).
[7] European Commission, Research Directorate-
General (http://europa.eu.int/comm/dgs/
research/index_en.html).
[8] European Commission Research Presentation,
Towards the Sixth Framework Programme,
page 9 of 62 (http://www.lnl.infn.it/pages/
Conferenze/Ferrini.pdf).
[9] Budget Breakdown for Seventh Framework
Programme (http://www.cordis.lu/fp7/
breakdown.htm).
[10] GSM World; GSM Facts and Figures
(http://www.gsmworld.com/news/
statistics/index.shtml).
[11] Top Patenting Corporations
(http://www.nsf.gov/sbe/srs/seind04/
c6/tt0603.htm).
[12] US Patent and Trademark Office Calendar Year
2004, Preliminary List of Top Patenting
Organizations (http://www.uspto.gov/web/
offices/ac/ido/oeip/taf/top04cos.htm).
[13] Ericsson Press Release, September 7, 2005
(http://www.3g.co.uk/PR/Sept2005/1867.htm).
[14] Telecommunications Industry Association White
Paper, Investing in Communications for Tomorrow’s
Innovations: The Case for Increased Communications
Research Funding, September 2005.
[15] Executive Office of the President – Office of
Science and Technology Policy, Memorandum for
the Heads of Executive Departments and Agencies, re:
FY 2007 Administration Research and Development
Budget Priorities, July 8, 2005.
January 2006 • Volume 4, Number 1
ADDITIONAL READING
dditional information sources used to
A develop this paper include:
• The Task Force on The Future of American
Innovation, The Knowledge Economy: Is the United
States Losing Its Competitive Edge?, February 16, 2005.
• N. Gehani, Bell Labs, Life in the Crown Jewel,
Silicon Press, 2003, ISBN 0 929306-27-9.
BIOGRAPHIES
Tim Statton is president of
Bechtel Telecommunications
and an executive vice president
and member of the Board of
Directors of Bechtel Group, Inc.
Before his present appointment,
Tim was president of Bechtel
Enterprises (BEn), the company’s
project development, financing,
and ownership subsidiary.
Before joining BEn, Tim was a member of the
Chairman’s Leadership Council and president of
Bechtel Energy, leading all energy-related engineering
and construction activities within the company. Before
this, he was managing director for energy and water
development at BEn. He also served as manager of Asia
Pacific operations, representing Bechtel’s engineering
and construction interests in the region. During his
extensive career in the power sector of the company,
Tim held numerous field and home office positions, was
a project manager and business development manager,
and was eventually elevated to president of Bechtel’s
power and industrial company.
Tim has been a member of the Board of the United
States Energy Association (USEA), the leading energy
trade organization in the US, and has volunteered for
numerous charitable and community activities. He also
served as chairman of the Board of Nexant, an energy
consulting business, and was a member of the Board
of Control of InterGen, an independent power
development company. He currently sits on the United
States Telecommunications Training Institute (USTTI)
Board of Directors.
Tim received his BS in Mechanical Engineering from
San Francisco State University, California, and his
BS in Business/Economics from Juniata College,
Pennsylvania. He is a member of the American
Cogeneration Society and the U.S. Energy Association.
Jake MacLeod is the chief
technology officer, Engineering
and Technology, for Bechtel
Telecommunications and a
Bechtel principal vice president.
Jake joined Bechtel in
May 2000 and is responsible
for expanding the scope of
Bechtel’s telecommunications
engineering services to include
all aspects of technical design, from network planning
to commercial system optimization. Jake initiated and
developed Bechtel’s RF and Network Planning
team, which has grown to over 150 world-class
11
 engineers. He also designed and established Bechtel’s
two world-class telecommunications laboratories to
provide clients with applied research and development
services ranging from interoperability testing to
product characterization.
Jake was the first Bechtel Telecommunications person to
enter Baghdad in 2003, immediately after the conflict
paused. He and his teams assessed the Iraqi
telecommunications network, then designed and
replaced 12 wire centers (equivalent to 240,000 POTS
lines) in a period of 4 months, an unprecedented
achievement in telephony. Jake and his teams also
analyzed and replaced the air traffic control system at
Baghdad International Airport. Under Jake’s purview,
Bechtel’s technology teams are currently in the final
stages of developing the Virtual Survey Tool, an
automated network planning tool that has the potential
to radically change the conventional methods of
network design. Jake’s laboratories are currently
working with two global wireless equipment
manufacturers to analyze and characterize UMTS,
HSDPA, Node B hotels, WiMAX, and intuitive
networks. Under Jake’s direction, the laboratories
annually produce two issues of the Bechtel
Telecommunications Technical Journal, an authoritative
technical publication focused on operational matters.
The laboratories also host semi-annual global
technology debates focused on the pros and cons of
the most advanced telecommunications technologies.
Jake, himself, provides an average of six to eight
keynote and technology-based presentations per year at
industry conferences.
12
Jake started his career in the telecommunications
industry in 1978, beginning in transmission engineering
with Southwestern Bell Telephone Company (SWBTC)
in San Antonio, Texas. His responsibilities at SWBTC
included design and implementation of radio systems
in Texas west of Ft. Worth. He participated in the
original cellular telephone system designs for SWBTC
in San Antonio, Dallas, and Austin. After SWBTC, Jake
became the second employee for PageNet/CellNet and
vice president of Engineering for their cellular division.
He designed over 135 cellular network systems,
including San Francisco’s, and filed them with the FCC.
In addition to his responsibilities at PageNet/CellNet,
Jake was asked to chair the FCC’s Operational
Relationships Committee.
Jake has held executive management positions with
NovAtel (Calgary), NovAtel (Atlanta), Western
Communications, and West Central Cellular. More
recently, Jake spent 9 years with Hughes Network
Systems (HNS), where he was instrumental in
establishing their cellular division. He designed and
established cellular and WLL systems in areas
ranging from central Russia to Indonesia, as well as
in 57 US markets.
Jake holds a BS degree from the University of Texas
in Austin.
Bechtel Telecommunications Technical Journal
IP MULTIMEDIA SUBSYSTEMS (IMS):
A STANDARDIZED APPROACH TO
ALL-IP CONVERGED NETWORKS

Issue Date: January 2006

Abstract—The IMS is a standardized approach to offering Internet services anywhere at any time using
cellular technology. The first release of the IMS is tailored for GPRS/UMTS, whereas later releases will allow
access independence, including WLANs and even fixed networks such as xDSL and cable modem.
The IMS is based on open interfaces and common elements that make it possible to provide integrated, secure,
IP-based, multimedia, multisession applications to mobile and fixed users, with guaranteed end-to-end QoS and
roaming capabilities, while providing the same services as the home network, and across different access
technologies. The IMS also allows a single sign-on authentication and provides a flexible architecture for
charging and billing. Deployment of the first version of IMS products is expected in early 2006.

INTRODUCTION
n recent years, both the Internet and wireless
I communications have experienced a surge of
activities and successes. The success of the
Internet stems from two main facts: It provides
useful applications such as e-mail, World Wide
Web (www), and instant messaging (IM), and it
uses readily available open protocols that
promote and facilitate the development of
various services and applications. The success of
mobile communications is self-evident. This
technology has experienced explosive growth
and provides wide coverage, touching almost
every aspect of peoples’ lives. Currently, there are
more than 1 billion mobile customers, and this
number is growing!
Wireless systems dubbed as beyond third-
generation (3G+) seek to merge these two
successful communications modes to provide the
power of the Internet—and all of its services and
applications—with the convenience of ubiquitous
wireless access. The new architecture expected to
accomplish this, the Internet Protocol (IP)
multimedia subsystem (IMS), is defined in a
standard created by the Third Generation
Partnership Project (3GPP™), one of the main
standards organizations for 3G networks.
Different aspects of the IMS (call control,
Rasoul
Safavian, PhD charging, roaming, etc.) have been formalized
and published in various 3GPP technical
srsafavi@bechtel.com
specifications [1–15]. The first version of the IMS
was published by the 3GPP in Release 5 of the
The 3GPP, formed in December 1998, is a
collaboration of telecommunications industry
organizations to produce globally applicable
technical specifications for 3G mobile
systems and to maintain and develop GSM
standards. Participants include major regional
telecommunications standards bodies such
as ARIB of Japan, CSCA in China, Committee
T1 of the US, and TTA of Korea, as well as
market representatives such as the UMTS
forum, 3G America, GSM associations, and
the IPv6 forum.
universal mobile telecommunications system
(UMTS). Release 6 of the UMTS, introduced in
March 2005, provided some enhancements to
the first release and introduced new concepts,
such as support for access independence, wireless
local area network (WLAN) integration, and
IM and presence services. Release 7, under
development with an expected release in mid-
2007, will focus primarily on fixed and mobile
convergence issues.
The Third Generation Partnership Project 2
(3GPP2), the main standardization body for 3G
networks based on cdma2000®, has established
the multimedia domain (MMD), its own version
of the IMS specification. The IMS and the MMD
are very similar, with the main differences
centering on two issues:

© 2006 Bechtel Corporation. All rights reserved. 13

 ABBREVIATIONS, ACRONYMS, AND TERMS
2G second generation, the original COPS common open policy service
digital mobile phone service CS circuit switched
2.5G “second-and-a-half” generation, CSCF call SCF
a marketing designation for an
intermediate level of digital CSEQ context sequence (number)
mobile phone service responding DEC decision
to demands for greater bandwidth
Diameter A protocol that provides a
3G third generation, enhanced digital framework for any services
mobile phone service at requiring AAA/policy support
broadband speeds enabling both across many networks and that
voice and nonvoice data transfer primarily supports mobile IP,
3G+ beyond 3G, envisioned as all- accounting, network access,
digital, entirely packet-switched and strong security
radio networks involving hybrid DiffServ differentiated services
networking and access technologies
that globally integrate services and DNS domain name system
technology while providing the DSCP differentiated services code point
multiple QoS of an ATM network
ECF event charging function
and the flexibility of an IP network
FTP file transfer protocol
AAA authentication, authorization, and
accounting GGSN gateway GPRS support node
ACA accounting answer GPRS general packet radio service
ACR accounting request GSM/gsm global system for mobile
communication
AoR address of record
GTP GPRS tunneling protocol
API application program interface
HLR home location register
AS application server
HSS home subscriber server
ATM asynchronous transfer mode
HTTP hypertext transport protocol
AUC authentication center
I-CSCF interrogating CSCF
AUTN authentication token
Ik integrity key
AV authentication vector
IKE Internet exchange key
BGCF breakout gateway control function
IM instant messaging
BICC bearer independent call control
IMS IP multimedia subsystem
CAMEL customized application for mobile
network enhanced logic IM-SSF IP multimedia service switching
function
CAP CAMEL application part
IntServ integrated services
capex capital expense
IP Internet Protocol
CCA credit control answer
IP-CAN IP connectivity access network
CCC credit control client
IPSec IP security
CCF charging collection function
IPv4 Internet Protocol version 4
CCR credit control request
IPv6 Internet Protocol version 6
CDF charging data function
ISC IMS service control (interface)
cdma2000® A family of standards, developed
through comprehensive proposals ISDN integrated services digital network
from Qualcomm, describing the ISIM IMS subscriber identity module
use of code division multiple
access technology to meet 3G ISUP ISDN user part
requirements for wireless IWF interworking function
communication systems
MAA multimedia authentication answer
CDR charging data record
MAP mobile application part
CGF charging gateway function
MAR multimedia authentication request
Ck cipher key
Megaco media gateway control

14 Bechtel Telecommunications Technical Journal

MGCF media gateway control function SCS service capability server
MGW media gateway S-CSCF serving CSCF
MIME multipurpose Internet mail extension SCTP streaming control transmission
MMD multimedia domain protocol

MRF multimedia resource function SDP session description protocol

MRFC MRF control SEG security gateway

MRFP MRF processor SGSN serving GPRS support node

MSC mobile switching center SGW signaling gateway

MTP message transfer part SIM subscriber identity module

NAI network access identifier SIP session initiation protocol

OCF online charging function SLF subscriber location function

opex operating expense SMS short message service

OSA open service access SMTP simple mail transfer protocol

OTA over the air SS7 Signaling System Number 7, a

common channel signaling system
PCM pulse code modulation defined by the ITU and used to
PCS personal communication system provide a suite of protocols that
enables circuit and noncircuit-
P-CSCF proxy CSCF related information to be
PDA personal digital assistant routed about and between
telecommunications networks
PDF policy decision function (same as
policy decision point [PDP]) SSP subscriber service profile
PDG packet data gateway TCP transmission control protocol
PDP packet data protocol; also: policy THIG topology hiding internetwork
decision point (same as PDF) gateway
PEP policy enforcement point TLS transport layer security
PHB per-hop behavior TUP telephone user part
PLMN public land mobile network UA user agent
PS packet switched UAA user authorization answer
PSTN public switched telephone network UAC UA client
PTT push to talk UAR user authorization request
QoS quality of service UAS UA server
RADIUS remote authentication dial-in user UDP user datagram protocol
service UICC UMTS integrated circuit card
RAND random number UMTS universal mobile
REQ request telecommunications system
RES response URI uniform resource identifier
RFC Request for Comments URL uniform resource locator
RSVP resource reservation protocol USIM UMTS subscriber identity module
RTCP real-time transport control protocol VoIP voice over IP
RTP real-time transport protocol WiMAX worldwide interoperability for
microwave access
RTSP real-time streaming protocol
WLAN wireless local area network
SA security association
www World Wide Web
SAA server assignment answer
xDSL term used for all forms of
SAR server assignment request
technology using a digital
SBLP service-based local policy subscriber line
SCF session control function XRES expected response
SCP service control point

January 2006 • Volume 4, Number 1 15

 • Mobility. The 3GPP2’s MMD is built on top of
mobile IP, whereas the 3GPP’s IMS manages
mobility through general packet radio
service (GPRS), which provides a Layer 2
tunneling mechanism.
• IP versions supported. The MMD supports
both IPv4 and IPv6, whereas the IMS initially
supported only IPv6. However, due to
pressure from telecommunications operators,
newer versions of the IMS also support both
IPv4 and IPv6.
The need to deploy
a new domain ORGANIZATIONS MENTIONED IN THIS PAPER
may naturally be 3GPP™ Third Generation Partnership
Project, formed in December
questioned, 1998 as a collaboration
especially at a time agreement bringing together a
number of telecommunication
when network
standards bodies to produce
operators are globally applicable technical
struggling with costs specifications for 3G mobile
systems and to maintain and
of deploying 3G
develop GSM standards
networks and are
3GPP2 A sister project to 3GPP that is a
also facing reduced collaboration agreement dealing
voice revenues. with North American and Asian
interests regarding 3G mobile
networks based on cdma2000
ARIB Association of Radio Industries
and Businesses (Japan)
CSCA Consortium on Standards and
Conformity Assessment (China)
IETF Internet Engineering Task Force
ITU International
Telecommunications Union
TTA Telecommunications
Technology Association (Korea)
This paper focuses on the 3GPP’s IMS. The
discussion opens with an overview of IMS and
the advantages it holds for network operators.
Since an IMS network is basically an advanced
mobile session initiation protocol (SIP) network,
the paper then explores SIP networks and their
main features and functionalities. Next, IMS basic
architecture, components, interfaces, etc., are
examined. The discussion then turns to IMS
functionalities and operations issues, such as IMS
prerequisites, registration, session setup,
roaming, security, quality of service (QoS),
and charging. Finally, key benefits offered
by the IMS approach are summarized in the
concluding remarks.
16
OVERVIEW OF THE IMS AND WHY IT IS NEEDED
s defined by the 3GPP, the IMS is a new
A subsystem that enables convergence of data,
speech, and mobile network technology over an
IP-based infrastructure. It provides an integrated
service control platform that allows the creation
of multimedia and multisession applications
using wireless (and recently wireline) transport
capabilities. It is a combination of new network
elements and interfaces, i.e., a new core network
domain, which creates a new service delivery
environment.
The need to deploy a new domain may naturally
be questioned, especially at a time when network
operators are struggling with costs of deploying
3G networks and are also facing reduced voice
revenues. After all, second-generation (2G)
terminals can act as modems to transmit IP
packets over a circuit, and 2G and 3G terminals
can use native packet-switched (PS) technology to
perform data communications. To evaluate the
need to deploy the IMS, the following discussion
examines conventional network domains, the
services they can offer, and how they handle
various new service offerings. The end users’
experiences are also considered.
Shortcomings of Conventional Network Domains
The circuit switched (CS) domain is used by
traditional CS networks, which offer simple user-
to-user voice services or short-message services
(SMSs). Even though voice services were, and
perhaps to some degree will continue to be, a
major source of revenue for network operators,
voice revenues have been dropping in the last
few years, mainly due to increased competition
for existing subscribers.
Operators have been looking for ways to
differentiate themselves by offering new and
creative data services. The PS domain of PS
networks or basically 2.5 and 3G networks has
helped PS network operators to introduce user-to-
server data services, where a user directly
addresses a specific server to execute the service
in question; furthermore, these services take
advantage of IP transport and provide “always
on” connectivity. Unfortunately, PS networks
have not been very successful, perhaps due to
insufficient bandwidth, lack of enticing
applications, confusing charging schemes, long
delays in service offerings, etc. Furthermore,
increasing the bandwidth alone may not prove
sufficient to enable the plethora of new and
desirable services that customers may demand.
Bechtel Telecommunications Technical Journal
Also, network operators, in a rush to offer nor provides synergy among the applications.
data services, deployed specialized isolated Furthermore, use of this traditional approach
island solutions that often did not integrate well increases not only capital expense (capex), but
with the other services. These dedicated solutions also operating expense (opex), since each
are typically proprietary and use dedicated application may have different operating and
components and interfaces that cannot be used maintenance requirements [16–18].
for other applications. This is particularly true
Thus, a new platform or domain is needed with
when the applications are provided by different
unified features; common elements; and open,
vendors. The island solutions also typically do
standardized interfaces that can be used by all
not support roaming. While the island approach,
existing and future applications and services. The
also known as the vertical application platform,
IMS—the domain of services—meets this need.
may be acceptable for deploying small-scale data
Figure 1 shows the position of the three domains
applications or services, it neither scales well in Thus, a new
(CS, PS, and IMS) in a mobile network [17].
deploying many diverse services or applications, platform or domain
is needed with
unified features;
common elements;
and open,
standardized
interfaces that can
be used by all
existing and future
applications and
services. The IMS—
the domain of
services—meets
this need.

Figure 1. CS, PS, and IMS Domains in a Mobile Network (After [16])

January 2006 • Volume 4, Number 1 17

 Desirable Characteristics of the IMS
The desirable features, requirements, and
architecture for this new domain are delineated in
the 3GPP’s technical specification for IMS service
requirements [1, 2].
The IMS domain supports the following key
requirements:
• IP multimedia sessions, i.e., delivery of
multimedia sessions over PS networks
• A mechanism to negotiate and enforce QoS
• Integration with Internet and CS networks
such as public switched telephone networks
(PSTNs) and existing cellular networks
• Full roaming capabilities
• Single sign-on and authentication
• Single converged billing
• Strong operator controls with respect to
services delivered to end users
• Rapid service creation without requiring
standardization
• Access independence, i.e., allows access
technologies other than GPRS and UMTS
(e.g., WLANs and x-type digital subscriber
line technology [xDSL])
The IMS also provides improved end-user
experience over that offered by the other two
domains. CS domain users can access only one
type of service per bearer or session. And while
multiple, parallel IP sessions may be available to
PS domain users, bandwidth, QoS, and charging
issues may hinder effectiveness. Implementation
of the IMS infrastructure will enable three general
types of services to be offered:
• User-to-user
• Multiuser
• Server-to-user
The user-to-user services enabled by IMS extend
beyond today’s simple voice call or SMS text
messaging to include services such as voice over
IP (VoIP), video telephony, chat sessions, and
push-to-talk (PTT). The IMS multiuser services
include one-to-many and many-to-one services
such as multimedia conferencing, group chat
sessions, and multiuser PTT services. These
services are enabled in the IMS via a dedicated
media server, the multimedia resource function
(MRF), which is discussed in the section of this
paper on IMS architecture. Finally, deploying
IMS enables operators to introduce many
innovative server-to-user services by virtue of
(1) its ability to locate users within the network,
18
i.e., its mobility management; (2) its signaling
abilities allowing servers to act as user agents
(UAs), i.e., initiating and receiving SIP messages;
and (3) its ability to integrate SIP-based enabling
services such as IM and presence. Server-to-user
services include click-to-dial, dynamic push
services, etc., and could be based on presence,
status, geographic location, device type and
capabilities, media preference, etc. To fully
appreciate these capabilities, a more thorough
understanding of SIP and SIP networks is needed.
Therefore, a discussion of SIP networks has been
provided in the next section of this paper.
Using IMS, end users also have much more
control over the services than they do under the
CS and PS domains. For instance, an end user can
initiate multiple services within a single session,
also known as dynamic media control; add or
modify some components of a session (e.g., by
adding a video component to an ongoing voice
call); or add or drop a user during an ongoing
conference call or chat session. In summary, IMS,
while not designed to create new services, offers
all the services, current and future, that the
Internet provides.
Protocols Used in the IMS
In any communications network, protocols used
fall into two basic categories: signaling and
control plane, and media or user plane. The IMS
is built based on IP protocols, and signaling and
control protocols for session initiation and control
are based on the SIP (Request for Comments,
[RFC] 3261 [19]) and session description protocol
(SDP) (RFC 2327 [20]). To transport IMS signaling
protocols, the reliable streaming control trans-
mission protocol (SCTP) or transmission control
protocol (TCP) is used. Media plane protocols
used for media delivery are based on the real-
time transport protocol/ real-time control
protocol (RTP/RTCP) (RFC 3550 [21]) for
transporting real-time media such as audio or
video. Near-real-time streaming media are
transported using the real-time streaming
protocol (RTSP). Both RTP/RTCP and RTSP
typically use the user datagram protocol
(UDP) as the transport protocol to avoid TCP’s
setup, teardown, and retransmission delays.
Non-real-time media are delivered using hyper-
text transport protocol (HTTP) (RFC 2616),
simple mail transport protocol (SMTP), or
file transfer protocol (FTP), with TCP as the
transport protocol.
IP protocols are developed mainly by the IETF
and the ITU and are published by their respective
organizations as RFCs and ITU-T documents.
Bechtel Telecommunications Technical Journal
Other protocols used in the IMS include:
• CS call control
— Telephony user part (TUP) (ITU-T Q.721)
— Integrated services digital network
(ISDN) signaling user part (ISUP) (ITU-T
Q.761)
— Bearer independent call control (BICC)
(ITU-T Q.1901). An evolution of ISUP,
BICC completely separates the signaling
plane from the media plane and can run
over asynchronous transfer mode (ATM),
Signaling System Number 7 (SS7), or IP.
• Multimedia session establishment and
control
— Packet-based multimedia communica-
tions systems ITU-T H.323. Unlike BICC,
this protocol is designed from scratch to
support IP technologies in establishing
multimedia sessions.
— Media gateway control (Megaco) (ITU-T
H.248). This protocol is used to control
the IMS media gateway (MGW).
• Authentication, authorization, and
accounting (AAA)
— Diameter (RFC 3588 [22]). An evolution
of remote authentication dial-in user
service (RADIUS) (RFC 2866 [23]),
Diameter:
– Is used by the network and the user
to authenticate and authorize each
other
– Has a base protocol, complemented
by so-called Diameter applications
that are customized extensions to the
base Diameter to suit a particular
application in a given environment
– Interacts with SIP during session
setup in one application, performs
credit control accounting in another
application, etc.
• Policy and QoS control
— Common open policy service (COPS)
(RFC 2748 [24]). This is a request/
response protocol used between the
policy server (the policy decision
function [PDF], also known as the policy
decision point [PDP]) and the policy
client (the policy enforcement point
[PEP]). COPS supports two modes:
outsourcing and configurable or
provisioning. In the outsourcing mode,
the PEP contacts the PDF each time a
January 2006 • Volume 4, Number 1
policy decision is needed. The PDF
makes the decision and communicates
this information to the PEP for
enforcement. In the configurable or
provisioning mode, the PDF configures
the PEP with the enforcement policy,
which the PEP stores and uses for
current and future decisionmaking.
Since the main protocols used for session
initiation, description, control, modification, and
termination in IP networks are SIP and SDP, the
IMS is considered basically an advanced mobile
SIP network. The essentials of SIP/SDP and SIP
In short, SIP
networks are examined next. supports the basic
aspects of the
multimedia session:
SIP NETWORKS
user location,
Purpose of SIP user availability,
SIP is a general-purpose application-layer user capabilities,
protocol designed to establish, modify, and session negotiation,
terminate multimedia sessions in IP networks [25,
and session
26]. It also allows other participants to be invited
to ongoing sessions. The main goal of SIP is to management.
deliver a session description to a user at the user’s
current location. Once the user has been located
and the initial session description delivered, SIP
can deliver new session descriptions to modify
the characteristics of the ongoing session or to
terminate the session. In short, SIP supports the
basic aspects of the multimedia session: user
location, user availability, user capabilities,
session negotiation, and session management.
Session Descriptions
A session description contains enough
information for a remote user to be able to
establish, join, modify, or terminate a session. A
session description could include information
such as the IP addresses and port numbers to
which the media services need to be sent and the
coder-decoders (codecs) used to encode the voice,
image, and video elements. SIP uses SDP, the
most common format to describe a multimedia
session. SDP has a text-based format and consists
of two basic parts: session-level information and
media-level information.
Even though SIP uses SDP to transport the
session description, SIP is completely
independent of the format of the objects it
transports. Objects that SIP transports could be
session descriptions written in formats other than
SDP, or any other piece of information.
SIP is a two-way session description exchange
protocol, also known as the offer/answer or
request/response transaction model. Since SIP is
19
 based on HTTP, it is text-based, which makes SIP
easier to debug and to extend, but less efficient
in terms of size and required time/bandwidth
to transmit.
SIP Transactions
A SIP transaction consists of a request from a
client or user agent (UA), usually referred to as a
UA client (UAC); zero or more provisional
responses; and a final response from a server,
usually referred to as a UA server (UAS). SIP
messages begin with a start line, a header field,
SIP uses a SIP URI and a message body [26].
(RFC 3261) to
Start Line
identify users. A SIP The start line is also called a request line in the
URI is similar to an request message and a response line in the
e-mail address and response message. It consists of a method name, the
consists of request-uniform resource identifier (URI), and the
protocol version, currently SIP/2.0. The method
the user’s name name indicates the purpose of the request, and
and the domain the request-URI contains the destination address
name of the home of the request. The response line consists of the
protocol version (e.g., SIP/2.0) and the status of the
network operator.
transaction, including both a number and its
equivalent readable phrase (e.g., 100 TRYING). A
sample SIP transaction is shown in Figure 2. In
this transaction, (1) a request is made by a UAC
(Bob) to establish a session with a UAS (Alice) via
an INVITE request (method), and (2) the reply
from the UAS (Alice) is 100 TRYING, confirming
that Bob’s request has been received.
Some of the major SIP methods (request types)
are listed in Table 1.
Header Field
Following the start line is the header field, which
is composed of a mandatory part and an optional
part. The mandatory part contains fields such as
To, which carries the destination URI; From,
which carries the originator’s URI; and Content
Sequence (CSEQ), which includes the sequence
number and method name used to match replies
to requests.
Figure 2. SIP Transaction: (1) INVITE Request Message and
(2) 100 TRYING Reply Message
20
Table 1. Sample SIP Methods
Method Description
INVITE Establishes a session
ACK Acknowledges the establishment of the session
CANCEL Cancels a pending request
REGISTER Maps a public URI with the user’s current location
UPDATE Modifies characteristics of a session
MESSAGE Carries an instant message
SUBSCRIBE Subscribes to an event
NOTIFY Notifies an IMS terminal about a certain event
BYE Terminates the session
Message Body
The message body carries multipurpose Internet
mail extension (MIME) encoded messages.
Message bodies starting with Content-Disposition
are session descriptions. Message bodies are sent
in their entirety, i.e., they are not parsed at proxy
servers in between and may be encrypted by
the UAC.
User Identification
SIP uses a SIP URI (RFC 3261) to identify users.
A SIP URI is similar to an e-mail mail address and
consists of the user’s name and the domain name
of the home network operator. It may also include
optional descriptions placed after a semicolon.
An example of a SIP URI is:
Alice.Smith@domain.com
This address, the public user identity, is assigned
by the home network operator. Public user
identities are used to route SIP signaling.
Since the PSTN can only interpret digits, and
since the PSTN is going to be in use for the
foreseeable future, the IMS also accommodates
public user identity telephone uniform resource
locators (URLs) (RFC 2806), which carry a
telephone number. An example of a telephone
URL is:
+1-123-456-7890
IMS operators typically assign at least one SIP
URI and one telephone URL to each subscriber.
Operators may assign more than one public user
identity to a user, i.e., one or more for personal
use and one or more for business use.
Bechtel Telecommunications Technical Journal
Along with a public user identity, subscribers
also receive a private user identity—basically a
user name and a password—that takes the form
of a network access identifier (NAI) (RFC 2486).
The private user identity is not used to route SIP
signals, but only to authenticate subscribers.
Private user identity is usually stored in the IMS
terminal’s UMTS integrated circuit card (UICC).
The UICC typically contains a subscriber identity
module (SIM) card needed for a global system for
mobile communication (GSM) call, a UMTS
subscriber identity module (USIM) card needed
for UMTS calls, and an IMS subscriber identity
module (ISIM) card needed for IMS calls. Without
a UICC, only emergency calls can be placed using
the terminal.
The ISIM card not only stores the user’s private
user identity, but also public user identity, home
network domain URI, and long-term secrets. The
long-term secret is used for authentication and for
calculating the integrity key (Ik) and cipher key
(Ck) used between the terminal and the network.
The IMS terminal uses the Ik to protect SIP
signaling integrity between the IMS terminal and
the network. IMS security is discussed in detail in
the IMS Operations section of this paper.
SIP Network Extensions and Entities
In the IMS, extensions are (and continue to be)
made to SIP. New methods and header fields can
be defined and easily integrated into the core
protocol. For instance, SIP has an extension to
deliver instant messages and an extension to
handle subscriptions to events. SIP uses either
TCP or UDP as the transport protocol. UDP is
usually preferred, because it does not have the
overhead associated with TCP setup, teardown,
and retransmission.
• Redirect server. This entity receives SIP
requests and returns an alternative location
where the user may be available.
SIP Registration
Registration is the process by which a user
identifies himself/herself and his/her current
location to the network. To illustrate how this
works, the basic registration process performed
in a SIP (not necessarily mobile or IMS) network
is outlined for a hypothetical user, Alice Smith
[26]. Alice has a public user identity—also
known as her address of record (AoR)—of
Alice.Smith@domain.com, where domain is the
domain of Alice’s home network operator. This
address is used by Alice no matter where she logs
on. When Alice logs on at her computer at work,
her workplace SIP URI is asmith@company.com;
when she logs on at her computer at school, she
uses a SIP URI of alice@university.edu. Alice’s
AoR is the address known by the public, i.e.,
the address that anyone trying to contact Alice
would use. Alice must always register her
current location with the registrar so that the
registrar knows where to forward incoming
requests to Alice.
All SIP requests arrive at the proxy server, which
may or may not be co-located with the registrar.
If the two are co-located, SIP messages arriving at
the proxy server are directly forwarded to Alice’s
new location. If the proxy server and the
registrar—more specifically, the location server of
the registrar—are separate, then the proxy server
needs to first contact the location server to learn
the current location of Alice, and then forward
the message. Figure 3 illustrates the latter process.

The main entities in a SIP network are:

• UA. The UA—that is, the SIP endpoint—is

the entity that initiates and receives SIP
requests and generates the provisional and
final responses.
• Registrar server. This entity keeps track of user
locations. The UA sends a registration
message to the registrar. This information is
saved in a location server for future use; this
server may or may not be co-located with the
registrar.
• Proxy server. This entity is simply a SIP router
that receives, processes, and forwards SIP
requests and responses. It receives SIP
messages from a UA or from another proxy
and routes them to its destination UA or Figure 3. Proxy Server and Registrar (including Registrar’s Location Server)
another proxy.

January 2006 • Volume 4, Number 1 21

 Conducting a SIP Session Setup
Having covered the basic SIP transactions, the
discussion turns to setup of a multimedia session
using SIP. Returning to the hypothetical user
example, Bob and Alice are assumed to have
already registered their current locations with the
network; now, Bob wants to establish a session
with Alice. Alice is using her personal digital
assistant (PDA), and the proxy server, registrar,
and location server are all co-located. The
following process takes place (see Figure 4):
1. Bob sends an INVITE request to Alice’s AoR
The IMS network
at the proxy server.
can basically
2. The proxy server sends a receipt via a
be divided into
100 TRYING message.
three layers:
3. The proxy server looks up Alice’s current
• Application or location and forwards the request to
service layer Alice’s PDA.
4. Alice’s PDA sends a confirmation reply via a
• Control or
100 TRYING message.
signaling layer
5. Alice’s PDA sends a 180 RINGING message
• Access or indicating that it has been alerted to the
connectivity request to establish a session.
layer 6. Alice’s PDA sends a 200 OK reply message
back to the proxy server. This reply includes
a Contact header field in the message (so
that Bob can use this URI to contact Alice
directly for all future transactions), along
with an SDP message describing Alice’s
PDA session components such as audio and
video; the codec used for each session
component; and the transport information
such as port numbers, IP addresses, and
transport protocol.
Figure 4. Conducting a SIP Session
22
7. The proxy server forwards this 200 OK reply
message along with the SDP to Bob.
8. Bob sends an ACK message directly to Alice.
At this point, Bob and Alice can initiate a
session, say, an audio session.
9. Bob decides to terminate the session and
sends a BYE message directly to Alice.
10. Alice sends an OK reply directly back to Bob
and the session terminates.
Note that if, in the middle of the session, either
Bob or Alice wants to modify the session (for
instance, by adding a video component), he or
she can send either an UPDATE request or
another INVITE request with an updated
session description.
Building on the foregoing discussion of SIP
networks, the following sections examine the
IMS network’s architecture; the entities
involved in the IMS; and the performance of
various operations such as registration, session
setups, roaming, security, QoS and policy
support, and charging.
IMS ARCHITECTURE
s depicted in Figure 5, the IMS network can
A basically be divided into three layers:
• Application or service layer
• Control or signaling layer
• Access or connectivity layer
The application or service layer contains
application servers (ASs) such as the SIP AS,
Bechtel Telecommunications Technical Journal

Figure 5. IMS Network Architecture (After [26])
third-party open service access (OSA) AS, and
legacy service control point (SCP) AS. The IMS
controls service via the subscriber’s home
network and those signaling network elements
distributed in the application layer and the
control layer. This arrangement enables
subscribers to receive the same types of services
while they are roaming.
January 2006 • Volume 4, Number 1
The heart of the
control layer
consists of the
CSCF servers,
also known as
SIP servers.
The control layer contains signaling network
elements or control servers for session setup,
modification, and termination or for managing
calls. The heart of the control layer consists of the
call session control function (CSCF) servers, also
known as SIP servers. This layer also includes the
home subscriber server (HSS) database,
subscriber location function (SLF) database, PDF,
and breakout gateway control function (BGCF).
23
 Next, elements of each layer are examined along
COMMONLY USED IMS INTERFACE with their respective roles or functions in
DESIGNATORS establishing, modifying, and terminating a
Cx between HSS and I-CSCF/ session; supporting roaming; providing desired
S-CSCF QoS; providing charging information; etc. The
Dx between SLF and I-CSCF/ discussion begins with the control layer elements,
S-CSCF since the IMS calls basically start from this point:

Gm between I-CSCF/S-CSCF and • CSCF server is also known as the SIP server,
user equipment since it processes SIP signaling. There are
Go between PDF and PEP three types of CSCFs: proxy (P-CSCF),
interrogating (I-CSCF), and serving
Gq between PDF and P-CSCF (S-CSCF).
Mb between IM-MGW and entities — P-CSCF is the first point of contact, in
such as user terminal, MRFP, AS
the signaling plane, between the IMS
Mi between BGCF and I-CSCF/ terminal and the IMS network. From
S-CSCF the SIP point of view, the P-CSCF acts
Mj between BGCF and MGCF as a SIP proxy server, i.e., all requests
initiated by the IMS terminal or destined
Mk within a given BGCF for the IMS terminal traverse the P-CSCF,
(intra-BGCF)
which provides data integrity and
Mn between MGCF and MGW confidentiality by using IP security
(IPSec) to maintain a security association
Mp between MRFC and MRFP
between itself and each IMS terminal.
Mw between CSCFs within the The P-CSCF handles the charging
same network records for billing purposes by creating
Ro between OCF and S-CSCF, and maintaining a charging data record
MRFC, SIP-AS, et al. (CDR) that can be consolidated at a
charging gateway function (CGF). The
Sh between HSS and AS
(SIP-AS or OSA-SCS) P-CSCF also provides QoS authorization
and control by providing the necessary
Si between HSS and IM-SSF information to the PDF for resource
Za between security domains authorization and QoS control. The IMS
(inter-domain) also supports roaming services via the
P-CSCF. (Roaming, QoS, charging, etc.,
Zb within a given security domain
(intra-domain) are discussed in more detail in the IMS
Operations section of this paper). Also,
Note: X/(Y) notation on the interfaces once the P-CSCF authenticates the user,
indicates that the interface is X and the
it asserts the identity of the user to the
protocol used on that interface is Y.
rest of the nodes (or IMS elements) in
the network, so those nodes do not
need to authenticate the user again.
The connectivity or access layer is used to
IMS networks usually have several
transport signaling traffic and media streams.
P-CSCFs for the sake of scalability and
This layer contains switches, router, and media-
redundancy, and each P-CSCF serves a
processing entities (MGWs, signaling gateways
certain number of IMS terminals, based
[SGWs], MRF controls [MRFCs], and MRF
on its capacity.
processors [MRFPs]). Since IMS is designed to be
access independent, it can connect to different — I-CSCF is a proxy server that is located
types of existing and emerging access networks at the edge of an administrative domain
as long as they have IP connectivity. Access and that interfaces with SLF and HSS
networks that can connect with the IMS include databases. These interfaces are based
GPRS/UMTS, 2G networks such as GSM via on the Diameter protocol. The I-CSCF
gateways, PSTNs via gateways, enterprise fixed retrieves user location information
networks via IP Centrex, residential fixed and routes the SIP requests to their
networks via xDSL or cable modem, WLANs, appropriate destinations, typically an
and worldwide interoperability for microwave S-CSCF. The I-CSCF also assigns an
access (WiMAX). S-CSCF if there is more than one. The

24 Bechtel Telecommunications Technical Journal

 I-CSCF selects the S-CSCF based on the
information queried from the HSS
through the Diameter-based Cx interface.
If there are multiple HSSs, the I-CSCF
must first contact the SLF (explained
below) through the Diameter-based Dx
interface to obtain the HSS addresses.
The I-CSCF may also act as a topology
hiding internetwork gateway (THIG) by
hiding sensitive information about the
domain, such as the number of servers,
their domain name system (DNS) names,
and their capacities. An IMS network
may have several I-CSCFs for the sake
of scalability and redundancy.
— S-CSCF is essentially a SIP proxy that
relays SIP messages, a SIP UA that
initiates and terminates SIP transactions,
and a SiP registrar that authenticates
users during registration. Most
importantly, the S-CSCF controls the
session. The S-CSCF interfaces with
the HSS via Diameter protocol and
downloads the user’s profile and
authentication vector (AV) to be used
in user authentication. The user profile
includes the subscriber service profile
(SSP), which has trigger points and
user-specific filter criteria. The S-CSCF
uses this information to control a user’s
access to different ASs. The S-CSCF
also collects data for charging purposes.
An IMS network may have several
S-CSCFs for the sake of scalability and
redundancy. Both the S-CSCF and the
P-CSCF maintain session timers, i.e.,
they are stateful proxies.
• HSS is a stateless Diameter server and a
database that holds all of the subscriber’s
information. It is basically an advanced home
location register (HLR) that holds user
information, including location, security data
(AV), user profile SSP, trigger points and
filter criteria, and the user’s allocated
S-CSCF. The HSS may also support HLR/
authentication center (AUC) functionality
and mobile-application-part (MAP)-based
interfaces for legacy 2G and 2.5G networks.
Subscriber data stored in the HSS is the
key enabler for service mobility across
different types of access networks and for
user roaming between different network
operators. A network may require more than
one HSS due to the number of subscribers
and the capacity of the HSS. Because of its
importance, the HSS is always implemented
in redundant configuration.
January 2006 • Volume 4, Number 1
• SLF is a Diameter-based redirect agent or
server that maps the user’s address to a
specific HSS. A network with a single HSS
does not require an SLF.
• PDF may be part of the P-CSCF or a
standalone entity. It interacts with the
P-CSCF via the Diameter-based Gq interface
and with the PEP at the packet data gateway
(PDG) via the COPS-based Go interface. The
PDG for the GPRS/UMTS network is the
gateway GPRS support node (GGSN).
• AS is a SIP entity that hosts and executes An application
services. New IMS-specific services are
server is a
expected to be developed in SIP ASs. An AS
may host several different applications. SIP entity that
hosts and executes
• IP multimedia service switching function
(IM-SSF) is a specialized AS that allows reuse services. New
of a GSM network’s customized application IMS-specific
for mobile network enhanced logic (CAMEL) services are
services in the IMS. The IM-SSF allows the
gsmSCF to control an IMS session.
expected to be
developed in
• MRF provides the source for media in the
home network. It enables the home network SIP ASs. An AS
to play announcements, mix media streams, may host several
transcode between different codecs, perform different
media analyses, and provide statistics. The
applications.
MRF is divided into two parts: MRFC and
MRFP. MRFC acts as a SIP UA and contains
an interface with the S-CSCF. It also controls
resources in the MRFP via a Megaco
interface. The MRFP implements all media-
related functions, such as playing media.
• BGCF is the SIP server with routing
functionality based on telephone numbers.
It is used in sessions that are initiated
by IMS terminals and addressed to users
in a CS network such as PSTN or other
cellular network. The BGCF’s main
functionality is to select an appropriate
PSTN/CS gateway.
• Media gateway control function (MGCF) is
the main node of the PSTN/CS gateway.
It has two primary functions: (1) to perform
control signal protocol conversion from SIP
to BICC or ISUP with the SGW, and (2) to
control the resources in the MGW. The MGW
converts the media formats between RTP on
the SIP side and pulse code modulation
(PCM) on the PSTN side.
IMS OPERATIONS
ome of the more important issues in IMS
S network operations, such as prerequisites,
registration, session setup, security, QoS and
25
 policy support, and charging, are examined in
this section.
IMS Operation Prerequisites
Before a user can access IMS service, the
IMS terminal must perform three major tasks:
Access the IMS home or visited network via the
IP connectivity access network (IP-CAN),
discover the P-CSCF address, and perform
IMS level registration.
The IP-CAN could be GPRS/UMTS, WLAN, or
xDSL, although Release 5 of the 3GPP (the first
To provide
release of the IMS) supports only GPRS/UMTS.
integrity and
Thus, the IMS terminal must perform a GPRS
confidentiality, Attach with the serving GPRS support node
the P-CSCF and (SGSN) and perform bearer level authentication
with the HLR. To discover the IP address of the
the IMS terminal
P-CSCF, which is stored at the GGSN, the
establish
terminal must perform a packet data protocol
two IPSec SAs (PDP) context activation with the GGSN. Finally,
between the terminal must perform IMS level registration
with the CSCF and the HSS. At this point, the IMS
themselves.
terminal can access various applications on ASs
via the CSCF.
IMS User Authentication and Security
Associations
Before a user can complete IMS level registration,
the user must perform user authentication
and establish security associations. User
authentication in the IMS is based on a
challenge/response algorithm and the secrets
stored in the ISIM part of the UICC and in the
HSS. The S-CSCF is the entity that performs the
authentication, and the P-CSCF is the entity that
Figure 6. IMS User Authentication
26
performs the integrity and confidentiality
protection of the messages.
The authentication process begins after the
primary PDP context activation, when the IMS
terminal sends a SIP register request to the
CSCF. The CSCF contacts the HSS and obtains the
user’s AV. The AV is composed of a random
number (RAND) challenge, an authentication
token (AUTN), the expected response (XRES), the
Ck, and the Ik. The XRES is calculated based on a
known algorithm RAND and a stored secret.
The CSCF replies with a SIP 401 message and
includes both the RAND and the AUTN. Upon
receiving this message, the IMS terminal
computes its own response (RES) based on the
received RAND, the same known algorithm used
by the CSCF, and the secret keys retrieved
from its ISIM card. The IMS terminal sends
another SIP register request message that
includes its own RES. Upon receiving this
message, the CSCF compares the XRES with the
RES, and, if they agree, the CSCF replies with the
SIP 200 OK message. This process is summarized
in Figure 6.
To provide integrity and confidentiality, the
P-CSCF and the IMS terminal establish two IPSec
security associations (SAs) between themselves.
One SA is established from the terminal’s client-
protected port to the P-CSCF’s server-protected
port, and the other SA is established from the
P-CSCF’s client-protected port to the terminal’s
server-protected port. Both SAs support traffic in
both directions. The P-CSCF and the IMS terminal
need to agree on a set of parameters to establish
the two IPSec SAs. The P-CSCF obtains the Ck
and the Ik in the SIP 401 unauthorized reply
message from the S-CSCF. The P-CSCF
removes both keys before relaying the
message to the IMS terminal. To negotiate
the rest of the IPSec parameters, the P-CSCF
and the IMS terminal use the same two SIP
register messages that are used for
authentication. After these steps are
completed, IPSec is permanently set up for
the duration of the session hop-by-hop
between the P-CSCF and I-CSCF, I-CSCF and
S-CSCF, S-CSCF and AS, AS and HSS,
P-CSCF and S-CSCF, and S-CSCF and other
entities such as BGCF and MGCF. Along
with the foregoing actions, encryption may
also take place between the IMS terminal and
the GGSN on the radio bearer end. This
process is shown in Figure 7.
Bechtel Telecommunications Technical Journal
IMS Level Registration
IMS level registration involves the following
basic actions:
• The P-CSCF and the S-CSCF are assigned to
the IMS terminal, and a path is established
between them.
• The IMS terminal is authenticated.
• Integrity and confidentiality security
associations are set up.
• The S-CSCF downloads the user profile
from the HSS.
• The S-CSCF stores the address information
in the HSS.
Specific steps in the IMS level registration process
are as follows:
Step 1: The IMS terminal sends a SIP
register request message to the
P-CSCF and compresses this
message to save transmission
bandwidth and minimize signaling
delays. The receiving P-CSCF
decompresses the request message.
Step 2: If there are multiple S-CSCFs, the
P-CSCF forwards the request to
the I-CSCF, so that the I-CSCF
can select an S-CSCF to serve the
user’s session.
Step 3: The I-CSCF sends a Diameter-based
user authorization request (UAR)
message to the HSS, requesting the
addresses of all available S-CSCFs.
Figure 7. IMS Security Associations
January 2006 • Volume 4, Number 1
Step 4: The HSS replies with a user authori-
zation answer (UAA) message that
includes a list of all available
S-CSCFs.
Step 5: The I-CSCF selects one S-CSCF and
forwards the register request to the
selected S-CSCF.
Steps 6–7: The S-CSCF retrieves the user AV
from the HSS via a Diameter-based
multimedia authentication request
(MAR) message and a Diameter-
based multimedia authentication
answer (MAA) message.
Steps 8–10: From the AV, the S-CSCF computes
the user-specific challenge data and
sends this information via a 401
UNAUTHORIZED message to the
IMS terminal through the I-CSCF,
P-CSCF, and IMS terminal.
Steps 11–15: The IMS terminal computes its
authentication response and sends
it to the S-CSCF via another register
request message.
Steps 16–17: The S-CSCF verifies the response
and, if correct, downloads the
subscriber profile from the HSS via
Diameter-based server assignment
request (SAR) and server assign-
ment answer (SAA) messages. The
S-CSCF may contact an AS (or ASs)
for service control as specified in
the SSP.
27
 Steps 18–20: Finally, the S-CSCF sends a 200 OK
message back to the IMS terminal
via the I-CSCF and the P-CSCF.
A graphic representation of the above steps is
shown in Figure 8.
IMS Session Setup
Session setup is the process of discovering
network nodes and signaling paths.
Building on the SIP session setup functionality
detailed earlier in this paper, the IMS
Within a security offers additional functionality, including
domain, network SIP signaling compression, SIP message
nodes use IPSec to integrity/confidentiality protection, QoS-related
capabilities and features, routing capabilities to
exchange traffic
the PSTN/CS networks, tracking and recording
with each other and of charging information, etc.
with the SEG.
IMS Security
IMS security can be divided into two areas
[9, 10, 26]: access security and network security.
Access security involves authentication of users
and networks and protection of traffic between
IMS terminals and networks. Network security
involves protection of traffic between security
domains. (A security domain is a network
managed by a single administrative authority.)
Figure 8. IMS Level Registration
28
For access security, as indicated in the earlier
discussion of IMS level registration, an IMS
terminal request message travels through the
P-CSCF and the S-CSCF during the registration
process. The S-CSCF performs the authenti-
cation, and the P-CSCF establishes two IPSec
associations to protect the traffic to and from
the terminal.
To provide network security, all traffic entering
or leaving a security domain passes through a
security gateway (SEG), as shown in Figure 9.
Security associations between SEGs are
established and maintained using the Internet
key exchange (IKE) protocol (RFC 2409). The
traffic between SEGs is protected using IPSec
encapsulation security payload (ESP) (RFC 2406)
running a tunnel mode.
Within a security domain, network nodes use
IPSec to exchange traffic with each other and with
the SEG. The inter-domain interfaces, Za, require
mandatory authentication, integrity protection,
and encryption. The intra-domain interfaces, Zb,
carry only intra-domain signaling traffic and
require mandatory integrity protection.
Encryption on these interfaces is optional.
Figure 10 illustrates this process.
Bechtel Telecommunications Technical Journal
 Figure 9. Network Security Between Domains

Since CS
networks and the
PSTN will remain
in use for the
foreseeable future,
the IMS
also provides
interworking
capabilities with
BICC- and
ISUP-based
Figure 10. Intra-Domain and Inter-Domain Network Security
legacy networks.
In addition to the mandatory network layer appropriate CS/PSTN gateway, if interworking
security (IPSec), the IMS also provides optional will take place in the same network in which the
transport layer security (TLS) and application BGCF resides. The MGCF is the component that
layer security (e.g., HTTP digest authentication controls the IM-MGW and that performs SIP-to-
for SIP). BICC or SIP-to-ISUP call-related signaling

The IMS also contains three interfaces, Cx, Dx, and

Sh, that use authentication functions. Interface Dx
is the interface between the SLF and the I-CSCF
and P-CSCF. Interface Cx is the interface between
the HSS and the I-CSCF and P-CSCF. And
interface Sh is the interface between the HSS and
the ASs. In all three interfaces, Diameter is the
authentication protocol. Diameter runs over
reliable transport protocols such as TCP and
streaming control transmission protocol (SCTP).

IMS-CS Interworking
Since CS networks and the PSTN will remain in
use for the foreseeable future, the IMS also
provides interworking capabilities with BICC-
and ISUP-based legacy networks (e.g., PSTN,
ISDN, and CS public land mobile networks
[PLMNs]). The main IMS functional entities
involved are the BGCF, MGCF, IM-MGW, and
SGW. IMS-CS interworking architecture is shown
in Figure 11.

The BGCF’s main functionality is to select an

appropriate network in which interworking with
the CS domain will take place, or to select an Figure 11. IMS-CS Interworking Architecture (Source [11])

January 2006 • Volume 4, Number 1 29

 interworking. The IM-MGW converts media
formats provided in one type of network to the
format required in another type of network; for
instance, it provides the interface between the PS
domain and the CS domain. The IM-MGW may
also be connected via the Mb interface to various
network entities such as a user terminal (via a
GPRS tunneling protocol [GTP] tunnel to a
GGSN), an MRFP, or an AS. The SGW performs
the call-related signaling conversion to or from
the BICC/ISUP-based MTP transport networks to
the BICC/ISUP-based SCTP/IP transport
The main drawback networks and forwards the converted signaling
of IntServ is that to or from the MGCF. The SGW may be
implemented as a standalone entity or located in
it does not scale
another entity in the CS network or the IM-MGW.
well. This is
primarily because IMS QoS Support Mechanism
(1) the network The QoS support mechanism ensures that the
critical elements of IP transmission such as
needs to store a
transmission rate, gateway delays, and error rates
large amount of
can be measured and guaranteed in advance. This
information, and function is performed mainly via the PDF, which
(2) routers need to interacts with and controls the underlying packet
network (i.e., the access network resources) via
look up large tables
the Go interface with an element in the PDG
before they can
called the PEP. The PDG for the GPRS/UMTS is
route the packets. the GGSN, which hosts the PEP. Policy-related
information is transmitted between the PDF and
the PEP using COPS (RFC 2478).
Currently, two basic methods support QoS
on the Internet: integrated services (IntServ)
Figure 12. QoS Authorization
30
(RFC 2215 [27]) and differentiated services
(DiffServ) (RFC 3260 [28]).
IntServ is designed to provide end-to-end QoS
with two classes of services: controlled load and
guaranteed. IntServ uses resource reservation
protocol (RSVP) to reserve resources with the
desired QoS. RSVP also ensures that the routers
receiving resource reservation requests are the
routers that will actually route the packets. This
function is performed via a two-way handshake,
in which one endpoint (endpoint A) sends a
PATH message to the other endpoint (endpoint
B), recording all the visited intermediate nodes.
Then, in the reverse direction starting at endpoint
B, a RES message is sent through all the nodes
recorded in the PATH message, this time actually
reserving the resources.
Note that a router in the path can reject a resource
reservation request either because it does not
have the required resources or because the
requester does not have the permissions to
reserve those resources. Thus, RSVP can be
considered as not only a resource reservation
protocol, but also an admission control protocol.
The main drawback of IntServ is that it does not
scale well. This is primarily because (1) the
network needs to store a large amount of
information, and (2) routers need to look up
large tables before they can route the packets.
To address these issues, DiffServ architecture
was proposed.
Bechtel Telecommunications Technical Journal
In DiffServ, routers identify packet treatment Step 7: The GGSN extracts the token and
without the need for table lookup. Packet the packet flow identifier and sends
treatments, known as per-hop behaviors (PHBs), this information to the PDF using
are identified by 8-bit codes called differentiated the COPS REQ (request) message.
services code points (DSCPs). Packets are marked The packet flow identifier contains
at the edge of the network with a certain DSCP, so the source address, the destination
that routers in the path apply the correct PHB to address, the source port number,
them. DSCPs are encoded in the Types of Service the destination port number, and
field of IPv4 and the Traffic Classes field of IPv6. the transport protocol used.
Two examples of PHBs are expedited forwarding
Step 8: The PDF responds to the GGSN
and assured forwarding. In expedited forwarding
with a COPS DEC (decision)
PHB, packets never experience congestion in
message that contains the QoS
the network. In assured forwarding PHB, On the GPRS/UMTS
characteristics of the IMS terminal’s
packet-drop precedence is determined, allowing access network,
authorized session. This is known
low priority packets to be discarded before
as service-based local policy (SBLP) the link-layer
high priority packets; some packets may in fact
information. The GGSN uses this resource reservation
be discarded.
information to install packet filters
is performed via
In the IMS, end-to-end QoS involves both QoS that allow only authorized packet
over the access network and QoS in the core flows to be transmitted over a given PDP context
network. This implies that QoS-required PDP context. activation, and the
resources have to be provisioned and enforced on GGSN maps link
Step 9: The GGSN (actually the PEP
both sides. This can be done by using a link-layer
residing in the GGSN) sends an layer resource
RSVP on the access network side and the DiffServ
RTP message to the PDF indicating reservation flows
method (or RSVP) on the network side. On the
that it will comply with the
GPRS/UMTS access network, the link-layer to DiffServ code
PDF’s policy.
resource reservation is performed via PDP points in the
context activation, and the GGSN maps link layer Step 10: The GGSN sends a PDP context network.
resource reservation flows to DiffServ code points response message back to the
in the network [7, 8, 26]. SGSN, authorizing the SGSN for the
requested PDP context.
Specific steps in the QoS provisioning process are
described below. Step 11: The SGSN forwards this response to
the IMS terminal.
For inbound sessions:
Figures 12 and 13 summarize the above process.
Step 1: An INVITE request message arrives
at the P-CSCF/PDF. For outbound sessions of the QoS provisioning
process, only the initial steps differ from those for
Step 2: The P-CSCF adds a media
inbound sessions. Replacing Steps 1 and 2 of
authorization token to the message
inbound sessions, the corresponding steps
and forwards this message to the
involved in outbound sessions are:
IMS terminal.
Step 1a: The IMS terminal sends an
Step 3: The IMS terminal creates a PDP
INVITE request message to the
context activation request message
P-CSCF/PDF.
and sends it to the SGSN.
Step 1b: The P-CSCF/PDF forwards this
Steps 4–6: The SGSN receives this message
message to the callee.
and checks the user’s subscription
information stored in the HSS using Step 2a: The P-CSCF sends a session
mobile application part (MAP) progress message to the callee.
protocol. If the IMS terminal
Step 2b: The P-CSCF also adds the
requests more resources than it is
authorization token to this message
allowed to use, the SGSN adjusts
and forwards it to the IMS terminal.
the requested resources to the
appropriate level and sends a PDP Steps 3 through 11 of outbound sessions are
context request message to the identical to those of inbound sessions of the QoS
GGSN, along with the authorization provisioning process.
token.

January 2006 • Volume 4, Number 1 31

 As was mentioned earlier, the GGSN/PEP must
also play the role of a DiffServ edge router and
map the required QoS to the appropriate DSCP to
provide the desired QoS on the core network side,
assuming DiffServ is used there.
IMS Service Delivery
One of the main features of the IMS is its service
delivery capabilities. The primary IMS network
elements involved in service delivery are the
S-CSCF, which acts as the central session control
point; the ASs, where the actual services reside
The IMS employs and service-specific SIP processing is performed;
three types of the HSS, which provides information on the
service delivery services and qualities accessible to the IMS user;
and the MRF (along with MRFC and MRFP),
platforms: SIP-AS,
which controls the media resources.
OSA-AS, and
The IMS employs three types of service delivery
legacy SCP.
platforms: SIP-AS, OSA-AS, and legacy SCP. The
subscriber’s S-CSCF interacts with the service
platforms through a SIP-based, intra-operator
interface known as the IMS service control (ISC).
ASs may access user data in the HSS for
application-specific information via the Sh
interface; as mentioned earlier, the S-CSCF
determines whether to invoke a particular AS
based on specific filter criteria from the user’s SSP
stored in the HSS. Multiple applications can
reside on a single AS.
Figure 13. QoS Authorization Flow (Inbound Session)
32
Service control in the IMS is home network based,
which means that the user receives the same
services whether operating in the home network
or roaming into a visited network.
IMS Roaming
As indicated earlier, roaming is handled via the
P-CSCF. The P-CSCF is the only node or element
in the IMS that must be located in either the home
network or the visited network. Also, when IP
connectivity access is via GPRS or UMTS, the
location of the P-CSCF is subordinate to the
location of the GGSN. In roaming scenarios,
GPRS and UMTS allow the GGSN to be located
either in the home network or the visited
network, while the SGSN is always located in the
home network. Furthermore, in the IMS, both the
GGSN and the P-CSCF share the same network,
and the P-CSCF controls the GGSN via the
Go interface.
Currently, the IMS allows two configurations for
roaming. In the first configuration, shown in
Figure 14a, the P-CSCF (and the GGSN) are
located in the home network. This arrangement
would probably be used in the early stages of IMS
deployment. In this configuration, the visited
network is not required to have an IMS-compliant
(i.e., 3GPP Release 5 or newer) GGSN. The visited
network only provides the radio bearers and the
SGSN services for the roaming terminal. The only
Bechtel Telecommunications Technical Journal

Figure 14a. P-CSCF and GGSN Located in a Home Network (After [26])
Figure 14b. P-CSCF and GGSN Located in a Visited Network (After [26])
negative aspect of this configuration is that media
are first routed to the home network and then to
their destination, introducing unnecessary and
undesirable delays. This configuration, however,
allow immediate deployment of the IMS and
IMS roaming [26].
In the second configuration, shown in Figure 14b,
the P-CSCF and the GGSN are located in the
visited network, i.e., the visited network has an
IMS-compliant GGSN. There may be a move to
this configuration after initial IMS deployment.
IMS Charging
The IMS supports both offline (or post-paid) and
online (or real-time pre-paid) charging services.
The IMS charging architecture is shown in
Figure 15 [12–14, 29].
The online charging services are handled by
the online charging function (OCF) AS; the
offline charging services are handled by
the charging collection function (CCF) AS. In
Release 6, the CCF is upgraded to the charging
data function (CDF).
January 2006 • Volume 4, Number 1
The IMS supports
both offline
(or post-paid)
and online (or
real-time pre-paid)
charging services.
The CCF/CDF is a stateless Diameter-based AS
that does not maintain session states, but keeps
track of transaction states. It creates, updates, and
closes charging data records (CDRs) based on
accounting request (ACR) messages it receives
from IMS elements. There are three types of ACR
messages: ACR (Start) to generate the CDR and
start accounting, ACR (Update) to update the
CDR, and ACR (Stop) to update and close the
CDR. Figure 16 depicts the signaling flow for a
session-based offline charging scenario in which
an ACR (Start) is triggered upon receipt of an
INVITE request message to start a session; an
ACR (Update) is started when the interim period
elapses, and an ACR (Stop) is triggered upon
receipt of a session termination BYE request
message.
The OCF is a stateful Diameter-based AS that
maintains both session states and transaction
states for online charging. IMS elements
interacting via the Ro interface with the OCF
include the S-CSCF, MRFC, and SIP ASs. The
S-CSCF interacts with the SCF of the OCF for
session-based pre-paid service control, while the
33
 Figure 15. IMS Charging Architecture

SIP AS and the MRFC interact with the event

charging function (ECF) of the OCF for content-
based pre-paid control. The SIP-Diameter
interworking function (IWF) provides SIP
to/from Diameter format conversion and also
acts as a credit control client (CCC).

Depending on the received SIP messages and the

service usage condition, the CCC sends a
Diameter-based credit control request (CCR)
message to the OCF/SCF, and the OCF/SCF
sends a credit control answer (CCA) reply
message. There are basically three types of CCRs
(and CCAs): CCR (Start), CCR (Update), and
CCR (Stop).

CCR (Start) corresponds to the initial CCR

Figure 16. Session-Based Offline Charging
message to start a session and is triggered when
the CCC receives an INVITE request to start the
session. Via this CCR message, the CCC checks
with the OCF/SCF to verify that the IMS terminal
has enough credit left for session initiation and
usage and then reports back in the CCA reply
message. At this point, the IWF/CCC starts
timing or counting session usage. When the
granted units or account balances near depletion,
the CCC sends a CCR (Update) message to the
OCF/SCF to request more credit. The OCF/SCF
replies to this request via a CCA and includes the
amount of additional credit, if any. Finally, if a
termination request is received via a BYE request
message, the CCC sends a CCR (Stop) message to

34 Bechtel Telecommunications Technical Journal

the OCF/SCF and reports the amount of credit
used and the amount left over. Of course, if the
credit is depleted before a SIP termination request
arrives, and additional credit cannot be obtained,
the CCC can force the session to be terminated.
CONCLUSIONS
he IMS allows an evolutionary move to
T all-IP converged networks. But in terms of
capabilities and features offered, the IMS is
truly a revolutionary approach to multimedia,
multisession service deployment. It provides
flexible session control with desirable features
such as guaranteed end-to-end QoS, roaming
capabilities, security, and easy and convenient
charging. It also allows horizontal service
deployment by offering a common platform with
reusable components and open interfaces. For
instance, the IMS will be able to offer PTT
with almost immediate user access. Deploying
PTT service will be as easy as uploading the
AS software in an AS and uploading the
client application software over the air (OTA) to
the user’s terminal. With the IMS in place,
services and applications can be deployed more
quickly, easily, and economically than ever
before. Thanks to the IMS, the vision of a personal
communication system (PCS) can finally become
a reality!

TRADEMARKS
3GPP is a trademark of the European
Telecommunications Standards Institute (ETSI)
in France and other jurisdictions.
cdma2000 is a registered trademark of
the Telecommunications Industry Association
(TIA-USA).
REFERENCES
[8] 3GPP TS 29.208, “End-to-End Quality of Service
(QoS) Signaling Flows.”
[9] 3GPP TS 33.203, “3G Security; Access Security
for IP-Based Services.“
[10] 3GPP TS 33.210, “Security Requirements.”
[11] 3GPP TS 29.163, “Interworking Between the IP
Multimedia (IM) Core Network (CN) Subsystem
and Circuit Switched (CS) Networks.”
[12] 3GPP TS 32.225, “Telecommunication
Management; Charging Management; Charging
Data Description for the IP Multimedia
Subsystem (IMS).”
[13] 3GPP TS 32.260, “Telecommunication
Management; Charging Management; IP
Multimedia Subsystem (IMS) Charging.” The IMS allows
[14] 3GPP TS 32.295, “Telecommunication an evolutionary
Management; Charging Management; Charging
Data Record (CDR) Transfer.” move to all-IP
[15] 3GPP TS 32.260, “IP Multimedia Subsystem converged networks.
(IMS) Charging.”
[16] “Siemens IP Multimedia Subsystems (IMS),” But in terms of
Siemens White Paper. capabilities and
[17] “IP Multimedia Subsystem IMS Overview and
Applications,” 3G America White Paper.
features offered,
[18] “IMS–IP Multimedia Subsystem,” Ericsson the IMS
White Paper, October 2004.
is truly a
[19] RFC 3261, “SIP: Session Initiation Protocol,”
June 2002. revolutionary
[20] RFC 2327, “SDP: Session Description Protocol,” approach to
April 1998.
multimedia,
[21] RFC 3550, “RTP/RTCP: Real Time Protocol/
Real Time Control Protocol,” July 2003. multisession service
[22] RFC 3588, “Diameter,” September 2003. deployment.
[23] RFC 2866, “RADIUS: Remote Authentication
Dial-In User Services,” June 2000.
[24] RFC 2478, “COPS: Common Open Policy
Services,” January 2000.
[25] G. Camarillo, SIP Demystified, McGraw-Hill,
2001.
[26] G. Camarillo and M.–A. Garcia-Martin, The 3G IP
Multimedia Subsystems (IMS): Merging the Internet
and the Cellular World, John Wiley & Sons,
August 2004.
[27] RFC 2215, “IntServ: Integrated Service,”
September 1997.
[28] RFC 3260, “DiffServ: Differentiated Services,”
April 2002.
[29] “IMS Signaling Architecture,” Ulticom White
Paper, 2005.

[1] 3GPP TS 23.228, “Service Requirements for the

Internet Protocol (IP) Multimedia Core Network
Subsystem (IMS), Stage 1.” BIOGRAPHY
[2] 3GPP TS 23.228, “IP Multimedia Subsystem Rasoul Safavian brings more
(IMS); Stage 2.” than 15 years of experience
[3] 3GPP TS 24.229, “IP Multimedia Call Control in the wired and wireless
Protocol Based on SIP and SDP.” communications industry to
[4] 3GPP TS 23.002, “Network Architecture.” his position as Bechtel
Telecommunications’ new vice
[5] 3GPP TS 32.299, “Diameter Charging
Applications.” president of Technology,
Americas Regional Business
[6] 3GPP TS 24.228, “Signaling Flows for the IP Unit. He is charged with
Multimedia Call Control Based on SIP and SDP.” establishing the overall
[7] 3GPP TS 23.207, “End-to-End Quality of Service technical vision for Bechtel’s American markets and
(QoS) Concept and Architecture.” providing guidance and direction to its specific

January 2006 • Volume 4, Number 1 35

 technological activities. In fulfilling this responsibility,
he will be well served by his background
in cellular/PCS, fixed microwave, satellite
communications, wireless local loops, and fixed
networks; his working experience with major 2G, 2.5G,
3G, and 4G technologies; his exposure to the leading
facets of technology development as well as its
financial, business, and risk factors; and his extensive
academic, teaching, and research experience.
Before joining Bechtel in June 2005, Dr. Safavian
oversaw advanced technology research and
development activities, first as vice president of the
Advanced Technology Group at Wireless Facilities, Inc.,
then as chief technical officer and vice president of
engineering at GCB Services. Earlier, over an 8-year
period at LCC International, Inc., he progressed
through several positions. Initially, as principal
engineer at LCC’s Wireless Institute, he was in charge of
CDMA-related programs and activities. Next, as lead
systems engineer/senior principal engineer, he
provided nationwide technical guidance for LCC’s XM
satellite radio project. Then, as senior technical
manager/senior consultant, he assisted key clients with
the design, deployment, optimization, and operation of
3G wireless networks.
Dr. Safavian is quite familiar with the Electrical
Engineering departments of four universities: The
George Washington University, where he has been an
adjunct professor for several years; The Pennsylvania
State University, where he is an affiliated faculty
member; Purdue University, where he received his
PhD in Electrical Engineering, was a graduate research
assistant, and was later a member of the visiting faculty;
and the University of Kansas, where he received both
his BS and MS degrees in Electrical Engineering and
was a teaching and a research assistant. He is a senior
member of the IEEE and a past official reviewer of
various transactions and journals.

36 Bechtel Telecommunications Technical Journal

A SURVEY OF MEMS-ENABLED OPTICAL
DEVICES – APPLICATIONS AND DRIVERS
FOR DEPLOYMENT

Issue Date: January 2006

Abstract—Telecommunications equipment manufacturers are deploying MEMS-based optical components in

next-generation equipment. These components, whether active or passive, enable gains in speed and reductions
in cost, power consumption, and size. They also can enable carriers to offer new services and reduce the cost of
deploying existing services. Components include VOAs, ROADMs, and optical switches. Carriers will face new
challenges, however, when deploying these devices.

BACKGROUND benefit of being a mature technology, as MEMS-

he beginning of this decade was marked by
T tales of the coming unlimited capacity and
endless bandwidth to be provided by ultra-fast,
ultra-cheap optical equipment. Multiple
equipment manufacturers launched new
photonic, or all-optical, switches, promising the
ability to switch hundreds of wavelengths at a
time—in fractions of the space and power, and at
a fraction of the cost—of existing equipment. The
components that made up the core of these new
switches would not be semiconductors based on
silicon, as in existing technologies, but novel
materials that enabled the devices to redirect the
path of light. This would eliminate optical-
electrical-optical (OEO) conversions, which
account for the bulk of the cost, as well as the
space and power consumption, in a modern
optical network [1].
based technology is widely deployed in other
industries, including the automotive, aerospace,
and medical fields.
Unfortunately for the proponents of all-optical
networks and photonic switches, MEMS-based
and otherwise, the next few years were marked
by the largest downturn in the history of the
telecommunications industry. Overcapacities on
routes, abundances of dark fiber, and excess
equipment in the network were noted time and
again, and the once emerging desires for huge
optical switches disappeared. Equipment
manufacturers scrapped photonic switch designs,
placed upgrades on hold, or even disappeared
entirely, as carriers suddenly realized their
growth forecasts were upside-down.

Several novel techniques for controlling and

directing an optical signal were proposed by
various manufacturers. These methods included
using polarized thin films, ceramic or polymer
materials, liquid crystals, or even bubbles.
However, the most popular method for optical
switching, micro-electrical mechanical systems
(MEMS), used tiny arrays of tilting mirrors.
Controlled electrical signals were used to adjust
the arrays of mirrors to the proper angle,
allowing for the desired output signal to appear
on the correct port. Figure 1 illustrates the MEMS
Brian Coombe mirror used in one manufacturer’s optical switch.
bcoombe@bechtel.com
MEMS-based systems allowed significantly
higher port-count switches than competing
technologies. These systems also offered the Figure 1. Optical MEMS Mirror Used in an Optical Switch [2]

© 2006 Bechtel Corporation. All rights reserved. 37

 leaner operating models, trying to do more with
ABBREVIATIONS, ACRONYMS, AND TERMS less. This is driving network operators to consider
2D MEMS two-dimensional (axes) a new set of requirements when evaluating optical
MEMS equipment. Features of interest now include
3D MEMS three-dimensional (axes) automatic provisioning, remote configurability,
MEMS and reduced power and space needs. These drive
the all-important reductions in operating expenses
DWDM dense WDM
that are key to positive cash flows [3].
FTTX fiber-based access network
topologies that include The need for reduced operating expenses does not
fiber-to-the-home/premise, enable equipment providers to sell “gold-plated”
fiber-to-the-curb, and equipment to carriers for their networks. The
fiber-to-the-node competitive landscape has also driven carriers’
Equipment GMPLS generalized MPLS capital budgets to lower levels, especially in core
providers must find networking and optical transports. Equipment
MEMS micro-electrical mechanical
systems providers must find ways to reduce their
ways to reduce
customers’ operational expenses, with less
their customers’ MPLS multiprotocol label switching expensive equipment.
operational ODF optical distribution frame
With the overbuilt core, carriers are paying more
expenses, with OEO optical-electrical-optical attention to access—literally the “on-ramps” to the
less expensive ROADM reconfigurable optical core network. Nearly every major local carrier has
equipment. add-drop multiplexer announced a fiber-to-the-node, fiber-to-the-curb,
VOA variable optical attenuator or fiber-to-the-home/premise (collectively known
as FTTX) strategy, with some carriers beginning
WDM wavelength division
construction, and others even offering services
multiplexing
over these networks. The sheer amount of data
that can be delivered over these networks will
As market demands catch up to the glut of drive capacity increases in the core transport
network capacity, carriers are beginning to build mechanisms as well. The new fiber infrastructures
out newer infrastructure. Not only must this put in place by these efforts will also require new
infrastructure support new bandwidth, it must methods of supervision and surveillance, as
also support new services, with reduced carriers have never managed this scale of optical
provisioning intervals and lower capital and connectivity.
operational costs. Together, these capabilities
make up what are known as “intelligent optical
networks.” New systems and subsystems are MEMS-BASED OPTICAL COMPONENTS
making use of the properties of MEMS to enable ptical networking systems manufacturers are
carriers to build these networks. O now using MEMS in a variety of components,
beyond the large, scalable optical switches that
This paper outlines several technologies based on
MEMS components, including optical switches, were seen in previous generations of equipment.
attenuation and equalization devices, and tunable MEMS components can be used to perform a
components. The applications and drivers for variety of functions, including redirecting,
deployment of these devices and systems are reflecting, and attenuating light. The systems
detailed. Included in the discussion are the made up by these components can be divided into
challenges, issues, and risks that network several subsets: core optical switches, automated
operators face in deploying these technologies. fiber management platforms, variable optical
attenuators (VOAs), and reconfigurable optical
add-drop multiplexers (ROADMs).
OPTICAL NETWORK MARKET DRIVERS
ith bankruptcy, consolidation, and Core Optical Switches
W overcapacity woes in carriers’ pasts, the
former glut of optical capacity has disappeared.
The initial application of MEMS components,
core optical switches, is still relevant; however,
Shifts in carriers’ business drivers and operating the opportunities for deployment are reduced
models have changed their requirements for from the initial projections. Optical switches can
optical equipment; no longer is there an outcry for enable extremely high bandwidth connections
larger and larger systems, handling endless and services, supporting high-performance
amounts of capacity. Instead, carriers are adopting applications such as high-speed simulation,

38 Bechtel Telecommunications Technical Journal

visualization, and grid computing (super-
computing technology using the resources of
many disparate networked computers).
Photonic switches use MEMS-based core
materials to provide all-optical light switching.
Tiny reflective components, resembling mirrors,
are adjusted to steer an optical signal. MEMS
switch cores come in two designs: 2D MEMS,
where the mirrors are arrayed on a single level
(and therefore can be adjusted only in two
dimensions), and 3D MEMS, where the mirrors
are on multiple planes. More flexible and scalable
than the 2D systems, 3D MEMS allow for more
lightpaths through the switch. However,
3D MEMS are more complex and costly than the
generally smaller and easier-to-manufacture
2D design [3]. These devices are usually referred
to as A x A in size, where A is the number of input
and output ports. Thus, a 32 x 32 switch can direct
any of 32 input signals to any of 32 output signals.
Due to their complexity, 3D MEMS devices
typically support much larger switch core sizes.
Figure 2a. 2D MEMS Operation [4]
Figure 2b. 3D MEMS Operation [4]
January 2006 • Volume 4, Number 1
Figure 2 illustrates the operation of 2D and 3D
MEMS switches.
Photonic switches can provide significant cost
savings over traditional OEO optical switches, as
a certain percentage of traffic at any node is
“express” traffic, i.e., it does not terminate locally.
In a traditionally designed OEO switch, this
traffic would be regenerated electrically before
being transmitted to its next destination. Photonic
switches can direct this traffic to the proper
interface without regeneration. The traffic
destined for local equipment can be connected to
a smaller OEO switch, sized just to handle the Photonic switches
local traffic, as shown in Figure 3. This hybrid can ensure that a
implementation can offer lower capital expenses
carrier’s investment
and a lower cost of ownership by reducing space
and power consumption. will be viable for a
longer term.
Another driver for photonic and hybrid
OEO/photonic switch applications is the
“agnostic” feature of photonic switches. A
photonic switch will direct a stream of light,
regardless of the number of wavelengths
or “color” of that stream, the underlying
protocol, the bitrate, etc. As such, photonic
switches are considered more future-proof, as
different OEO devices are required for different
wavelengths, different protocols, different
bitrates, etc. When new protocols or higher
bandwidth services are deployed, existing OEO
infrastructure can become worthless, requiring
new equipment to be purchased. Photonic
switches can ensure that a carrier’s investment
will be viable for a longer term.
Photonic switches also enable the framework for
generalized multiprotocol label switching
(GMPLS). GMPLS is an extension of the signaling
protocols of MPLS to lower-layer entities in the
network, including optical and physical layer
devices. GMPLS-enabled photonic switches allow
automated provisioning and bandwidth-on-
demand services, as well as new services like
optical virtual private networks.
39
 GMPLS-enabled
photonic switches
allow automated
provisioning and
bandwidth-on-
demand services, as
well as new services
like optical virtual
private networks.

Figure 3. Photonic Switch Integration Used to Terminate Local Traffic

Although the applications and drivers for
photonic switches are strong, several pitfalls have
hindered their deployment. Carriers must
consider the loss across the fabric of the switch.
Each component does have a finite loss, and this
loss must be carefully calculated into the link
budget of the optical path. This often has the
unintended consequence of requiring additional
amplification, or worse, OEO regeneration—a
problem for designers, who had sought to avoid
the need for OEO regeneration in the first place!
The complexity of MEMS-based devices is also an
issue—each mirror needs very complex
electronics to drive its movement and stabilize its
position. These devices add cost to the system as
well, counteracting some of the potential cost
replaced quickly has hindered deployment.
Finally, carriers have cited concerns about dust,
dirt, and even resiliency in the face of vibration,
such as in an earthquake zone. Several manu-
facturers of optical components are countering
this claim by certifying their equipment with
standards institutes or organizations, such as
Telcordia® Technologies, Inc.
Table 1 summarizes the similarities and
differences between optical and photonic
switches.
Table 1. Optical Versus Photonic Switches
Optical Photonic
Characteristic
Switch Switch

savings of the technology. Another complexity Cost per port Higher Lower
consideration is the finite switching time
involved in repositioning the reflective Limited by size of
Capacity Virtually unlimited
switch core
components. This time must be carefully
considered when examining applications, as the Granularity Circuit Wavelength
switching for these devices is not “wire speed”
Wire-speed,
(data processed or switched at its native rate). Switching/Routing Out-of-band
processes signals

Complexity is also cited as an issue in terms of Deployed in large

Maturity Limited deployments
numbers
durability. While the photonic core is future-
proof, concern that the equipment controlling Limited by granularity
Flexibility Future-proof
of core
the core configuration may need to be

40 Bechtel Telecommunications Technical Journal

Automated Fiber Management Platforms
Photonic switches can be used for more than just
core optical nodes; a new application is emerging
for MEMS-based switching systems. As carriers
deploy FTTX technologies and continue to
extend the reach of their optical infrastructure,
they face the challenge of managing literally
hundreds of fiber interfaces in a single location.
Managing these interfaces is typically done
manually via an optical distribution frame (ODF)
or patch panel. Manual patch cords are installed
by a technician, placing the proper signal on the
proper interface. Any moves, additions, or
changes must be done manually, and
configuration records are often developed by
hand as well. Testing of fiber cables terminated
on the ODF panel is also done manually, and a
costly error can occur if the technician
accidentally tests an in-service fiber pair!
This manual work is costly and labor-intensive
and also introduces many opportunities for
errors. In the case of an unmanned site,
repairs, testing, and provisioning cannot occur
until a technician arrives on site. This delay can
increase the mean time to repair or time to
turn-up a new circuit.
Automated fiber management systems make use
of MEMS components to mechanize this process.
Traffic can be switched or rerouted electronically
and remotely. Since signals do not need to be
switched dynamically or in real time, lower speed
(and lower cost) components are used.
Automated fiber management platforms also
enable advanced features. They allow for a circuit
layout inventory to be developed electronically,
so carriers can monitor exactly what signal or
circuit is riding on what fiber, and when and
Figure 4. Automated Versus Manual Fiber Management
January 2006 • Volume 4, Number 1
where it takes place. This can virtually eliminate
the outages that result when the “wrong” fiber
pair is disconnected. In addition, remote,
automated testing as well as test equipment
consolidation is made possible. A single piece of
test equipment can be used in one site, and the
automated fiber management platform can be
used on the tested fibers to direct the signals from
that site to the correct strands [5].
While automated fiber management is a potential
beneficial application for carriers, several
obstacles may hinder its deployment. ODF panels
Systems
are entirely passive components—the only active
manufacturers will
component is the technician who is cabling
them—while these platforms are active. This have to justify a
means that the system will require power and large gap in capital
telemetry and that, typically, some experience in
savings with a much
using its management platform will be necessary.
lower cost of
Carrier personnel will require training in using
this equipment as opposed to the ODF panel, ownership, or
which requires a very low skill level. reduce that gap
Figure 4 highlights the differences between with maturing,
automated and manual fiber management. mass-produced
A final hurdle is deployment cost. While carriers technology.
are eager to reduce operational expenses, which
clearly is possible with this platform, they are
unlikely to pay a premium for equipment
functions to do so. Systems manufacturers will
have to justify a large gap in capital savings with
a much lower cost of ownership, or reduce that
gap with maturing, mass-produced technology.
VOAs
OAs are components that allow the
V attenuation of selected optical signals or
wavelengths. These components are integral in
41
 dense wavelength division multiplexing
(DWDM) systems, as they are used to
dynamically compensate for skewed gains in
network wavelength amplification. The
amplifiers used in typical DWDM systems
have varying degrees of “flatness”—signals of
certain wavelengths are amplified more or
less than signals of other wavelengths. After
several chains of amplification, the wavelengths
receiving the most amplification may be
powerful enough to saturate the receive
electronics, while the wavelengths receiving the
Potential least amplification are just strong enough to
applications for register at the receiver. VOA components can
then attenuate only the most powerful
MEMS-based
wavelengths, bringing the entire signal into
components in conformance with DWDM receiver specifications.
distribution
VOA components are also crucial in protection
environments switching functions, as a cable cut can reduce the
include amplifiers number of aggregate wavelengths on a fiber.
Optical amplifiers have a fixed amount of
for cable television
amplification, which is typically linked to
systems, and gain
amplifier pump laser current; the amplification
control devices for is then spread across all of the relevant
active FTTX wavelengths. If a broken or disconnected fiber
results in the amount of wavelengths being cut in
deployments.
half, the remaining wavelengths will be amplified
twice as much. This effect is often cascaded
through multiple amplifiers; the end receiver
component may then receive a signal that is
significantly higher than the component’s
operating specification for optical power,
rendering the received signal unintelligible. VOA
components can prevent failure during a
protection switching event by attenuating the
offending wavelengths until they are in the
operating range for receiver input power.
MEMS-based devices use tuned components that
can be adjusted, typically by raising or lowering
them, to accomplish attenuation by partially or
completely blocking a stream of light. These
components can be thought of as being very
similar to a mechanized camera shutter, albeit on
a much smaller scale. Multiple actuators can be
used to provide a very highly resolved range of
attenuation; this arrangement is potentially
superior to alternative optical attenuation
technologies. Figure 5 illustrates a MEMS-based
VOA. The cylindrical items are two optical fibers;
attenuation is achieved by raising a metallic
shutter in the gap [6].
As they are significantly smaller and less power-
hungry than competing technologies, MEMS-
based devices do not require temperature
compensation, which affords several advantages.
Temperature control requires additional
42
Figure 5. MEMS-Based VOA [6]
electronics, which adds cost, complexity, power
consumption, and real estate requirements to the
component. Temperature control requirements
also typically prevent non-MEMS-based VOA
components from being deployed outside
the central office or other controlled,
telecommunications-oriented environment. On
the other hand, MEMS-based VOA components
can be deployed in a wider variety of
environments, including outside plant and access
networks. Potential applications for MEMS-based
components in distribution environments include
amplifiers for cable television systems, and gain
control devices for active FTTX deployments.
For these reasons, MEMS VOA components are
experiencing widespread market acceptance;
several design wins have been announced in
recent months. Systems designers must consider
other issues in addition to the cost and
complexity of this particular subsystem. The
electronics required to control and drive these
devices must also be assessed and taken into
account as part of an entire design. When
evaluating a particular design or component, the
system architect must also examine items such as
scalability. Unlike wideband optical components
such as optical switches, these devices may have
to attenuate individual wavelengths or small
bands of wavelengths. Systems designers must
ask themselves: Will one device be sufficient as
larger numbers of wavelengths are deployed?
ROADMs
OADMs have received significant attention
R from carriers in recent months. These
components allow for remote provisioning and
management of DWDM systems. Selected
wavelengths can be transparently transported or
serviced—added to and dropped from client
equipment—at particular nodes. Changes can
Bechtel Telecommunications Technical Journal
be made remotely and in real time via an
integrated control plane such as GMPLS. These
components hold great implications for
improving service velocity and reducing
operational expenses. Figure 6 illustrates the
functionality of a ROADM [7].
In addition to these advantages, several ROADM
designs offer further benefits, such as remote
wavelength testing, dynamic power balancing
(also known as equalization), and remote signal
surveillance. The dynamic power balancing
feature of a ROADM may seem very similar to
the features of a VOA; this is not entirely
coincidental. VOAs are one of the components
that make up a typical ROADM.
Two types of component subsystems are used
in ROADMs—wavelength blockers and
wavelength-selective switches. Wavelength
blockers are devices offering the capacity to
add/drop wavelengths and that have some
tunability—the ability to add/drop selected
wavelengths. Wavelength-selective switches are
more widely tunable and flexible. Most
wavelength-selective switches can be configured
to add/drop or pass through any number or
combination of wavelengths at a node, allowing
for more modularity and deployment ease while
reducing spacing requirements. It should be
noted that this flexibility typically comes at a
price premium [8].
ROADM technologies were first deployed in
metropolitan networks, where channel counts are
low and wavelength spacing is accordingly high.
Integrated silicon components, using no moving
parts, perform well for these applications. As
Figure 6. ROADM Subsystem Functionality [7]
January 2006 • Volume 4, Number 1
metropolitan networks grow, systems designers
typically increase wavelength counts by
interleaving channels in between existing
channels, allowing for retention of amplifier
components. This approach presents a problem
for silicon-based ROADMs, as flatness
requirements for these interleaved systems are
generally higher. Additionally, ROADM
applications are being proposed for long-haul
networks. Long-haul networks typically feature
higher capacity, higher channel count
wavelength division multiplexing (WDM)
ROADM designs
systems, which present the same challenges as the
interleaved metropolitan systems [9]. offer further
MEMS-based wavelength-selective switches are benefits, such as
emerging that provide solutions to this flatness remote wavelength
problem. Subsystems are based on either testing, dynamic
diffractive MEMS components or MEMS mirrors.
power balancing
Diffractive MEMS components use arrays of
(also known as
piezoelectric materials suspended over a
substrate; in this condition, the device “looks” equalization), and
like a mirror to light. Applying a voltage to the remote signal
materials causes them to move in the direction of
surveillance.
the substrate and thus become a diffraction
grating, attenuating a specific frequency of light.
Arrays of MEMS materials can be used to
selectively reflect or attenuate specific channels.
Diffractive MEMS do not require contact to
actuate (instead relying on applied voltage) and
are relatively simple to manufacture. They are
also more reliable than mirror-based MEMS, due
to the lack of moving parts, and can be packaged
in a smaller subsystem. The only limitation is
scalability—arrays of tuned components become
more complex by orders of magnitude when
channel counts increase.
43
 Mirror-based MEMS subsystems, long used for
optical switches, can be built for much higher
channel-count applications. Designers have to
contend with the same challenges found in
optical switches—cost, complexity, reliability,
and size. Systems designers are now developing
hybrid techniques in an attempt to take
advantage of the positives of both technologies.
Several manufacturers now offer hybrid
diffractive MEMS/mirror MEMS wavelength
selective switches.
To achieve an While optical systems using ROADM
technologies are an enabler for scalable,
optimum network
dynamically reconfigurable networks, systems
design, network designers must pay additional attention to
engineers must implementation details. Granular power
measurement and control are required for each
continue to use
wavelength. Fault isolation at the wavelength
strong planning
and component level is a necessity, since a single
methodologies, misdiagnosed fault can cause networkwide
combined with instability. Additional hardware and software
required to perform these functions can entail
detailed knowledge
additional cost, real estate, and complexity. As
of ideal applications
such, ROADM technologies may hold a “first-
for various cost” penalty for smaller, lower channel-count
technologies and networks when compared to more mature
the associated technologies. Carriers must consider the growth
forecast and timeframe when evaluating choices
cost-versus-benefit
for new optical networks.
tradeoffs.
CONCLUSIONS
fter much early hype, MEMS-based optical
A devices are experiencing continued
deployment in next-generation optical networks.
These components enable a number of advanced
networking features, such as dynamic
reconfigurability, while potentially reducing
power consumption, network real estate, and
capital and operational expenses. The benefits
offered by systems using these components often
come at a cost, however. Also, not every
application or deployment is suitable for MEMS-
based systems. To achieve an optimum network
design, network engineers must continue to use
strong planning methodologies, combined with
detailed knowledge of ideal applications for
various technologies and the associated cost-
versus-benefit tradeoffs.

TRADEMARKS
Telcordia is a registered trademark of Telcordia
Technologies, Inc., in the United States, other
countries, or both.
44
REFERENCES
[1] T. Freeman, “MEMS Devices Put Their Stamp on
Optical Networking,” Fibre Systems Europe,
September 2004.
[2] J. Ford, J. Walker, and K. Goosen, “Optical
MEMS: Overview and MARS Modulator,”
Presentation made by Lucent Technologies.
[3] Goldman Sachs and McKinsey & Company, “US
Communications Infrastructure at a Crossroads:
Opportunities Amid the Gloom,” August 2001.
[4] M. Fernandez and E. Kruglic, “MEMS Technology
Ushers in a New Age in Optical Switching,”
Lightwave, August 2000.
[5] C. Matsumoto, “Calient Patches Its Strategy,”
Light Reading, February 14, 2005
(http://www.lightreading.com/document.asp?
site= lightreading&doc_id=68031).
[6] D. Horsley, “Image Gallery”
(http://mae.ucdavis.edu/faculty/
horsley/photoalbum.html).
[7] “ROADM Architectures and Implementations,”
Metconnex
(http://www.metconnex.com/Products-
page/roadm.htm).
[8] C. Matsumoto, “ROADMs: Almost Famous,”
Light Reading, May 20, 2005
(http://www.lightreading.com/document.asp?
doc_id=74210).
[9] C. Matsumoto, “ROADMs Roll On,”
Light Reading, June 7, 2005
(http://www.lightreading.com/document.asp?
doc_id=75148&site=supercomm).
BIOGRAPHY
Brian Coombe joined Bechtel
Telecommunications in 2003.
Currently, he is a systems
engineer for the federal
telecommunications group
within Bechtel National, Inc.,
Bechtel’s government sub-
sidiary. He designs and
analyzes carrier, large-scale
enterprise, and government and
military communications systems.
Previously, as a lead Bechtel Telecommunications
engineer for the Verizon fiber-to-the-premises program,
Brian managed the planning and network design of two
wire centers. His Bechtel experience also includes work
as an RF engineer providing design solutions for the
AT&T Wireless GSM overlay program.
Before joining Bechtel, Brian was a systems engineer
and applications specialist at Tellabs®, an optical
networking equipment manufacturer. He was a
member of the team that launched Tellabs’ DWDM
platform and supported Tellabs’ first DWDM customer.
Brian received a BS with honors in Electrical
Engineering from The Pennsylvania State University.
He is a member of the Institute of Electrical and
Electronics Engineers and also Eta Kappa Nu, the
national electrical engineering honor society. Brian
authored “Reliable Electric Power Transmission and
Distribution under National Restructuring,” which
appeared in the IEEE/WISE Journal of Engineering and
Public Policy in August 1999.
Bechtel Telecommunications Technical Journal
 I N V I T E D PA P E R

PHY/MAC CROSS-LAYER ISSUES

IN MOBILE WiMAX

Issue Date: January 2006

Abstract—After the IEEE 802.16-2004 standard was published, much attention was drawn to providing
broadband access in rural and developing areas over fixed wireless channels. Now, the IEEE 802.16e standard
for Mobile WiMAX is about to be published. It is known that Mobile WiMAX will incorporate error-correction
capability and will be an enhanced version of the IEEE 802.16 standard with mobility support. Therefore, it is
expected that Mobile WiMAX will not only compete with the broadband wireless market share in urban areas
with DSL, cable, and optical fibers, but also threaten the hot-spot-based Wi-Fi™ and even the voice-oriented
cellular wireless market. This paper first provides an overview of Mobile WiMAX, especially on OFDMA/TDD
systems. Then, the paper addresses some PHY/MAC cross-layer issues that need to be resolved through radio
resource management to increase throughput, cell coverage, and spectral efficiency.

INTRODUCTION data rates of 5 Mbps or better. Hence, Mobile

EEE 802.16e, the standard for Mobile WiMAX,
I is expected to be published by the end of 2005,
and Mobile WiMAX service will launch in South
Korea and possibly in North America in late 2006
or early 2007. Originally, the IEEE 802.16
standard was developed for fixed wireless in the
search for a new tool to allow homes and
businesses to link with the worldwide core
networks. It was envisioned that the IEEE 802.16
standard would offer a better solution for last-
mile connections, compared to fiber, cable, or
digital subscriber line (DSL) links, because
wireless systems are less costly to deploy over
wide geographic areas. The publication in
June 2004 of the IEEE 802.16d (IEEE
802.16-2004) standard provided assurance that
the WiMAX market and its competitiveness for
non-line-of-sight (NLOS) wireless broadband
access are maturing [1–4].
Jungnam Yun, PhD Since the IEEE 802.16e standard (the mobile
POSDATA America version of the IEEE 802.16-2004 standard) will be
R&D Center published soon, the focus of WiMAX is expected
jnyun@posdata-usa.com to shift from fixed subscribers to mobile
subscribers with various form factors: personal
Professor Mohsen digital assistant (PDA), phone, or laptop. The
Kavehrad, PhD IEEE 802.16e standardization group promises to
The Pennsylvania support mobility at speeds of up to
State University
(CITCTR) approximately 40 to 50 mph; some vendors are
mkavehrad@psu.edu already claiming success in testing prototype
systems with mobility at speeds over 50 mph for
WiMAX is not only expected to compete with
other last-mile connections such as fiber, cable,
and DSL; it also threatens Wi-Fi™ and code
division multiple access (CDMA) voice
communications with voice-over-Internet-
Protocol (VoIP) services.
Unlike wired networks, wireless networks are
highly dependent on communications channels;
radio channels are dynamic, correlated,
unreliable, and very expensive. This is why
performance will be highly dependent on how
well radio resource management supports
quality-of-service (QoS) requirements, even if
QoS might be a luxury in the early stages of the
Mobile WiMAX market. Therefore, several cross-
layer issues between the medium access control
(MAC) layer and the physical (PHY) layer need to
be optimally resolved on the radio resource
management side of Mobile WiMAX systems.
In multiuser environments, especially on wireless
fading channels, multiuser diversity is a key radio
resource management element for maximizing
throughput. Multiuser diversity is a form of
selection diversity. Since different users
experience independent time-varying fading
channels in wireless networks, resources are
allocated to the user with the best channel quality
to maximize system throughput. Multiuser
diversity has drawn attention since tracking user
channel fluctuations is becoming more accurate

© 2006 Bechtel Corporation. All rights reserved. 45

 ABBREVIATIONS, ACRONYMS, AND TERMS
AAS adaptive antenna system MPEG Moving Picture Experts Group
AMC adaptive modulation and coding MS mobile station
ARQ automatic repeat request MSB most significant bit
ASCA adjacent subcarrier allocation NLOS non-line of sight
ATM asynchronous transfer mode nrtPS non-real-time polling service
BE best effort OFDM orthogonal frequency division
multiplexing
BS base station
OFDMA orthogonal frequency division
CC convolutional coding
multiple access
CDMA code division multiple access
OFUSC optional full usage of
CP cyclic prefix subchannels
CQI channel quality indicator OPUSC optional partial usage of
subchannels
CRC cyclic redundancy code
PDA personal digital assistant
CS convergence sublayer
PDU protocol data unit
CTC convolutional turbo coding
PHY physical
DC direct current
PMP point to multipoint
DCD downlink coding descriptor
PN pseudorandom noise
DHCP dynamic host configuration
protocol PUSC partial usage of subchannels
DL downlink QoS quality of service
DLFP DL frame prefix QAM quadrature amplitude
modulation
DSCA distributed subcarrier allocation
QPSK quadrature phase shift keying
DSL digital subscriber line
RTG receive/transmit transition gap
D-TDD dynamic TDD
rtPS real-time polling service
ertPS extended real-time polling service
SDU service data unit
FEC forward error correction
SINR signal-to-interference+noise
FBSS fast BS switching
ratio
FCH frame control header
SISO single input, single output
FDD frequency division duplex
SNR signal-to-noise ratio
FRF frequency reuse factor
STC space-time coding
FUSC full usage of subchannels
TDD time division duplex
HARQ hybrid ARQ
TSA time slot allocation
H-FDD half-duplex FDD
TTG transmit/receive transition gap
HO handover
TUSC tile usage of subchannels
HT header type
UBC uplink coding descriptor
IR incremental redundancy
UGS unsolicited grant service
LDPC low density parity check
UL uplink
LSB least significant bit
VoIP voice over Internet Protocol
MAC medium access control
Wi-FiTM wireless fidelity
MAN metropolitan area network
WiMAX worldwide interoperability for
MAP mobile application part microwave access
MIMO multiple input, multiple output
MISO multiple input, single output

46 Bechtel Telecommunications Technical Journal

and faster. Hence, diversity gain increases when
the dynamic range of the fluctuation increases,
but the gain is limited in environments with slow
fading [5]. In slow fading, multiuser diversity
hardly satisfies all QoS parameters at the same
time, especially fairness among users. Ultimately,
radio resource management needs to implement
a combined form of multiuser diversity and
fairness scheduling [6].
As broadband wireless networks encompass
various services such as World-Wide Web
(www), voice, video, and data, network traffic
becomes very dynamic and unbalanced between
the uplink (UL) and downlink (DL) stream
volumes. To provide the highest transport
efficiency in broadband networks, time division
duplex (TDD) is preferred over frequency
division duplex (FDD) because it offers more
flexibility in changing the UL and DL band-
width ratio according to the dynamic traffic
pattern [7]. It has been assumed that the
switching points for the UL and DL of TDD
schemes are determined by network operators
and are not changeable. Moreover, switching
points in adjacent cells have been synchronized
to avoid severe inter-cell interference. Having the
same switching points in adjacent cells, a
centralized controller could set switching points
for all cells by observing the traffic characteristics
[8]. For the duplex scheme, FDD, TDD, and half-
duplex FDD (H-FDD) are all available options in
Mobile WiMAX. This paper addresses only the
TDD mode for the orthogonal frequency division
multiple access (OFDMA) PHY layer. Even
though it seems too early to adopt dynamic
changes in the DL/UL ratio for Mobile WiMAX
systems, several investigations show that
appropriate time slot allocation (TSA) and beam
forming can bring higher spectral efficiency in
dynamic TDD (D-TDD) systems compared to
conventional TDD systems [9, 10, 11].
This paper first describes the general PHY
and MAC layers and then provides views on
several cross-layer issues related to Mobile
WiMAX, specifically focusing on the point-to-
multipoint (PMP) mode of OFDMA/TDD
systems, multiuser diversity, zone adaptation for
interference cancellation, hybrid automatic
repeat request (HARQ), and variable DL/UL
ratio or D-TDD.
PHY LAYER IN OFDMA/ TDD MOBILE WiMAX
he three different PHY layers in Mobile
T WiMAX are single carrier, orthogonal
frequency division multiplexing (OFDM), and
January 2006 • Volume 4, Number 1
OFDMA. The WiMAX Forum’s Mobile Task
Group is developing a Mobile WiMAX profile
based on the OFDMA PHY layer only [1–4]. At
the same time, the WiMAX Forum’s Evolution
Task Group is developing technical specifications
for the evolution of OFDM-based networks from
fixed to nomadic connections. Hence, the
OFDMA PHY layer will be the baseline for
Mobile WiMAX.
OFDMA
The wireless metropolitan area network (MAN)-
OFDMA PHY layer based on OFDM modulation OFDMA can
is designed for NLOS operation in the frequency fully use multiple
bands below 11 GHz. OFDMA inherits OFDM’s user channel
immunity to intersymbol interference and
variations via
frequency selective fading.
two-dimensional
An inverse Fourier transform may be used to
resource
synthesize the OFDMA waveform during a
symbol time. A small end-portion of the symbol allocation.
time, called a cyclic prefix (CP), is copied at the
beginning of the symbol time duration to collect
multipath while maintaining orthogonality
among the subcarriers.
Within the OFDMA symbol time frame, the
active subcarriers are split into subsets of
subcarriers; each subset is termed a subchannel.
On the DL, a subchannel may be intended
for different (groups of) receivers. On the UL,
a transmitter may be assigned one or
more subchannels and several transmitters
may transmit simultaneously. The subcarriers
forming one subchannel may, but need
not, be neighbors. A slot—the minimum
possible OFDMA data allocation unit—
consists of one or more symbols in the time
domain by one subchannel in the frequency
domain. Hence, OFDMA can fully use multiple
user channel variations via two-dimensional
resource allocation.
TDD
The three different duplex modes for OFDMA
Mobile WiMAX systems are TDD, FDD, and
H-FDD (see Figure 1). TDD systems use the same
frequency band for DL and UL, and the frame is
divided into DL subframes and UL subframes in
the time domain. FDD systems use different
frequency bands, and DL and UL subframes are
overlapped in the time domain. H-FDD systems
have two different frequency bands for DL and
UL, and DL and UL subframes do not overlap in
the time domain.
Mobile WiMAX has an optional channel-
sounding feature for TDD systems. Channel
47
 Figure 1. Duplex Modes TDD, FDD, and H-FDD
sounding is a signaling mechanism that
There are
enables the base station (BS) to estimate BS-to-
two main types mobile station (MS) channel response based on
of subcarrier the UL signals transmitted by the MS. Channel
allocation sounding works only under the assumption of
TDD reciprocity.
techniques:
distributed and Due to channel reciprocity and DL/UL ratio
adaptability, TDD is the most favored duplex
adjacent.
mode in Mobile WiMAX [7]. It is the only mode
addressed in this paper.
OFDMA/TDD Frame Structure
Figure 2 shows an example of the OFDMA frame
structure for the TDD mode. Each frame is
divided into DL and UL subframes by
transmit/receive transition gaps (TTGs) and
receive/transmit transition gaps (RTGs). Each DL
subframe has a preamble in the first OFDMA
symbol and then starts with the frame control
header (FCH) in the second symbol. The FCH
specifies the subchannel groups used for the
segment, the burst profile, and the length of the
DL-mobile application part (MAP) message,
which directly follows the FCH. The UL-MAP
message is carried by the first burst allocated in
the DL-MAP. Each UL subframe may have one or
more ranging slots, which are used for the
network entry procedure. UL subframes may
have fast feedback channels for fast channel
quality indicator (CQI) reports or other fast
operational requests or responses. (Fast feedback
channels are not shown in Figure 2.)
Subcarrier Allocation
There are three types of subcarriers: data, pilot,
and null. Data subcarriers are used for data
transmissions, pilot subcarriers are used for
channel estimation and various synchronization
purposes, and null subcarriers are used for the
direct current (DC) carrier and guard bands
transmitting no signals. Multiple data sub-
carriers are grouped into a subchannel, and a
subchannel forms a slot with one or more
OFDMA symbols. A slot is a channel and MAP
allocation unit; it contains 48 data subcarriers.
48
There are two main types of subcarrier allocation
techniques: distributed and adjacent. In general,
the distributed allocations perform very well in
mobile applications, while adjacent subcarrier
permutations can be properly used for fixed,
portable, or low mobility environments. These
options enable the system designers to trade
mobility for throughput.
Distributed Subcarrier Allocation
In a distributed subcarrier allocation (DSCA),
multiple data subcarriers are grouped into a
subchannel. Although subcarriers in a
subchannel are not usually adjacent to each other,
they may be in some cases. DSCAs for the DL are
DL-partial usage of subchannels (PUSC), full
usage of subchannels (FUSC), optional FUSC
(OFUSC), and tile usage of subchannels (TUSC)
1 and 2. DSCAs for the UL are UL-PUSC and
UL-optional PUSC (OPUSC).
• DL-PUSC: The default DL subcarrier
allocation method. All DL subframes start in
the DL-PUSC zone. Subchannels are grouped
into six major groups and assigned to three
segments (three sectors) of a cell. Assigning
two major groups to each segment, the cell
can be viewed as frequency reused by a
factor of three. By switching to a DL-PUSC
zone that assigns all subchannel groups to
each segment, the cell can realize a frequency
reuse factor of one. DL-PUSC is designed to
minimize the probability of using the same
subcarrier in adjacent sectors or cells.
• FUSC: Uses all subchannels and minimizes
the performance degradation of fading
channels by frequency diversity. FUSC is also
designed to minimize the probability of
using the same subcarrier in adjacent sectors
or cells. FUSC pilots are in both variable and
fixed positions.
• OFUSC: Also designed to fully use
frequency diversity. One difference from
FUSC is that OFUSC uses a bin structure
like band adaptive modulating and coding
(AMC).
Bechtel Telecommunications Technical Journal
 • TUSC: For use in the adaptive antenna
system (AAS) zone; similar in structure to
UL-PUSC.
• UL-PUSC: The default UL subcarrier
allocation method. It is not necessary to start
the UL subframe in the UL-PUSC zone.
UL-PUSC has a tile structure, with each
tile comprising four subcarriers by three
symbols. The four corner subcarriers are used
as pilots, and the remaining eight subcarriers
are used as data subcarriers.
• OPUSC: Also has a tile structure, with each
tile comprising three subcarriers by three
symbols. The center subcarrier is used as a
pilot, and the remaining eight subcarriers are
used as data subcarriers.
On the DL side, DL-PUSC with all subchannel
groups performs similarly to FUSC and to DL-
PUSC with partial subchannel groups, which can
avoid co-channel interference by deploying a
frequency reuse factor of three. On the UL side,
UL-PUSC, with its four pilot subcarriers, has a
better channel estimation performance than
OPUSC. However, OPUSC has more data slots
than UL-PUSC because it has a smaller tile size
with the same number of data subcarriers.
Adjacent Subcarrier Allocation
Band AMC and PUSC-adjacent subcarrier allo-
cation (ASCA) are ASCA techniques. While
DSCAs can gain frequency diversity in frequency
Figure 2. Example of the OFDMA Frame Structure for the TDD Mode
January 2006 • Volume 4, Number 1
selective slow fading channels, ASCAs can gain
multiuser diversity in frequency non-selective
fading channels. In the adjacent subcarrier
permutation, symbol data within a subchannel is
assigned to adjacent subcarriers, and the pilot and
data subcarriers are assigned fixed positions in
the frequency domain within an OFDMA symbol.
This permutation is the same for both the UL
and DL.
• Band AMC: In defining a band AMC
allocation, a bin—the set of nine contiguous
subcarriers within an OFDMA symbol—is
While DSCAs
the basic allocation unit on both the DL and
UL. A group of four rows of bins is called a can gain
physical band. An AMC slot consists of six frequency diversity
contiguous bins in the same band, and four in frequency
types of AMC slots are defined in the IEEE
selective slow
802.16-2004 standard. But in Mobile WiMAX,
fading channels,
only one type of slot, defined as two bins by
three symbols, is used. ASCAs can gain
• PUSC-ASCA: PUSC-ASCA uses distributed multiuser diversity
clusters for the PUSC mode. The symbol in frequency
structure uses the same parameters as those
non-selective
of the regular PUSC, and the same cluster
structure is maintained; only the subcarrier fading channels.
allocation per cluster is different from that of
the regular PUSC.
Adjacent subcarrier allocations are preferred in
the AAS zone.
49
 Ranging interleaving, and modulation. A pseudorandom
For the purposes of network entry, connection noise (PN) sequence generator is used to
maintenance, bandwidth request, and efficient randomize each FEC data block. Multiple FEC
handover (HO), Mobile WiMAX provides types are available for encoding randomized
ranging channels with CDMA-like signaling. A data: tail-biting convolutional coding (CC), zero-
maximum of 256 sets of 144-bit pseudo-noise tailed CC, convolutional turbo coding (CTC), and
ranging codes are generated and divided into low density parity check (LDPC). Among these
four groups: initial, periodic, bandwidth request, channel coding schemes, tail-biting CC and CTC
and HO ranging. are mandatory; the others are optional. CC is
used for the FCH DL frame prefix (DLFP) and is
One or more groups of six subchannels (for
mandatory. Except for the FCH, it is highly likely
PUSC) or eight subchannels (for OPUSC)
that CTC will be used for all control information
constitute a ranging channel. For initial and HO
MIMO and AAS and data bursts. FEC encoded data is interleaved
ranging, two OFDMA symbols are used, and the
are seriously in two steps: the scattering step and the least-
same ranging code is transmitted on the ranging
significant-bit (LSB)/most-significant-bit (MSB)
being considered channel during each symbol, with no phase
switching step. For modulation in OFDMA
for implementation discontinuity between the two symbols. Hence,
systems, quadrature phase shift keying (QPSK)
initial and HO ranging have a wide range for the
in the early stage and 16 and 64 quadrature amplitude modulation
purpose of timing adjustments. Meanwhile,
to boost spectral (QAM) are available.
periodic and bandwidth request ranging are
efficiency and to transmitted over one OFDMA symbol because
active MSs are aligned mostly via frame time and Multiple Input, Multiple Output/Adaptive Antenna
promote success
the timing deviations are very small. System
in the broadband Since single input, single output (SISO) systems
wireless market. cannot achieve high spectral efficiency, multiple
HARQ
input, multiple output (MIMO) systems draw
HARQ greatly increases the data rate when the
much attention, and they are included in the
signal-to-noise ratio (SNR) is very low; hence, it
WiMAX profile as optional features. MIMO
increases the coverage of Mobile WiMAX
systems have various advantages over SISO and
systems. The major difference between
multiple input, single output (MISO) systems;
conventional ARQ and HARQ is that a
these include multiplexing gain, diversity gain,
conventional ARQ discards erroneous packets
interference suppression, and array gain. In a
when retransmitting lost and/or subsequent
highly scattering channel, transmitting inde-
packets, whereas an HARQ does not discard the
pendent data from different antennas increases
erroneous packets, but combines them with
capacity linearly. Also, there are receiver
retransmitted packets to gain time diversity. For
diversity gains with multiple receiver antennas
every MS, each packet transmission over a radio
and space-time coding (STC) gains with multiple
channel faces different channel characteristics.
transmitter antennas. As is true of smart antenna
Therefore, HARQ can also be used to maximize
systems, beam forming is also an advantage in
the throughput from the time diversity gain.
MIMO systems when channel information is
There are two types of HARQ: chase and available.
incremental redundancy (IR). For chase HARQ,
Even though MIMO and AAS are optional Mobile
each retransmission is identical to the original
WiMAX features, they are seriously being
transmission; hence, implementation complexity
considered for implementation in the early stage
is lower than for IR HARQ. Meanwhile, an IR
to boost spectral efficiency and to promote
HARQ transmits a different redundancy version
success in the broadband wireless market.
for different subpackets. An IR HARQ is flexible
in adapting the subpacket transmission rate
according to the most recent channel quality
MAC LAYER IN OFDMA/ TDD MOBILE WiMAX
feedback; this obviously has the potential of
he MAC layer of Mobile WiMAX provides a
achieving better performance than that of a chase
HARQ. However, chase HARQ using CTC is the T medium-independent interface to the PHY
layer and is designed to support the wireless PHY
preferred HARQ option because it is less complex
than IR HARQ using CTC. layer by focusing on efficient radio resource
management. The MAC layer supports both PMP
and mesh network modes; this paper focuses only
Channel Coding
on the PMP mode.
Mobile WiMAX has four channel coding steps:
randomization, forward error correction (FEC),

50 Bechtel Telecommunications Technical Journal

The MAC layer schedules data transmission
based on connections. Each MS creates one or
more connections having various service classes:
unsolicited grant service (UGS), real-time polling
service (rtPS), extended real-time PS (ertPS), non-
real-time PS (nrtPS), and best effort (BE) service.
The MAC layer is intended to manage radio
resources efficiently to support QoS for each
connection; to maintain link performance using
AMC, ARQ, and other methods; and to maximize
throughput. The MAC layer handles network
entry for the MS and creates the MAC protocol
data unit (PDU). Finally, the MAC layer provides
two convergence sublayer (CS) specifications:
asynchronous transfer mode (ATM) CS and
packet CS.
Service Classes
When created, each connection is assigned to a
certain service class based on the type of
QoS guarantees required by the application. The
IEEE 802.16e standard provides the following
service classes:
• UGS: Designed to support real-time service
flows that periodically generate fixed-size
data packets, such as T1/E1 and VoIP
without silence suppression.
• rtPS: Designed to support real-time service
flows that periodically generate variable-size
data packets, such as Moving Picture Experts
Group (MPEG) video.
• ertPS: A scheduling mechanism that builds
on the efficiency of both UGS and rtPS. The
BS provides unsolicited unicast grants as in
UGS, thus saving the latency of a band-
width request. However, UGS allocations
are fixed in size, whereas ertPS allocations
are dynamic.
• nrtPS: Offers regular unicast polls, which
ensures that the service flow receives request
opportunities even during network
congestion.
• BE: Intended to provide efficient service for
BE traffic. Typical BE service is Web surfing.
Network Entry
When an MS wants to enter the network,
it follows the network entry process: (1) down-
link channel synchronization, (2) initial ranging,
(3) capabilities negotiation, (4) authen-
tication message exchange, (5) registration,
(6) IP connectivity, and (7) periodic ranging
afterwards. The following paragraph amplifies
this process.
January 2006 • Volume 4, Number 1
(1) The MS first scans for a channel in the defined
carrier frequency list and detects its frame
synchronization, using the preamble at the PHY
layer. (2) Once the PHY level is synchronized, the
MS can obtain the DL-MAP, downlink coding
descriptor (DCD), and uplink coding descriptor
(UCD) for the DL and UL parameters. When the
MS has all parameters and information regarding
the UL ranging allocation, it starts sending a
CDMA ranging code, followed by several MAC
messages, and then adjusts timing and power
according to the BS command. (3) After initial
ranging is completed, the MS negotiates with the There are
BS regarding its modulation level, coding five service classes
scheme, MAP support, and other capabilities, so
in Mobile WiMAX:
that the BS knows exactly what the MS is capable
of and can allocate resources efficiently. (4) Once UGS, rtPS, ertPS,
capabilities are negotiated, the BS authenticates nrtPS, and BE.
the MS and sends important key material for data
ciphering. (5) After the MS is authenticated, it
finally registers onto the networks and starts the
dynamic host configuration protocol (DHCP) to
obtain the IP address and other parameters
needed to establish IP connectivity. (6) After that,
transport connections are made. The BS initiates
the establishment of pre-provisioned connections
while the MS initiates the establishment of non-
pre-provisioned connections. (7) Finally, the MS
conducts periodic ranging as needed.
MAC PDU Construction and Transmission
The MAC layer of Mobile WiMAX supports both
fragmentation and packing of MAC service data
units (SDUs) for ARQ-enabled and non-ARQ-
enabled connections. Also, multiple MAC PDUs
can be concatenated into a single transmission of
either DL or UL connections. For ARQ-enabled
connections, fragments are formed for each
transmission by concatenating sets of ARQ blocks
with consecutive sequence numbers. Even
though ARQ implementation is mandatory, ARQ
may be enabled on a per-connection basis.
Furthermore, a connection cannot have both ARQ
and non-ARQ traffic. A service flow may require
that a cyclic redundancy code (CRC) be added to
each MAC PDU carrying data for that service
flow. In this case, for each MAC PDU with header
type (HT) = 0, a CRC32 is appended to the
payload of the MAC PDU; i.e., request MAC
PDUs are unprotected. The CRC covers the
generic MAC header and the payload of the MAC
PDU. The CRC32 calculation methods for the
OFDM mode and the OFDMA mode are
different, which makes fixed wireless OFDM
systems using the IEEE 802.16d standard and
Mobile WiMAX (mobile OFDMA systems) using
the IEEE 802.16e standard incompatible.
51
 Packet Scheduling and Radio Resource
Management
The main goal of packet scheduling and radio
resource management is to maximize throughput
while satisfying QoS requirements.
The BS packet scheduler works closely with the
radio resource management entity to guarantee
QoS while maximizing throughput by efficiently
using the opportunistic characteristics of wireless
channels. Meanwhile, the MS packet scheduler
only schedules packets from the connection
queues into the transmission buffer so that the
The addition of
an HO scheme
MS can transmit packets when the BS allocates
bandwidth in certain frames.
makes Mobile
Radio resource management is the main task of
WiMAX much the scheduler, whose functions are to allocate
different from DL and UL bandwidth, construct MAPs in the
DL and UL subframes, and decide on the best
fixed broadband
burst profile for each connection. Bandwidth
wireless systems.
allocation and MAP construction should be done
jointly, and the burst profile should be
determined per connection beforehand, based on
the signal-to-interference+noise ratio (SINR)
report from each MS. For MAP construction, DL
and UL subframes can be divided by multiple
zones, i.e., a normal zone with multiple choices of
subcarrier allocation, STC, AAS, and MIMO,
based on how the optional features may be
implemented and used.
Handover
The addition of an HO scheme makes Mobile
WiMAX much different from fixed broadband
wireless systems. Seamless HO is a must when
the connection is for real-time service such as
VoIP. There are three types of HO: hard HO, fast
BS switching (FBSS), and macro-diversity HO.
Also, either the MS or the BS can initiate HO. The
first type of HO, hard HO, disconnects the MS
from the previous serving BS before the MS
connects with the target BS. The second type of
HO, FBSS, uses a fast switching mechanism to
improve HO quality. The MS can only
transmit/receive data to/from one active BS at
any given frame, and all active BSs should
be ready for downlink data for the specific MS at
any frame. For FBSS, all BSs should be
synchronized based on a common time source
and use the same frequency channel. The BSs are
also required to share or transfer MAC context
through networks. The third type of HO, macro-
diversity HO, allows one or more BSs to transmit
the same MAC/PHY PDUs to the MS so that the
MS can perform diversity combining. This macro-
diversity HO is also called a soft HO. Due to the
complexity of the macro-diversity HO, it is not
52
likely to be implemented in the early stage of
Mobile WiMAX service.
To prepare and expedite a potential HO in the
near future, the MS should be able to scan nearby
BS signals and associate with possible target BSs
to acquire and record ranging parameters and
service availability information.
PHY/MAC CROSS-LAYER ISSUES IN MOBILE
WiMAX
he challenges inherent in implementing
T Mobile WiMAX arise from its deployment
of a frequency reuse of one and its adoption of
many state-of-the-art technologies such as
HARQ, MIMO, and AAS. This section describes
some cross-layer issues that need to be addressed
in the radio resource allocation management of
Mobile WiMAX.
Zone Switch and Frequency Reuse
Since the OFDMA PHY layer has many choices
of subcarrier allocation methods, multiple zones
can use different subcarrier allocation methods
to divide each subframe. One benefit of using
zone switching is that different frequency reuse
factors (FRFs) can be deployed in a cell (or
sector), dynamically.
Figure 3 shows an example of deploying different
FRFs in one frame. For the first half of each frame,
the entire frequency band is divided by three and
allocated in each sector. For the second half of
each frame, the whole same frequency band is
used in each sector. The benefits of deploying
different FRFs in one frame are: (1) the FCH and
DL-MAP are highly protected from severe
co-channel interference; (2) edge users, who are
receiving co-channel interference from other
sectors in other cells, also have suppressed
co-channel interference; and (3) users around
the cell center have the full frequency band
because they are relatively less subject to
co-channel interference.
Multiuser Diversity
In the wireless multiuser environment, it is well
known that multiuser diversity is a very
important leveraging factor of resource allocation
management [5, 6, 12]. Each MS faces a different
fading channel; hence, radio resource
management can use multiuser diversity to
maximize system throughput. The difficulty lies
in the fact that radio resource allocation also
should satisfy fairness among subscribers.
Moreover, in slow fading, multiuser diversity
Bechtel Telecommunications Technical Journal

Figure 3. Zone Switch for FRF Change
hardly satisfies all QoS parameters at the same
time, especially fairness among users. Ultimately,
radio resource management should follow a
combined form of multiuser diversity and
fairness scheduling.
Figure 4 illustrates the deployment of multiuser
diversity with a band AMC zone. Similarly,
multiuser diversity can also be deployed with a
DSCA zone such as PUSC, FUSC, optional FUSC,
and TUSC. The only thing different from a band
AMC zone is that the multiuser diversity gain can
be obtained only from time domain allocations.
Dynamic TDD Usage
Each MS and BS experiences not only different
channel characteristics, but also various data
traffic. In other words, UL and DL stream
volumes that have been considered symmetrical
for conventional voice transmissions are
unbalanced, and the ratio is time varying. To
provide the highest transport efficiency in
broadband networks, TDD is preferred to FDD
because it enables real-time adaptation of UL and
DL bandwidth according to the dynamic traffic
pattern. Even though FDD can also be used for
asymmetric traffic, DL and UL channel bands
January 2006 • Volume 4, Number 1
Ultimately,
radio resource
management
should follow a
combined form
of multiuser
diversity and
fairness
scheduling.
should be matched to the ratio of DL and UL
traffic. Moreover, FDD channel bands cannot be
adjusted dynamically in response to the varying
ratio of DL and UL traffic, due to hardware
limitations. The ratio of UL to DL streams is fixed
for FDD.
It has been assumed that network operators
determine the switching points for TDD UL and
DL schemes and that once such systems are
deployed, the DL/UL ratio is not changeable.
Moreover, switching points in adjacent cells
must be synchronized to avoid severe inter-
cell interference.
Figure 5 shows various co-channel interference
cases. In conventional TDD systems, only DL/DL
and UL/UL cases can occur. However, if the
DL/UL ratio is changed dynamically frame by
frame and independently cell by cell, co-channel
interference can exist in all four cases.
A cross-layer D-TDD scheme considering traffic
and channel condition together may be adopted.
Since each cell can have different offered loads for
UL and DL, cell switching points are set
independently. Although this may cause severe
co-channel interference at time slots around the
53
 Although D-TDD
is complex,
it has the
potential of
offering higher
bits-per-hertz
efficiency when
needed.

Figure 4. Usage of a Band AMC Zone Based on Multiuser Diversity

DL/DL UL/UL

DL/UL UL/DL

Figure 5. Various Co-channel Interference Cases

54 Bechtel Telecommunications Technical Journal

switching points, it still produces more
throughput than the conventional TDD scheme
with a fixed DL/UL ratio.
Switching points can be updated daily or by
frame. In the daily approach, the network
operator monitors for a certain amount of
time, then updates the switching point to allocate
resources based on switching point information.
In the frame approach (used in this paper), the
switching point is changed dynamically for each
frame based on traffic characteristics and channel
status. Based on the resource allocation
algorithm, each user is given a number of time
slots, with corresponding indices and a
modulation format. Modulation is determined by
the estimated channel state, and the time slot
indices are determined by the TSA algorithm.
Figure 6 illustrates the deployment of different
DL/UL ratios for different cells. In this example,
for the early symbols of the UL subframe in
Cell 2, the BS experiences severe co-channel
interference coming from Cells 1 and 3 because
they are all in the DL period and signals from
their BSs interfere with the UL signals of users in
Cell 2. To suppress interference in this case, a
good TSA algorithm with beam forming is
necessary. Several investigations of TSA
algorithms show that D-TDD systems out-
perform conventional TDD systems when the
dynamic traffic has unbalanced DL/UL
characteristics [9, 11]. Although the complexity
of D-TDD makes its adoption in the early
stages of Mobile WiMAX unlikely, it has the
potential of offering higher bits-per-hertz
efficiency when needed.
CONCLUSIONS
his paper has provided an overview of the
T IEEE 802.16e standard and Mobile WiMAX
and has provided some simple suggestions to
address certain PHY/MAC cross-layer issues.
Mobile WiMAX is expected to bring fast, broad,
seamless data communications, not only for fixed
home and small business subscribers, but also for
mobile subscribers. Mobile WiMAX will begin
competing in fixed broadband markets to link
homes and businesses with worldwide core
networks, before ultimately penetrating mobile
communication market shares. To strengthen the
market power of Mobile WiMAX, radio resource
management that deals with PHY/MAC cross-
layer issues needs to be developed accurately and
close to optimally.

January 2006 • Volume 4, Number 1
Mobile WiMAX
Figure 6. Various DL/UL Ratios for Different Cells
is expected to
bring fast, broad,
TRADEMARKS seamless data
Wi-Fi is a registered trademark of the Wireless communications,
Ethernet Compatibility Alliance, Inc. not only for fixed
home and small
REFERENCES business
[1] Wireless MAN Working Group subscribers, but
(http://www.wirelessman.org/). also for mobile
[2] IEEE Std 802.16-2004, “IEEE Standard for Local
and Metropolitan Area Networks – Part 16:
subscribers.
Air Interface for Fixed Broadband Wireless
Access Systems,” October 2004.
[3] IEEE P802.16-Cor1/D5, “Draft Corrigendum
to IEEE Standard for Local and Metropolitan
Area Networks – Part 16: Air Interface for
Fixed Broadband Wireless Access Systems,”
September 2005.
[4] IEEE P802.16e/D11, “Draft Amendment to
IEEE Standard for Local and Metropolitan Area
Networks – Part 16: Air Interface for Fixed and
Mobile Broadband Wireless Access Systems —
Amendment for Physical and Medium Access
Control Layers for Combined Fixed and Mobile
Operation in Licensed Bands,” September 2005.
[5] P. Viswanath, D.N.C. Tse, and R. Laroia,
“Opportunistic Beamforming Using Dumb
Antennas,” IEEE Transactions on Information
Theory, Vol. 48, No. 6, pp. 1277–1294, June 2002.
[6] H. Fattah and C. Leung, “An Overview of
Scheduling Algorithms in Wireless Multimedia
Networks,” IEEE Wireless Communications,
pp. 76-83, October 2002.
[7] TDD Coalition white paper, “The Advantages
and Benefits of TDD Broadband Wireless Access
Systems,” September 2001.
[8] D.G. Jeong and W.S. Jeon, “Time Slot Allocation
in CDMA/TDD Systems for Mobile Multimedia
Services,” IEEE Communications Letters, Vol. 4,
No. 2, February 2000.
[9] J. Li, S. Farahvash, M. Kavehrad, and
R. Valenzuela, “Dynamic TDD and Fixed
Cellular Networks,” IEEE Communications
Letters, Vol. 4, pp. 218–220, July 2000.
[10] W.C. Jeong and M. Kavehrad, “Co-channel
Interference Reduction in Dynamic-TDD Fixed
Wireless Applications, Using Time Slot
Allocation Algorithms,” IEEE Transactions on
Communications, Vol. 50, No. 10, pp. 1627–1636,
October 2002.
55
 [11] J. Yun and M. Kavehrad, “Adaptive Resource
Allocations for D-TDD Systems in Wireless
Cellular Networks,” Proceedings of MILCOM,
Vol. 2, pp. 1047–1053, November 2004.
[12] M. Ergen, S. Coleri, and P. Varaiya, “QoS Aware
Adaptive Resource Allocation Techniques for
Fair Scheduling in OFDMA Based Broadband
Wireless Access Systems,” IEEE Transactions on
Broadcasting, Vol. 49, No. 4, pp. 362–370,
December 2003.
Dr. Kavehrad has been a consultant to a score of major
corporations and government agencies. He has
published over 300 refereed journal and conference
papers, several book chapters, and books. His research
interests are in the areas of technologies, systems,
and network architectures that enable the vision
of the information age, e.g., broadband wireless
communications systems and networks and optical
fiber communications systems and networks. He holds
several key issued patents in these areas.

Dr. Kavehrad received three Exceptional Technical

BIOGRAPHIES
Jungnam Yun is a member of
the Technical Staff at the
POSDATA America R&D
Center, Santa Clara, California,
where he works as a system
engineer for the OFDMA PHY
algorithm and MAC scheduler.
His research interests are signal
processing, communication
theories, and cross-layer
optimization for radio resource allocations in
broadband wireless communication systems.
Dr. Yun received his BS and MS in Electrical
Engineering from Korea Advanced Institute of Science
and Technology (KAIST), Taejun, South Korea, and his
PhD in Electrical Engineering from The Pennsylvania
State University, State College, Pennsylvania.
Contributions Awards for his work on wireless
communications systems while he was with Bell
Laboratories; the 1990 TRIO Feedback Award for his
patent on a “Passive Optical Interconnect”; the 2001
IEEE Vehicular Technology Society Neal Shepherd Best
Propagation Paper Award; three IEEE Lasers and
Electro-Optics Society Best Paper Awards; and a
Canada National Science and Engineering Research
Council (NSERC) PhD dissertation Gold Medal award,
jointly with his former graduate students, for work on
wireless and optical systems. He is a fellow of the IEEE
and has lectured worldwide as an IEEE distinguished
lecturer and as a plenary and keynote speaker at
leading conferences.
Dr. Kavehrad received his PhD in Electrical Engineering
from Polytechnic University (Brooklyn Polytechnic),
Brooklyn, New York.

Mohsen Kavehrad was with

the Space Communications
Division, Fairchild Industries,
and Satellite Corporation and
Laboratories, GTE, from 1978 to
1981. In December 1981, he
joined Bell Laboratories, and
in March 1989, he joined
the Department of Electrical
Engineering, University of
Ottawa, Ottawa, Ontario, Canada, as a full professor. At
the same time, he also was director of the Broadband
Communications Research Laboratory, director of
Photonic Networks and Systems Thrust, project leader
in Communications and Information Technology
Ontario (CITO), and director of the Ottawa-Carleton
Communications Center for Research (OCCCR). He
was an academic visitor (senior consultant) at NTT
Laboratories, Yokosuka, Japan, in the summer of 1991.
He spent a 6-month sabbatical term as an academic
visitor (senior consultant) with Nortel, Ottawa,
Ontario, Canada, in 1996. In January 1997, he joined
the Department of Electrical Engineering, The
Pennsylvania State University, University Park, as
the AMERITECH (W.L. Weiss) professor of Electrical
Engineering and director of Communications
Engineering. Later, in August 1997, he was appointed
founding director of the Center for Information and
Communications Technology Research (CICTR). From
1997 to 1998, he also was chief technology officer and
vice president with Tele-Beam Inc., State College,
Pennsylvania. He visited, as an academic visitor (senior
consultant), Lucent Technologies (Bell Laboratories),
Holmdel, New Jersey, in the summer of 1999. He spent
a 6-month sabbatical term as an academic visitor (senior
technical consultant) at the AT&T Shannon Research
Laboratories, Florham Park, New Jersey, in 2004.

56 Bechtel Telecommunications Technical Journal

ANSI/TIA STANDARD 222 – STRUCTURAL
STANDARD FOR ANTENNA SUPPORTING
STRUCTURES AND ANTENNAS:
A COMPARISON OF REVISIONS F AND G
Issue Date: January 2006

Abstract—A new revision of ANSI/TIA Standard 222 will take effect on January 1, 2006. The new standard—
Revision G—is the most comprehensive revision of Standard 222 since its first publication in 1949. A revision
on this order can cause anxiety in the industry as to its impact on tower design and can raise questions.
This paper explains the differences in the basic design philosophies of the standard (Revision F) and the new
Revision G. It also discusses the impetus behind this major revision triggered by the latest understanding and
state-of-the-art practices of the current codes and standards in the building industry.

INTRODUCTION obtained from available sources and

his paper describes the provisions of the
T ANSI/TIA/EIA-222-F-1996 standard [1] for
the design, purchase, fabrication, and installation
of telecommunication tower superstructures and
foundations and compares these provisions with
those of the new ANSI/TIA-222-G-2005 standard
[2]. The new Revision G standard was issued in
August 2005 with an effective date of January 1,
2006. This standard is the governing document
for telecommunication towers in the United
States. The contents of the document are being
applied extensively as new telecommunication
tower sites are built and existing sites are
upgraded to accommodate the growth in the
wireless communication industry. This paper
provides insight into new Revision G of the
standard and how it will affect projects involving
wireless telecommunication tower sites.
The Telecommunications Industry Association
(TIA) subcommittee TR-14.7, which was
responsible for preparing the standard, states:
The objective of this Standard is to provide
recognized literature for antenna supporting
structures and antennas pertaining to: (a)
minimum load requirements as derived from
ASCE 7-02, “Minimum Design Loads for
Buildings and Other Structures,” [3] and (b)
Peter Moskal
design criteria as derived from AISC-LRFD-99,
pmoskal@bechtel.com
“Load and Resistance Factor Design
Specification for Structural Steel Buildings,”
Krishnamurthy [4] and ACI 318-05, “Building Code
Raghu Requirements for Structural Concrete” [5]. The
kraghu@bechtel.com information contained in this Standard was
represents, in the judgement of the
subcommittee, the accepted industry
minimum structural standards for the design
of antenna supporting structures and
antennas. While it is believed to be accurate,
this information should not be relied upon for
a specific application without competent
professional examination and verification of
its accuracy, suitability and applicability by a
licensed professional engineer. This standard
utilizes loading criteria based on an annual
probability and is not intended to cover all
environmental conditions which could exist at
a particular location.
The standard provides the requirements for the
structural design of new and the modification of
existing structural antennas and antenna-
supporting structures, i.e., towers, mounts,
structural components, guy assemblies,
insulators, and foundations.
This paper compares the provisions of Revisions
F and G of ANSI/TIA Standard 222 as they relate
to five key aspects of tower superstructure and
foundation design:
• Design method evolution from a traditional
working strength design approach to a
contemporary load resistance factor design
approach for the tower superstructure
• Wind load definition
• Seismic requirements
• Ice loading definition
• Design of existing structures

© 2006 Bechtel Corporation. All rights reserved. 57

 Revision F
ABBREVIATIONS, ACRONYMS, AND TERMS This standard uses the working stress design
ACI American Concrete Institute method. The American Institute of Steel
AISC American Institute of Steel Construction (AISC) refers to this approach as the
Construction allowable stress design (ASD) method. Stresses in
AISI American Iron and Steel a structure are determined by application of
Institute defined service loads using the principles of
ANSI American National Standards statics and dynamics. The loads are applied to a
Institute suitable arrangement of structural elements
ASCE American Society of forming a stable assembly. The assembly has
Civil Engineers sufficient strength and is designed using
ASD allowable stress design members and connections of defined geometric
A significant change shape and known material properties in
COA Certificate of Authorization
in the standard accordance with applicable code provisions. The
ERC engineer in responsible
resulting actual member stresses are compared
reflects the charge
to allowable member stresses, based on the
migration from the IBC International Building Code
code, which are less than the member material
typical working NWS National Weather Service
yield stresses. The allowable stresses provide
TIA Telecommunications Industry an inherent factor of safety to account for
stress design to Association uncertainties related to typical simplifying
the contemporary
assumptions and the use of nominal or average
limit-states calculated stresses as the basis for manual
design approach In each case, the two standards are contrasted by methods of analysis. The margin between the
briefly describing the provisions of Revision F allowable stresses and the material yield stresses
for structural steel.
and then those of Revision G, followed by a provides the margin of safety. In applying this
summary of the significance of the changes. approach, engineering judgment must be
exercised. The working stress approach has
This paper also discusses related areas of interest.
served as the principal design philosophy for
Figure 1 shows a typical guyed lattice
a century.
telecommunication tower structure governed by
the ANSI/TIA-222 standard. Allowable stress design may be formulated as
follows:

DESIGN METHOD
φRn = ∑Q
significant change in the standard reflects (1)
A the migration from the typical working
stress design to the contemporary limit-states
γ i

design approach for structural steel. This brings
the standard into compliance with current codes.
In this elastic design approach, all loads Qi are
assumed to have the same variability, and the
term φ/γ may be thought of as the safety factor
applied to the material resistance. For materials in
direct tension, the allowable stress is limited to
0.60 times the material yield strength. This results
in a safety factor of 1/0.60 = 1.67. The AISC
specification defines the allowable stresses
for tension members, columns and other
compression members, beams and other flexural
members, plate girders, bolts, and welds.

Per Revision F, design for structural members,

Figure 1. Guyed Lattice Tower Structure Governed by
ANSI/TIA-222
unless otherwise noted, must be in accordance
with the appropriate AISC or American Iron and
Steel Institute (AISI) specification. American
Society of Civil Engineers (ASCE) Standard 10 is
used to adjust the AISC allowable compression
stresses for the effects of eccentric axial loading
and partial end restraint for structural steel
single-angle compression members.

58 Bechtel Telecommunications Technical Journal

Revision G
The new standard is based on limit-states design.
Structural design has been moving toward this
more rational probability-based approach over the
past 25 years. The generic term “limit states” used
in the new standard is synonymous with
the AISC’s use of load and resistance factor
design terminology.
Structures (tower superstructure and foundation),
members, and connections must have adequate
strength to function safely over their service life.
Reserve strength must exist in a structure to
account for possible overload and under-strength
conditions. Overloads may occur because the
structure’s use may be changed or the design loads
may be judged to be less than those actually
experienced. Under strength may result from
variations in member sizes due to manufacturing
or construction practices, or from installed
material strengths that are less than specified.
A structure must contain an adequate safety
margin. Safety has been studied using
probabilistic methods to assess the chance of
failure or the limit state. Limit states are reached
when a structure no longer performs its intended
function. Limit states are divided into two
categories: strength and serviceability. Strength
refers to the ability of elements to sustain intended
loads and maintain stability in conditions such as
sliding or overturning. Serviceability involves
consideration of requirements such as deflection
and permanent deformation.
The loads acting on a structure and the
structure’s resistance to the loads must balance
with an appropriate allowance for a margin of
safety. The variables that affect loads and
strength have been studied probabilistically to
assess variability. The limit-state approach
assumes that the load Q and the resistance R are
random variables. The goal is to have the
resistance R exceed the load Q by a reasonable
margin of safety. Unless the margin of safety is
very large, some probability exists that R may be
less than Q. When assessing the relationship for
the natural logarithm of resistance divided by
load, i.e., ln(R/Q), a probability distribution
function is obtained. Values for the number of
standard deviations that are appropriate between
the mean value of ln(R/Q) and the lower
specification limit are established and designated
as the reliability index [6] (see Figure 2).
This leads to the following limit-state requirement:
φRn = ∑ γi Qi (2)
January 2006 • Volume 4, Number 1
The left side of this equation represents resistance
or strength, i.e., strength of concrete, steel, or other
material. The right side represents the loads that
the structure is designed to carry. On the left side of
the equation, the nominal resistances Rn are
multiplied by a strength reduction factor φ to
obtain a design strength limit. On the right side, the
applied loads Qi (e.g., wind, ice, and seismic) are
multiplied by a load factor γi to obtain the factored
design loads for which the structure is designed.
The strength reduction factors are less than one to
establish material strength limits with an
appropriate conservatism applied based on The limit-state
statistical assessments of materials. The load factors design approach is
are greater than one to allow margin in the design
the contemporary
loads that are also based on statistical assessments.
method for
Additional discussion and comparison of the
structural analysis
working stress design and limit-state design
approaches can be found in Chapter 1 of [7]. and design.
The new standard specifies the requisite loads,
load combinations, strength reduction factors,
and load factors for tower design and
modification use.
Design Method Summary
The limit-state design approach is the
contemporary method for structural analysis and
design. It is now the method of design for
telecommunication towers as well as concrete
and structural steel. This is a rational approach
that will facilitate incorporation of additional
information that becomes available on loads and
resistances and their variation. It provides a
framework for handling unusual loads that may
not be covered by specifications. Safer structures
may result because the method should lead to
improved awareness of structural behavior. In
most cases, it should also improve the economy
of structures.
Figure 2. Reliability Index β
59
 As shown in Figure 3, the value of 0.02 indicates
a 2 percent chance that the 70 mph speed will be
exceeded in a year or a 98 percent chance that it
will not be exceeded. Wind determination
requires a probabilistic formulation, and the
necessary probability measures must be
estimated on the basis of experimental
formulations. The average time between two
consecutive annual occurrences of the wind
exceeding 70 mph is the reciprocal of the
probability of the event within one time unit, i.e.,
1/0.02 = 50 years. Therefore, the probability that
the wind will exceed 70 mph is a 2 percent chance
in a given year, and the chance that this speed
will be exceeded in consecutive years occurs once
Figure 3. Largest Annual Fastest-Mile Velocity (V) every 50 years. The National Oceanic and
Atmospheric Administration’s National Weather
Service (NWS) branch collects and publishes
WIND LOADS wind speed data throughout the US.

nother significant change relates to the

A migration from fastest-mile wind speed to a
3-second-gust wind speed definition, which
Revision G
In the new standard, wind speed definition is
brings the standard into compliance with other based on the 3-second gust. Wind speeds are
current standards, codes, and guides. developed for 3-second gusts at 33 feet above
ground in open-country exposure Category C for
a 50-year mean recurrence interval. The NWS has
Revision F
phased out the measurement of fastest-mile wind
Section 2.6.1 of the standard defines the basic
speeds, and the basic wind speed has been
wind speed as the fastest-mile wind speed at
redefined as the peak gust, which is recorded and
33 feet above the ground corresponding to an
archived for most NWS stations. The wind speeds
annual probability of 0.02 (50-year recurrence
are not representative of speeds at which ultimate
interval) for Exposure C. This may be thought of
limit states are expected to occur. Load factors or
as the average speed obtained during the passage
allowable stresses used in design equations lead
of 1 mile of wind. For a 70 mph wind, this would
to structural resistances substantially higher than
mean that 1 mile of wind passes in 51.4 seconds.
the design wind speeds. Figure 4 illustrates a
The wind speed is measured by a weather device
wind speed time history record and provides an
called an anemometer at a height of 33 feet above
interpretation of the two speed definitions.
the ground. The annual probability is based on
Figures 5 and 6 show lattice towers with
the probability distribution function.
significant coaxial cable wind load areas.

Figure 4. Wind Speed Record

60 Bechtel Telecommunications Technical Journal

Wind Loads Summary
The fastest-mile wind speed of 70 mph mentioned
previously is equivalent to an 85 mph 3-second-
gust wind speed. The new standard provides a
table of wind speed equivalencies.
This approach to wind loads makes the standard
consistent with ASCE Standard 7 and codes such
as the International Building Code (IBC). It is the
contemporary wind load approach in the US.
This standard enables accurate site-specific wind
modeling using exposure categories and
topographic effects.
The new standard will use an averaging period
of 3 seconds for the US. Other countries use
different averaging periods (e.g., 1 hour or
10 minutes). Weather data in the US is collected at
465 stations throughout the country.
A description of how to convert the fastest-mile
wind speed to the 3-second-gust definition can be
found in Section 4.0 of Reference [8].
SEISMIC LOADS
nother significant change in the standard is
A the inclusion of seismic provisions. This
brings the standard into compliance with other
current standards, codes, and guides.
Revision F
The standard does not contain seismic provisions.
When seismic loads were considered to apply to a
specific tower design, an equivalent static load,
response spectrum modal analysis, or time
Figure 5. Coaxial cable is a major contributor of wind load
on a tower. Wind computations are handled
in detail in Revision G.
January 2006 • Volume 4, Number 1
history analysis would be applied based on the
provisions contained in applicable building
codes.
Revision G
The new standard defines four methods of
structural analysis for seismic loads. A table is
provided to identify which method is to be used
for which tower type in accordance with mass and
stiffness irregularities. The four approaches are:
• Method 1 – Equivalent Lateral Force
The total weight of the structure, including The new
appurtenances, is determined; the total standard defines
seismic shear is determined and distributed four methods
as specified in the standard. The structure is
then analyzed statically using the seismic
of structural
forces as external loads. analysis for
• Method 2 – Equivalent Modal Analysis seismic loads.
The fundamental frequency of the structure
is determined along with the seismic forces
for each level of the structure. The structure is
then analyzed statically using the seismic
forces as external loads.
• Method 3 – Modal Analysis
A mathematical model of the structure is
created that represents the spatial
distribution of the mass and stiffness. The
mode shapes—the period, shape factor, and
participation factor—are determined. The
design response spectrum is established in
accordance with the standard. The base
shears and forces at each level of the
Figure 6. Towers can be loaded with a very complex array
of coaxial cable. Revision G addresses
several load scenarios.
61
 structure for each mode are determined and
then combined by the square root of the sum
of the squares of the modal values.
• Method 4 – Time History Analysis
A mathematical model of the structure is
created that represents the spatial distribution
of the mass and stiffness considering
structural damping to be 5 percent of critical
damping. Two orthogonal ground motion
time histories are selected from recorded
events with seismicity compatible to the site,
The standard or synthetic time histories are developed
in accordance with accepted practice. A
includes provisions
5 percent horizontal response spectrum is
for the design of constructed, and the foregoing time histories
existing tower are combined according to rules specified in
superstructures the standard to develop two scaled, site-
specific time histories. These two time
and foundations.
histories are used to perform a structural
analysis. Load effects for design are
determined by selecting the maximum values
from the time history analysis.
Seismic Loads Summary
Under the new standard, wireless
telecommunications structures will typically be
analyzed using either Method 1 or 2.
The standard contains provisions for ground
motions, torsional moments, mathematical model
requirements, and earthquake loads for
structures supported on buildings or other
supporting structures. A seismic approach
considering the uniqueness of towers will now be
included and considered as part of the design.
ICE LOADS
he inclusion of ice loading provisions
T represents a significant change that brings the
standard into compliance with the current
standards, codes, and guides.
Revision F
Though it recognizes a consideration of radial solid
ice uniform thickness with a density of 56 lb/ft3,
the standard does not specifically state an ice
requirement. The standard does recognize that
ice may be a significant load for structures to be
located in areas of significant ice accumulation and
provides information for consideration in an annex.
Revision G
The design ice thickness specified in the new
standard is a uniform radial thickness of
glaze ice at 33 feet above the ground in exposure
62
Category C for a 50-year mean recurrence
interval. Escalation of ice thickness and wind on
ice over the height of the structure is required.
Ice is assumed to be glaze ice with a density of
56 lb/ft3. Accumulation of ice is considered on the
structure, guys, and appurtenances. All elements
are assumed to be covered with a uniform
thickness of ice that results in a wind drag.
Design ice thickness is also escalated with height
and is based on regional climatic data. For
engineering design, all members are traditionally
assumed to be covered with a uniform thickness
of ice, which together with the ice density may be
used to calculate the ice weight as well as the
wind drag.
Ice Loads Summary
The new standard provides a rational approach
for considering ice thickness in the tower and
foundation design.
EXISTING STRUCTURES
he standard includes provisions for the
T design of existing tower superstructures and
foundations.
Revision F
The standard indicates that towers and
supporting structures should be analyzed when
changes in the original design or operating
conditions take place. Recommended criteria
appear in an annex to the standard.
Revision G
A section of the new standard addresses the
structural analysis of existing structures. Another
section discusses the evaluation of structures
regardless of the standard used in the original
design. Existing structures are to be evaluated in
the following circumstances:
• A change in type, size, or number
of appurtenances such as antennas,
transmission lines, platforms, and ladders
• Any structural modification (except for
maintenance) made to the structure
• A change in serviceability requirements
• A change in structure classification in
accordance with the categories identified
in the standard
Existing Structures Summary
The new standard indicates that existing
structures need not be re-evaluated for each
Bechtel Telecommunications Technical Journal
revision of the standard except in the
circumstances listed above. Analyses and
modifications will definitely include modeling,
analysis, design, and modification using the
provisions of the standard in computer models
and calculations.
SOFTWARE
he commercially available software products
T typically used for tower analysis, design,
and modification are Power Line Systems,
Inc.’s TOWER® and PLS-POLE®, and RISA
Technologies’ RISATower. Both companies’
software packages are now standard computer
applications in accordance with engineering
department procedures. Both companies are
including the provisions for the new standard in
their software; this capability will be available in
January 2006.
TRAINING
raining in the following subjects is available
T for engineers who work in this area.
Customers may also wish to take advantage of
this training.
• Telecommunication Tower Overview
The overview training is an orientation for all
engineering personnel, regardless of
discipline, involved in wireless site project
tower activities. Practical background is
provided to enhance understanding of the
different types of towers and foundations,
the codes and standards that apply, and the
work associated with tower activities.
• ANSI/TIA-222-G-2005 Structural Standard
for Antenna Supporting Structures and
Antennas
This offering focuses on the meaning and use
of the Revision G standard.
• Telecommunication Tower Structural
Analysis and Design
This course provides instruction for
civil/structural engineers in the use of
Revisions F and G of the ANSI/TIA-222
standard. Training is provided in the use of
the TOWER, PLS-POLE, and RISATower
software packages. Participants model and
design a lattice steel guyed tower and a self-
supporting, tapered tubular steel pole tower.
The course provides the background
necessary for an engineer to independently
analyze and design telecommunication
towers. Upon successful completion of the
January 2006 • Volume 4, Number 1
course and after suitable experience,
participants will possess the skill set necessary
to assess, evaluate, analyze, and modify
towers using the ANSI/TIA-222 standard.
LICENSED PROFESSIONAL ENGINEER SEAL
he primary responsibility of a registered
T licensed Professional Engineer is the
protection of public, life, safety, health, welfare,
and property. All tower work is to be carried out
under the direction of a licensed Professional
Engineer in the Civil and/or Structures
The primary
disciplines in responsible charge as defined in
state licensing laws, rules, and regulations. In a responsibility of a
majority of states, all companies performing registered licensed
engineering must possess a Certificate of Professional
Authorization (COA) with an engineer in
Engineer is the
responsible charge (ERC) designated on the
certificate. The COA is the state corporate license protection of public,
to practice engineering. In states that do not life, safety, health,
require a COA, all companies performing welfare, and
engineering must have an ERC with
property.
the appropriate Professional Engineering
registration.
The drawings, calculations, and reports prepared
by civil/structural engineers for owners are
manually wet-stamp sealed or manually
embossed with a seal and manually signed and
dated by a licensed Professional Engineer
registered in the state where the tower structure
is located. States are also allowing electronic
seals, signatures, and dates on engineering
deliverables.
CONCLUSIONS
he comprehensive Revision G of Standard 222
T governing telecommunication tower
activities in the US will continue to guide the
creation of tower structures that successfully
perform their intended function. The standard
addresses both the tower superstructure and
the foundation, as both are necessary for the
successful performance of a tower in service.
The term “tower“ encompasses both the
superstructure above the ground and the
foundation below the ground functioning
together to support loads under all design
conditions.
The standard has been prepared by individuals
and companies who work and practice in this
area, i.e., fabricators, erectors, consultants,
contractors, and architect-engineers. Designs are
expected to be in line with those that result from
application of Revision F of the standard.
63
 In summary, ANSI/TIA Standard 222 Revision G
reflects contemporary engineering practice and
incorporates the latest appropriate technical
methods and academic information. It is a
welcomed evolution that brings the science and
art of tower design and analysis into conformance
with current provisions of the AISC, ACI, ASCE,
IBC, and other national codes, standards, and
practices.

TRADEMARKS
and selects and provides the software, design guides,
specifications, training, and tools needed to perform
the work. He also regularly supports business
development activities.
Peter is a voting member of ANSI/TIA-222-G-2005
standard committee TR-14.7. He holds Professional
Engineer licenses in multiple states and serves as the
designated engineer in responsible charge on a number
of corporate licenses to practice engineering.
Peter received a BS in Civil Engineering and an MS
in Civil Engineering (Structures and Soil Mechanics)
from the University of Pittsburgh and a Business
Management Certificate from Golden Gate University.

PLS-POLE and TOWER are registered

trademarks of Power Line Systems, Inc.
REFERENCES
[1] American National Standards Institute
(ANSI)/Telecommunications Industry
Association (TIA)/EIA-222-F, Structural Standards
for Steel Antenna Towers and Antenna Supporting
Structures, June 1996.
[2] ANSI/TIA-222-G, Structural Standard for Antenna
Supporting Structures and Antennas, August 2005.
[3] Structural Engineering Institute (SEI)/American
Society of Civil Engineers (ASCE) 7-02, Minimum
Design Loads for Buildings and Other Structures,
2003.
[4] American Institute of Steel Construction (AISC),
Manual of Steel Construction, Load and Resistance
Factor Design.
[5] American Concrete Institute (ACI) 318, Building
Code Requirements for Structural Concrete.
[6] A.H.-S. Ang and W.H. Tang, Probability Concepts
in Engineering Planning and Design, Volume 1 –
Basic Principles, John Wiley & Sons, Inc., 1975.
[7] C.G. Salmon and J.E. Johnson, Steel Structures
Design and Behavior, Fourth Edition, Harper
Collins, 1996.
[8] M.K.S. Madugula, “Dynamic Response of
Lattice Towers and Guyed Masts,” ASCE,
Reston, Virginia, 2002.
Krishnamurthy Raghu is an
engineering supervisor in
Bechtel Telecommunications
currently responsible for
supporting Telecommunications
staff engineering and self-
performance engineering in
Frederick, Maryland. A
civil/structural engineer, he
has also worked in Bechtel’s
Power and Petroleum & Chemical business lines. He is
experienced in wireless tower engineering, site
development, and zoning as well as in the transmission
and distribution industry. He is a technical specialist in
tower and pole structures and telecommunications
standards.
Krishnamurthy is a member of ANSI/TIA-222-G-2005
standard committee TR-14.7.
Krishnamurthy received a BE in Civil Engineering from
Bangalore University, India, and an MS in Civil
Engineering from Virginia Polytechnic Institute and
State University (Virginia Tech).

BIOGRAPHY
Peter Moskal is chief engineer,
civil, structures, and architecture,
for Bechtel Telecommunications.
He is responsible for the wire-
less and wire line infrastructure
engineering deliverables, the
people who prepare those
deliverables, the processes
employed, and the tools used to
execute the work on behalf of
Bechtel Telecommunications customers. He oversees
the work of civil engineers, structural engineers,
architects, and CAD designers and drafters within the
Telecommunications organization.

Peter implements defined standard work processes in

accordance with engineering department procedures

64 Bechtel Telecommunications Technical Journal

REDUCING THE AMOUNT OF FIBER IN
FIBER-TO-THE-HOME NETWORKS
Issue Date: January 2006
Abstract—Reducing the amount of unnecessary fiber deployed in fiber-to-the-home networks can significantly
reduce capital costs. Two principal causes of fiber waste—excessive serving area size and poor hub location—
are examined. Methods to detect and quantify excessive fiber usage attributable to these causes are developed.
INTRODUCTION
n excessive amount of fiber is being
A deployed in today’s fiber-to-the-home
(FTTH) networks. In many cases, more than half
the fiber deployed may be unnecessary. Although
the excess fiber is increasing the cost of networks,
this phenomenon is occurring largely unnoticed
for several reasons:
• Fiber is generally regarded as “cheap.”
Little attention is paid to conserving fiber,
since its cost is generally considered
insignificant. While it is inexpensive, fiber is
not free.1 The cost of fiber in a network can
reach hundreds of dollars per household
passed. Significant savings can be realized if
excess fiber is eliminated.
• There is a common belief that deploying
extra fiber is “good.” It is considered
less expensive to have extra fiber in place
to meet future demands than to install
additional fiber later on. Indeed, there
is some economy in deploying extra fiber
for maintenance and foreseeable growth
opportunities. A threshold exists, though,
between extra fiber and excessive fiber
deployment. Installing fiber in excess of
the foreseeable demand is an inefficient use
of capital.
Brian Perkins ____________________________
bmperkin@bechtel.com 1 While the price of fiber-optic cable can vary based on
the number of fibers, cable construction, and purchase
commitments, $0.01 per fiber-foot is a useful rule-of-
thumb approximation in today’s market.
© 2006 Bechtel Corporation. All rights reserved.
• It is difficult to determine when a network
design contains excessive fiber. Each design
is a unique reflection of an engineer’s
judgment applied to the peculiarities
of differing neighborhoods. No two
neighborhoods will yield the same design,
nor will two engineers produce the same
design for the same neighborhood.
Identifying excessive fiber is a challenging
task and the impetus for this paper. Two
metrics are developed in this paper that can
be simply applied to any network design to
indicate the presence of excessive fiber.
BACKGROUND
iber is the principal component of any FTTH
F network, as the name implies. Fiber from
homes is routed to an aggregation point in the
neighborhood. At this aggregation point or hub,
traffic from multiple fibers is combined onto
fewer fibers for transport back to the central
office. The hub can combine traffic passively via
optical splitters in a passive optical network
(PON) or actively via an Ethernet switch in an
active Ethernet network. Within the central office,
traffic is routed to the appropriate destination.
Traffic flows in both directions: home to central
office (upstream) and central office to home
(downstream). The hub not only aggregates
upstream traffic but also distributes downstream
65
 homes suited to the capacities of the available
ABBREVIATIONS, ACRONYMS, AND TERMS hubs. Once the serving areas are defined, a hub
FTTH fiber to the home site is selected within each serving area. Hubs are
HC hub centricity best located near the center of the serving areas,
PON passive optical network but safety, accessibility, and aesthetic concerns
often cause hubs to be located near the serving
PSTN public switched telephone
network area edges (see Figure 2).

The completed plan, with serving area

traffic from the central office across the multiple
fibers to the homes (see Figure 1).
The design of FTTH networks typically occurs in
Two principal two phases: planning and engineering. During
causes of excessive the planning phase, homes are grouped into
serving areas. The serving area boundaries are
fiber deployment:
established to contain an “ideal” number of
•Overly large
serving area
boundaries and hub locations established, is used
as a template for the engineering phase. The
network design is completed by routing and
sizing fiber cables that connect every household
to the hub location within each serving area. Only
minor modifications of the plan, to compensate
for constructability issues, are considered during
the engineering phase.

•Poor hub
location

Figure 1. FTTH Network Architecture

CAUSES OF EXCESSIVE FIBER DEPLOYMENT

eviews of several FTTH network designs
R identified two principal causes of excessive
fiber deployment:

• Overly large serving area

• Poor hub location
The following sections discuss serving area size
and hub centricity (HC) and their effect on the
amount of fiber deployed. Models and metrics
are presented to detect and assess the impact of
these causes.

Figure 2. Hub Locations Within Serving Areas

66 Bechtel Telecommunications Technical Journal

Serving Area Size
The serving area size for each network design is
X
chosen during the planning phase. Armed with
ONT ONT ONT ONT
guidelines and an approved list of hub sizes, the
planner uses “good judgment” to group an
acceptable number of households into serving
ONT ONT ONT ONT
areas. Previous work has suggested that expected
subscription rate (or take rate) should also be
considered when establishing serving area size to
maximize resource utilization [1]. In general, Figure 3. Linear Serving Area with Eight Households
resource utilization improves as serving area size

X A larger serving area

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT
requires additional
fiber per household
12 Fiber Cable 12 Fiber Cable
to be installed.
ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

Figure 4. Linear Serving Area with 24 Households

increases, lowering the cost per subscriber. linear serving areas of any size n, the total fiber
However, there is a penalty to be paid for TFl required is:
increasing serving area size. A larger serving area
requires additional fiber per household to be
⎞ n2 ⎞
installed. This principle is easily demonstrated TFl = ⎟⎠ 4 – n⎟⎠ x (1)
and quantified with two simplistic examples.

Linear Serving Area

Consider a serving area composed of eight
households arranged as shown in Figure 3.
Each household lot has frontage of x feet. With
the hub placed in the center, four fibers would
exit east and west and traverse one lot,
connecting to drops to each of the four
households on either side of the hub. The total
fiber required is 8x fiber-feet or x fiber-feet
per household.
Consider now a serving area of 24 households, as
shown in Figure 4.
The dimensions are the same as before; each lot
has a frontage of x feet. The hub is centered with
12 households to either side. In this case, 12 fibers
exit east and west from the hub and traverse
5 lots to connect all the households.2 The
total fiber required for this serving area is
120x fiber-feet or 5x fiber-feet per household.
The previous two examples demonstrate that
tripling the serving area size results in a fivefold
increase in the fiber required per household. For
____________________________
2 The fiber count could be reduced by four at each drop
location. However, it is more expensive to splice smaller
cables at each taper point than to continue the original
12 fibers the entire length.
The total fiber per household TFl / n is simply:
TFl ⎞ n – 1⎞ x
n = ⎟⎠ 4 ⎟⎠
TFl ≈ n x n >>4 (2)
n 4
For large linear serving areas, total fiber per
household is directly proportional to the serving
area size—splitting the serving area in half halves
the fiber cost per household.
The linear serving area model may be too severe
a characterization of the real world. Therefore, it
may be useful to investigate the other extreme—
an overly generous characterization.
Ideal Mesh Serving Area
Consider a serving area composed of
16 households arranged in a square, as shown
in Figure 5.
Each household lot is x by x feet. The hub is
located at the center of the serving area with eight
households north and south of the hub. A main
distribution cable of eight fibers exits north and
south of the hub and traverses one lot.
Subtending distribution cables of four fibers

January 2006 • Volume 4, Number 1 67

 traverse one lot east and west of both the north
and south main cable to connect all the
households. The total fiber used is 32x fiber-feet
(16x main cable + 16x subtending cable) or
2x fiber-feet per household.
X
ONT ONT ONT
Now consider a similar serving area of
144 households, as shown in Figure 6.
The lot dimensions are the same: x by x feet. The
hub is located at the center. The main distribution
cable is now 72 fibers and traverses 5 lots.
There are 12 subtending cables of 12 fibers,
each traversing 5 lots. The total fiber is
ONT 1,440x fiber-feet or 10x fiber-feet per household.

X
In general, the total fiber TFim required for an
ideal mesh serving area is:
ONT ONT ONT ONT

TFim = n ⎞⎟⎠ n – 2⎞⎟⎠ x

ONT ONT ONT ONT
(3)

ONT ONT ONT ONT

Figure 5. Ideal Mesh Serving Area with 16 Households

X
ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT
X

12 Fiber Cable 12 Fiber Cable

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

12 Fiber Cable 12 Fiber Cable

72 Fiber Cable

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

12 Fiber Cable 12 Fiber Cable

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

ONT ONT ONT ONT ONT ONT

HUB ONT ONT ONT ONT ONT ONT

12 Fiber Cable 12 Fiber Cable

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

12 Fiber Cable 12 Fiber Cable

72 Fiber Cable

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

12 Fiber Cable 12 Fiber Cable

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

Figure 6. Ideal Mesh Serving Area with 144 Households

68 Bechtel Telecommunications Technical Journal

The total fiber per household TFim /n is:

TFim = ⎞ n – 2⎞ x
n ⎟⎠ ⎟⎠

TFim ≈ nx n >>4 (4)

n

For large ideal mesh serving areas, total fiber per

household is proportional to the square root of
the serving area size. Splitting a serving area 32 232 432 632 832 1032
in half reduces total fiber per household by
30 percent. Series 2 Rural Linear Series 3 Rural Mesh
Series 4 Urban Linear Series 5 Urban Mesh
Optimum Serving Area Size
Both models indicate a direct relationship Figure 7. Total Fiber per Household Versus Serving Area Size
between the size of the serving area and the total (Rural and Urban)
fiber per household. The expectation is for real
world behavior to be bounded by the predictions
of the two models. However, the models reveal
another important relationship. The total fiber per
Fiber Cost per Subscriber

household and the geographic density—the x

factor—are directly related. While the x factor is
very specific to geographic locations, some useful
generalizations can be made. In more rural
environments with households separated by
greater distances, the total fiber per household is
greater than in urban environments. Figure 7
plots the relative total fiber per household versus
32 232 432 632 832 1032
serving area size for rural and urban Serving Area Size
environments. Linear Mesh
To determine the optimum serving area size for a
Figure 8. Fiber Cost per Subscriber Versus Serving Area Size
specific region, the fiber cost per subscriber
F$ /sub of a specific design must be calculated:

F$ TF$ • 1
Equipment Cost per Subscriber

sub = n τ (5)

Where:
TF$ = total fiber cost
n = number of households in serving area
τ = expected take rate

Using the models, the fiber cost per subscriber 32 132 232 332 432 532
can now be extrapolated for any serving area size, Serving Area Size
as shown in Figure 8.
Figure 9. Equipment Cost per Subscriber Versus Serving Area Size
An equipment cost per subscriber plotted against
serving area size can be derived from equipment
cost and calculated utilization (see Figure 9).

January 2006 • Volume 4, Number 1 69

 The optimal serving area size is obtained by
finding the minimum total cost—fiber cost per
subscriber + equipment cost per subscriber (see
Figure 10).
Prior to calculating optimum serving area size,
excess fiber associated with HC must be
identified and removed.
Hub Centricity
FTTH network planning, as described previously,
is a two-step process: establish the service area
boundaries, then locate the hub. There is a
general perception that hub location has little
impact on cost. As such, little effort is expended
on selecting the least cost location. Locations at
the edge of the serving area—considered more
“convenient”—are often chosen. However,
serving areas with hubs located away from the
center require additional fiber strands to be
engineered to gain full connectivity. To
understand the impact, the earlier models are
revisited.
Common Equipment + Fiber Cost per Subscriber vs. Serving Area Size
Common Equipment + Fiber Cost per Subscriber
Optimal Serving Area Size
32 132 232 332 432
Serving Area Size
Equipment + Fiber Linear Model
Equipment + Fiber Mesh Model
Figure 10. Total Cost per Subscriber Versus Serving Area Size
ONT ONT ONT ONT
ONT ONT ONT ONT
Figure 11. Linear Serving Area with 24 Households with Hub Off-Center
70
Linear Serving Area
From Figure 4, with the hub centered in the linear
serving area, the total fiber required was
120x fiber-feet or 5x fiber-feet per household. If
the hub is moved to the far right, as shown in
Figure 11, additional fiber is required.
In this case, 20 fibers must now exit to the west of
the hub and traverse 10 lots. The total fiber
required is 200x fiber-feet or 8.3x fiber-feet per
household—a 66 percent increase.
Ideal Mesh Serving Area
From Figure 6, with the hub centered in the ideal
mesh serving area, the total fiber required was
1,440x fiber-feet or 10x fiber-feet per household. If
the hub moved to the extreme south end of the
serving area, as shown in Figure 12, additional
fiber is required.
In this case, 120 fibers exit the hub to the north as
the main distribution cable and traverse 10 lots.
The subtending distribution cables remain the
same. The total fiber required is 1,920x fiber-feet
or 13.33x fiber-feet per household—a 33 percent
increase.
Hub Centricity Metric
To quantify the degree to which the hub is
centered, an HC metric is created. HC is defined
as the ratio of the total number of fibers exiting
the hub less the greatest number of fibers exiting
the hub in one direction to the greatest number of
fibers exiting in one direction.
In the Figure 4 example, HC = (24 – 12)/12 = 1.0.
In the Figure 11 example, HC = (24 – 20)/20 = 0.2.
In the Figure 6 example, HC = (144 – 72)/72 = 1.0.
532 In the Figure 12 example, HC = (144 – 120)/120 =
0.17.
An HC metric of 1 (or higher) indicates that the
hub is well-centered within the serving area. An
HC metric below 1 indicates that the hub is not
ONT ONT ONT ONT ONT ONT ONT ONT
ONT ONT ONT ONT ONT ONT ONT ONT
Bechtel Telecommunications Technical Journal
 X
X ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

12 Fiber Cable 12 Fiber Cable

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

12 Fiber Cable 12 Fiber Cable

120 Fiber Cable

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

A very low HC metric

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT
indicates a
12 Fiber Cable 12 Fiber Cable significant amount
of excess fiber—over
ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT
50 percent in some
ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT
instances.
12 Fiber Cable 12 Fiber Cable
120 Fiber Cable

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

12 Fiber Cable 12 Fiber Cable

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT ONT

12 Fiber Cable 12 Fiber Cable

ONT ONT ONT ONT ONT ONT HUB ONT ONT ONT ONT ONT ONT

Figure 12. Ideal Mesh Serving Area with 144 Households with Hub Off-Center

centered; the lower the metric, the more the hub

is offset from center.
0.6
While the two models predict an increase in total 0.5
fiber per household, it is difficult to predict how
much excess fiber is present in a design with the 0.4
Waste

given HC metric. To understand the behavior of 0.3

the metric, 50 actual designs were reviewed. The
0.2
metric was calculated for each design. Where the
hub was not centered, a better, more central, hub 0.1
location was sought. The fiber required was 0.0
recalculated for each design with improved hub 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
location. The difference in fiber requirements Centricity
between the two designs was classified as excess
fiber and is plotted against the HC metric of the
Figure 13. Excess Fiber Versus Hub Centricity
original design (see Figure 13).

The plot suggests a good correlation between the excess fiber—over 50 percent in some instances.
HC metric and the amount of excess fiber. A very The decision to find a more suitable hub location
low HC metric indicates a significant amount of or alter serving area boundaries to center the hub

January 2006 • Volume 4, Number 1 71

 should be made after evaluating the potential REFERENCES
savings to be realized. Reworking a network [1] B. Perkins, “Optimal Splitter Placement in PONs,”
design in an urban area with an HC of 0.6 and a Bechtel Telecommunications Technical Journal, Vol. 2,
fiber cost of $30 per household may not be No. 2, September 2004, pp. 49–52.
warranted to realize a potential savings of $0 to $3
per household. In contrast, a design with an HC
of 0.1 and a fiber cost of $200 per household with BIOGRAPHY
Brian Perkins joined Bechtel
the potential to recover $100 per household most
Telecommunications in 2004
certainly warrants rework. with more than 15 years
of experience in the tele-
communications industry. As
senior director of technology—
CONCLUSIONS
optical networking, he
The two primary
he two primary causes of excessive fiber
causes of excessive
fiber deployment—
T deployment—creation of too large a serving
area and poor hub centering—originate in the
evaluates new products and
technologies and establishes
best practices for designing and
implementing fiber-optic networks in support of
creation of too large planning phase. Bechtel’s global projects.
a serving area Fiber costs of today’s FTTH networks can be Before joining Bechtel, Brian worked for Quantum
hundreds of dollars per household. Identifying Bridge Communications, where he was heavily
and poor hub involved in the deployment of PONs. Before Quantum
and removing excessive fiber can save tens to
centering— Bridge, he led a networks integration department
hundreds of dollars per household. at Lucent Technologies. Brian began his
originate in the telecommunications career with AT&T Bell
The amount of fiber per household is directly Laboratories. He has an MS and a BS in Electrical
planning phase.
proportional to the size of the serving area. Engineering from Worcester Polytechnic Institute.
Reducing a serving area by half reduces fiber per
household by 30 to 50 percent. Equipment
utilization is also affected by serving area size.
Fiber cost and equipment cost should be jointly
considered to arrive at an optimum serving area
size.

The network requires the least amount of fiber

when the hub is centered in the serving area.
When the hub is not centered, excess fiber is
required. An HC metric has been developed to
identify how well the hub is centered within the
serving area and to estimate the fiber impact. A
poorly centered hub may double the fiber per
household requirements. With an understanding
of the cost impact, an informed decision can be
made on possible adjustment of hub location or
serving area boundaries.

72 Bechtel Telecommunications Technical Journal

THE IMPACTS OF ANTENNA AZIMUTH
AND TILT INSTALLATION ACCURACY ON
UMTS NETWORK PERFORMANCE

Issue Date: January 2006

Abstract—Inconsistencies in setting up antenna azimuth and tilt during installation may reduce overall
network performance. However, the degree of quality degradation depends on the amount of the discrepancy
between the designed and installed parameters. The paper investigates the effect of these errors on UMTS
RF KPIs, including coverage, signal quality (Ec /Io), and soft-handoff areas. Two examples are studied that
include real measurement data. The studies show the effect of azimuth and tilt installation inaccuracies on
UMTS network quality.

INTRODUCTION coverage, the ratio of chip energy to interference

Esmael Dinan, PhD
ehdinan@bechtel.com
Aleksey A. Kurochkin
aakuroch@bechtel.com
ntenna azimuth and downtilt are two
A important optimization parameters in
universal mobile telecommunications system
(UMTS) networks. Optimization of these two
parameters can significantly improve system
performance. However, new networks sometimes
use inefficient optimization techniques and
implement default values. Furthermore, incon-
sistencies in setting these parameters during
installation vary the network coverage and
capacity. This paper presents the results of a
quantitative study that investigated the effect of
these parameters on UMTS network performance.
Many techniques are used to measure antenna
azimuth and tilt during installation. The accuracy
in setting up the azimuth and tilt depends on
the antenna installation processes and human
and instrumentation errors. Inefficient imple-
mentation and rigging processes may also cause
azimuth or tilt errors. The overall accuracy is
within ±10 degrees using most traditional
techniques. Usually, antenna azimuth errors are
independent for antennas belonging to different
sectors. New processes and instruments may
reduce these errors by several degrees, reduce
randomness in antenna orientations, and bring
errors consistently within the set tolerance.
This paper investigates the effects of azimuth and
tilt inaccuracies on network coverage and
performance and considers the three main UMTS
network system quality parameters: service
(Ec/Io), and soft handoff areas. Two exercises
are defined. A variety of errors are introduced
for all antennas, and a simulation is performed
for each case. At the end, the results are
compared and analyzed. Consistent use of the
new antenna installation processes is promoted to
limit the impact of inconsistencies. Suggestions
are also provided on acceptable installation
error limits for use as a baseline to develop
implementation processes.
ANTENNA AZIMUTH AND TILT SETTINGS AND
INCONSISTENCIES
ntenna azimuth and tilt errors (Figure 1) are
A randomly distributed among the sites and
sectors. For the purpose of this paper, azimuth
error is measured as the absolute difference
between the actual azimuth installed in the field
and the designed azimuth, as illustrated in
Figure 1a. In this definition, all azimuth errors
are positive. Tilt errors can be positive or
negative—uptilt errors are considered negative,
while downtilt errors are considered positive, as
shown in Figure 1b.
An antenna installation technician sets up the
azimuth using a compass and alignment tool. On
the top of the tower, the technician can use
several mechanisms to install the antenna.
However, the technician’s capabilities are
restricted by uncomfortable climbing status,
limited time, limited available tools, and

© 2006 Bechtel Corporation. All rights reserved. 73

 environmental factors. An example of an
ABBREVIATIONS, ACRONYMS, AND TERMS installation mechanism using landmarks and an
Ec/Io ratio of chip energy to optical alignment tool is shown in Figure 2. This
interference figure shows two pre-specified landmarks for the
GPS global positioning system technician to use from the top of the tower. In
KPI key performance indicator this example, the respective angles between
QoS quality of service the antenna aim point and Landmarks A and B
are set to 40 degrees (counterclockwise) and
RF radio frequency
–25 degrees (clockwise) from aim point to target.
RSCP received signal code power
Once the alignment is set, antenna tilt is adjusted
UMTS universal mobile
using a mechanical tilt bracket. Antenna tilt errors
telecommunications system
are caused by imperfect vertical adjustment of the
antenna support structure.
The accuracy in
setting up the Designed Tilt

azimuth and tilt

Negative
depends on Error
Positive
the antenna Error

installation
processes and
human and Positive Error
Field Azimuth
instrumentation
errors.
Designed Azimuth

(a) (b)

Figure 1. Antenna Azimuth and Tilt Errors

(a) Azimuth Error; (b) Tilt Error

Target A
True North

Optical
50° Actual Bearing
Alignment Tool
le
ng
tA
f fse
°O
40
Antenna Aim Point

-25 90° Specified Antenna

°O Azimuth
f fse
t An
gle
Target B

115° Actual Bearing

Antenna Support Structure

Figure 2. Example of an Antenna Azimuth Setup and Installation

74 Bechtel Telecommunications Technical Journal

Using the Six Sigma process improvement
methodology, Bechtel initiated a task force to
measure antenna installation accuracies [1]. The
implementation team analyzed the data related
to repeatability and reproducibility of different
antenna azimuth adjustment mechanisms. The
results demonstrated up to 10 degrees of error in
simple global positioning system (GPS)-based
adjustment methods. More advanced mecha-
nisms can provide accuracies within 5 degrees
with 95 percent probability of confidence.
Figure 3 illustrates another element used in the
study that is the subject of this paper: the
correlation of errors between sectors of the same
site. Scenario A illustrates the traditional
technique of pointing antennas individually,
leading to independent error in each sector.
This paper proposes using a technique that
offers a consistent error or the same error for
antennas belonging to the same site. In this
technique, shown in Scenario B, the azimuths of
the second and third antennas are adjusted
relative to the azimuth of the first-installed
antenna. This paper shows that this scenario,
offered by recent installation techniques,
provides better network performance than the
traditional method.
SIMULATION MODEL AND ASSUMPTIONS
his paper examines two example network
T clusters—one with 20 sites and one with
42 sites—that were simulated using planning and
optimization tools. These clusters are shown in
Figure 4. The simulation results help to analyze
the effect of azimuth and tilt settings on some
aspects of network performance. The following
tasks were included in the study:
• Select cluster areas, antenna types, default
site configuration, and system parameters
• Develop simulation scenarios, objectives,
and plans
• Develop project setup in the planning and
optimization tools and configure all the
parameters

Error = γ Error = ∝ Error = ∝ Error = ∝

Field Azimuth Field Azimuth

Designed Azimuth Designed Azimuth

Error = β Error = ∝

Scenario A Scenario B

Figure 3. Correlation of Errors Between Sectors of the Same Site

Scenario A – Traditional Azimuth Setting; Scenario B – Proposed Azimuth Setting

(a) (b)

Figure 4. Cluster Area Elevation Map

(a) 20 UMTS Sites – Traffic and Coverage Relevant Area: 17.17 km2
(b) 42 UMTS Sites – Traffic and Coverage Relevant Area: 26.14 km2

January 2006 • Volume 4, Number 1 75

 • Optimize all antenna azimuths and tilts using In the next step, a series of simulations were
recursive optimization algorithms (This performed to investigate the effect of azimuth
design will be considered to be the baseline and tilt errors on network performance. For both
design.) Scenarios A and B, a variety of errors were
• Execute the simulation and record the introduced for all the antennas. These errors were
statistics for the above scenarios and error randomly distributed among the cells. For each
parameters error set, the simulation was executed repeatedly
until a steady, consistent result was achieved.
• Analyze the data and compile the final Then the performance statistics, including
graphs coverage, interference, and soft handoff area
A standard default site configuration was were calculated and compared. Performance
considered. Cell sites included in the test cluster statistics were recorded and then analyzed to
UMTS network had the following configuration parameters: produce the final graphs.
performance
• Antenna radiation center heights in the The exercises described above were performed
sensitivity to range of 20 to 25 meters multiple times, each using a different antenna
azimuth and tilt type. The results help provide an understanding
• Node B transmission power = 20 watts
error increases of the effect of antenna types on the performance
• Pilot power = 2 watts graphs and conclusions. Overall behavior is
as beamwidth
• Traffic load = 50 percent, uniform distribution consistent with antennas having the same
is reduced. horizontal and vertical beamwidth. UMTS
• Total antenna feeder loss = 3 dB
network performance sensitivity to azimuth and
• Frequency = 2,150 MHz (downlink) tilt error increases as beamwidth is reduced. The
Two example projects were created in the relationship between error type and beamwidth
planning and optimization tools using the above is as follows:
configuration parameters. Other UMTS system • Horizontal beamwidth ↔ Azimuth error
parameters were set to default values. In the
baseline design, antenna azimuth and tilt • Vertical beamwidth ↔ Tilt error
configurations were optimized for maximum Simulation results presented in this paper were
overall performance of the test cluster. Therefore, performed with antennas that have 65-degree
changes in these parameters would result in horizontal beamwidth and 7-degree vertical
reduced network performance. Antenna azimuth beamwidth, which is considered to be a typical
and tilt were optimized using an automated antenna type in most UMTS networks.
recursive optimization tool (Radioplan GmbH’s
Wireless Network System [WiNeS]). The tool
prediction parameters and path loss matrix were SIMULATION RESULTS
tuned using drive test data. For the baseline imulation results are presented in Figures 5, 6,
design, a simulation was performed, including
coverage, interference, and soft handoff analysis.
S and 7. Figure 5 considers a simple single site

Single Site Coverage Versus Antenna Azimuth Error Single Site Coverage Versus Antenna Tilt Error
100 105
RSCP < –86 dBm RSCP < –86 dBm

99.5 100

99
Normalized Coverage Area

Normalized Coverage Area

95
98.5
90
98
85
97.5
80
97

96.5 75

96 70
0 5 10 15 20 25 30 –3 –2 –1 0 1 2 3
Average Antenna Azimuth Error Average Antenna Tilt Error
(a) (b)

Figure 5. Network Performance Versus Antenna Azimuth and Tilt Installation Error in a Single-Site Configuration
(a) Azimuth Error; (b) Tilt Error

76 Bechtel Telecommunications Technical Journal

 Coverage Gap Versus Antenna Azimuth Error Coverage Gap Versus Antenna Tilt Error
7 16
RSCP < –86 dBm, A RSCP < –86 dBm
RSCP < –86 dBm, B 14 RSCP < –92 dBm
6 RSCP < –92 dBm, A

Increase in Coverage Gap (Percentage)

Increase in Coverage Gap (Percentage)

RSCP < –92 dBm, B 12

5
10
4 8

3 6

4
2
2
1
0

0 –2
0 5 10 15 20 25 30 –3 –2 –1 0 1 2 3
Average Antenna Azimuth Error Average Antenna Tilt Error

(a) Area with RSCP < –86 dBm = 12.32%, Area with RSCP < –92 dBm = 4.80%

QoS Gap Versus Antenna Azimuth Error QoS Gap Versus Antenna Tilt Error
4.5 2.5
Ec /Io < –12 dB, A Ec /Io < –12 dB
4 Ec /Io < –12 dB, B
Increase in Service Quality Gap (Percentage)

Ec /Io < –13 dB

Increase in Service Quality Gap (Percentage)

Ec /Io < –13 dB, A 2

Ec /Io < –13 dB, B
3.5

3 1.5

2.5
1
2

1.5 0.5

1
0
0.5

0 –0.5
0 5 10 15 20 25 30 –3 –2 –1 0 1 2 3
Average Antenna Azimuth Error Average Antenna Tilt Error

(b) Area with Ec /Io < –12 dB = 4.0%, Area with Ec /Io < –13 dB = 1.01%

Soft Handoff Area Versus Antenna Azimuth Error Soft Handoff Area Versus Antenna Tilt Error
10 1
SHO Margin = 3 dB
9 SHO Margin = 5 dB, A SHO Margin = 5 dB
Increase in Soft Handoff Area (Percentage)

SHO Margin = 5 dB, B

Increase in Soft Handoff Area (Percentage)

0
SHO Margin = 3 dB, A
8
SHO Margin = 3 dB, B
7 –1

6
–2
5
–3
4

3 –4
2
–5
1

0 –6
0 5 10 15 20 25 30 –3 –2 –1 0 1 2 3
Average Antenna Azimuth Error Average Antenna Tilt Error

(c) Soft Handoff Area = 28.05% (Soft Handoff Margin = 5 dB),

Soft Handoff Area = 17.58% (Soft Handoff Margin = 3 dB)

Figure 6. Performance Graphs for 42-Site Cluster

January 2006 • Volume 4, Number 1 77

 Coverage Gap Versus Antenna Azimuth Error Coverage Gap Versus Antenna Tilt Error
2 7
– RSCP < –86 dBm
RSCP < –86 dBm, A
1.8 RSCP < –92 dBm
RSCP < –86 dBm, B 6

Increase in Coverage Gap (Percentage)

Increase in Coverage Gap (Percentage)

RSCP < –92 dBm, A
1.6 RSCP < –92 dBm, B
5
1.4
4
1.2

1 3

0.8 2
0.6
1
0.4
0
0.2

0 –1
0 5 10 15 20 25 30 –3 –2 –1 0 1 2 3
Average Antenna Azimuth Error Average Antenna Tilt Error

(a) Area with RSCP < –86 dBm = 5.76%, Area with RSCP < –92 dBm = 2.0%

QoS Gap Versus Antenna Azimuth Error QoS Gap Versus Antenna Tilt Error
4.5 4
Ec /Io < –12 dB, A Ec /Io < –12 dB
4 3.5
Increase in Service Quality Gap (Percentage)

Increase in Service Quality Gap (Percentage)

Ec /Io < –12 dB, B Ec /Io < –13 dB
Ec /Io < –13 dB, A
3.5 Ec /Io < –13 dB, B
3
3
2.5
2.5
2
2
1.5
1.5
1
1

0.5 0.5

0 0
0 5 10 15 20 25 30 –3 –2 –1 0 1 2 3
Average Antenna Azimuth Error Average Antenna Tilt Error

(b) Area with Ec /Io < –12 dB = 4.42%, Area with Ec /Io < –13 dB = 0.94%

Soft Handoff Area Versus Antenna Azimuth Error Soft Handoff Area Versus Antenna Tilt Error
8 1
SHO Margin = 5 dB, A
SHO Margin = 5 dB, B SHO Margin = 3 dB
7 SHO Margin = 3 dB, A 0
Increase in Soft Handoff Area (Percentage)

SHO Margin = 5 dB
Increase in Soft Handoff Area (Percentage)

SHO Margin = 3 dB, B

6
–1
5
–2
4
–3
3

–4
2

1 –5

0 –6
0 5 10 15 20 25 30 –3 –2 –1 0 1 2 3
Average Antenna Azimuth Error Average Antenna Tilt Error

(c) Soft Handoff Area = 36.34% (Soft Handoff Margin = 5 dB),

Soft Handoff Area = 23.0% (Soft Handoff Margin = 3 dB)

Figure 7. Performance Graphs for 20-Site Cluster

78 Bechtel Telecommunications Technical Journal

configuration to provide an initial reference result
for comparison purposes. In this example no
interference or inter-cell soft handoff areas exist;
only coverage plots are shown. Only Scenario A
was considered because in Scenario B all the site’s
antennas were rotated with the same azimuth
error; therefore, overall coverage performance
did not change.
As illustrated in Figure 5a, coverage shrinks
when azimuth error increases. Coverage is
reduced by 4 percent when there is a 30-degree
error in azimuth setting. The coverage area has an
almost inverse linear relationship with azimuth
error. Figure 5b shows that coverage is also very
sensitive to downtilt errors. Coverage changes up
to 29 percent when downtilt error varies in the
range of –3 to +3 degrees. This example shows a
system with no interference and inter-cell soft
handoff coverage. To study and capture real
network performance behavior, multiple sites
are needed.
Figure 6 shows the results for a 20-site cluster,
and Figure 7 shows the results for a 42-site
cluster. These provide realistic examples in
performance graphs.
Azimuth errors in the range of 0 to 30 degrees
were considered for both Scenarios A and B. Tilt
errors varied between –3 and +3 degrees. The
areas are represented as the percentage of the
cluster area. The performance graphs are
categorized by coverage area, coverage quality,
and soft handoff area.
Coverage Area
Coverage area is measured in reference to
received signal code power (RSCP). Two
definitions were considered for coverage gap: the
area with less than –86 dBm RSCP and the area
with less than –92 dBm RSCP. Figures 6a and 7a
show the variations in coverage gaps when there
are inconsistencies in antenna azimuth and tilt
settings. A higher coverage percentage and
fewer coverage gaps is desirable when
implementing a UMTS network.
Coverage Quality
Quality of service (QoS) or coverage quality is
measured by Ec/Io. Two definitions were
considered for QoS gap: the area with Ec/Io less
than –12 dB and the area with Ec/Io less than
–13 dB. Figures 6b and 7b show the variations in
areas with QoS gaps when there are incon-
sistencies in antenna azimuth and tilt settings. A
higher QoS and fewer QoS gaps is desirable when
implementing a UMTS network.
January 2006 • Volume 4, Number 1
Soft Handoff Area
Soft handoff area is defined as the area covered
by more than one sector belonging to different
Node Bs. Two different settings were considered
for soft handoff threshold. Performance graphs
are shown for soft handoff areas when the soft
handoff margin is 3 dB and 5 dB. Figures 6c and
7c show the variations in soft handoff areas when
there are inconsistencies in antenna azimuth and
tilt settings. It is desirable to achieve the target
soft handoff area recommended by the service
operator when implementing a UMTS network.
A smaller soft handoff area results in increased Azimuth error
call drop rate, and a higher soft handoff area in the range of
results in inefficient use of radio resources and
6 to 8 degrees
excessive interference.
is tolerable,
Careful investigation of the results of the graphs
depending on
in Figures 6 and 7 leads to the following
conclusions: the installation
scenario and initial
• Antenna Azimuth: Network performance
coverage area.
variations depend on antenna azimuth error
variations and the installation process.
Overall degradation in Scenario B is 40 to 60
percent less than in Scenario A. Therefore,
the same error in all sectors is preferable.
Azimuth error in the range of 6 to 8 degrees
is tolerable, depending on the installation
scenario and initial coverage area.
Performance degrades noticeably if the error
is greater than 10 degrees. Soft handoff
areas are the least sensitive to azimuth
error. The coverage gap is 30 percent
greater with 30 degrees of error in antenna
azimuth. A comparison of the coverage
graphs in Figures 6 and 7 shows that when
the coverage/quality gap is smaller, its
sensitivity to error is higher.
• Antenna Tilt: Both coverage and quality
performances are very sensitive to antenna
tilt variations. There is up to a 100 percent
increase in coverage and quality gaps with
±3 degrees of tilt error. Soft handoff areas are
the least sensitive to tilt error. The graphs in
Figures 6c and 7c show less than a 10 percent
variation in soft handoff area with ±3 degrees
of tilt error.
SUMMARY AND CONCLUSIONS
oth the 20- and 42-site examples produce
B consistent network performance behavior
and lead to the same conclusions. If equal errors
are introduced to cell site sectors, there is less
network performance degradation (Scenario A),
compared with random errors (Scenario B). For
79
 practical purposes, azimuth error in the range
of 6 to 8 degrees is tolerable for network
performance. Performance degradation is
noticeable if the azimuth error is greater than
10 degrees. Network performance is almost ten
times more sensitive to antenna tilt variations,
compared with azimuth variations. Both
coverage and quality gaps increase by up to
100 percent with ±3 degrees of tilt error.
If possible, only one antenna should be oriented
and the other antenna azimuths set in reference
to that one (Scenario B). However, rooftop size
Both coverage
and configuration may interfere with this
and quality gaps
recommendation. If Scenario B installation
increase by up techniques can be applied to the site, simpler
to 100 percent methods (instead of the more expensive methods)
with ±3 degrees have the same effect on network performance.
of tilt error. Considering these conclusions, the following
UMTS network implementation standard can be
Tilt setting
practically recommended for antenna azimuth
tolerance of
and tilt tolerances:
±0.5 degrees
1. For the Scenario A technique: Azimuth
is recommended.
setting tolerance of ±6 degrees
2. For the Scenario B technique: Azimuth
setting tolerance of ±8 degrees
3. For both scenarios: Tilt setting tolerance of
±0.5 degrees
The cluster with more sites experiences less
network quality degradation due to azimuth and
tilt errors. However, this could be a subject for
further studies.

ACKNOWLEDGMENTS
he authors would like to thank Lacy Kiser
T from the Bechtel Six Sigma Team and Jeff
Bryson from the Bechtel Construction Team for
the valuable data and information they provided.
Special thanks go to Radioplan GmbH for
providing WiNeS software for this study.
REFERENCES
[1] Six Sigma PIP TI-81, Report and Data Analysis,
Bechtel Telecommunications, 2005.
[2] E. Dinan, “UMTS RF Network Optimization
Process,” Document Number 3DP-T04G-50009,
Bechtel Telecommunications Network Planning
Department, 2005.
80
BIOGRAPHIES
Esmael Dinan, a senior
RF technologist with Bechtel
Telecommunications, has
been instrumental in many
aspects of the business unit’s
research activities and
the Cingular RF engineering
project. He has designed
and engineered an RF
engineering data management system, developed
Cingular project RF engineering processes and
procedures, designed UMTS networks, and
verified and tested Dupont cryogenic TMA
performance.
Before joining Bechtel in 2002, Dr. Dinan
was product manager for the GMPLS control
plane of the RAYStar DWDM optical switch
at Movaz Networks, and lead network
architect at MCI. He has conducted research
and development on access methods and
performance modeling of 3G wireless commu-
nications and high-speed optical networks.
Dr. Dinan received his PhD in Electrical
Engineering from George Mason University,
Fairfax, Virginia, and is a registered Professional
Engineer in Maryland. He has authored more
than 25 conference papers and journal articles
and has filed a patent on a novel signaling
mechanism developed for 3G cellular networks.
He is a member of the Institute of Electrical and
Electronics Engineers.
Aleksey Kurochkin is
currently senior director,
Site Development and
Engineering, in the Bechtel
Telecommunications
Technology group, a group
that he originated. He is
experienced in international
telecommunications
business management and network imple-
mentation. Before joining Bechtel, he worked
at Hughes Network Systems, where he built
an efficient multi-product team focused on
RF planning and system engineering. His
engineering and marketing background has
given him both theoretical and hands-on
knowledge of most wireless technologies.
Aleksey has an MSEE/CS degree in Automatic
Telecommunications from Moscow Technical
University of Communications and Informatics,
Russia.
Bechtel Telecommunications Technical Journal
2.4 GHz Wi-FiTM PHASED ARRAY
ANTENNA EVALUATION

Issue Date: January 2006

Abstract—This paper examines the application, coverage, and security of a 2.4 GHz IEEE 802.11b (Wi-Fi TM)
phased array antenna system. Two antennas are installed in real-world scenarios. One antenna provides
outdoor-to-indoor and outdoor-to-outdoor coverage, and the other antenna provides outdoor-to-outdoor coverage
only. Throughput is measured using a variety of typical application methods. The results show both the
advantages and disadvantages of using a phased array antenna system.

INTRODUCTION transmitter and a single receiver, is provided by

merging advances in wireless communication
E technology and its application and the ever-
increasing demand for the availability of wireless
high-bit-rate devices have resulted in numerous
high-tech innovations. Recently, phased array
antennas have been proposed to reduce the
time and cost of providing high-bit-rate
wireless fidelity (Wi-Fi™) to indoor and outdoor
wireless users.
This paper examines the application, coverage,
and security of 2.4 GHz IEEE 802.11b (Wi-Fi)
phased array antennas. Two of these antennas are
installed in real-world scenarios on the roof of
Bechtel Park Building 5 (BP5) in Frederick,
Maryland. One antenna covers Bechtel Park
Building 2 (BP2) (outdoor to indoor) and the
business area behind it (outdoor to outdoor), and
the other antenna covers a large outdoor
residential area (outdoor to outdoor).
Throughput is measured using a variety of
typical application methods.
This paper describes the theory, test architecture,
and setup of these 2.4 GHz Wi-Fi phased
array antennas; provides extensive experimental
results and a technical interpretation of
these results; summarizes the advantages and
disadvantages of using this technology; and
provides practical suggestions for its use.
Glenn A. Torshizi
THEORY
gtorshiz@bechtel.com
nformation capacity I, which limits the amount
I of information that can be sent between a single
Claude Shannon in the following formula [1]:
I = log (1 + S/N) bits/s/Hz
In the above formula, S is the received signal
power and N is the noise power, and the
information is measured in bits per second per
hertz of transmission bandwidth available. Due
to the maximum power limitation set by the
Federal Communications Commission (FCC) on
an already crowded unlicensed spectrum,
Shannon’s formula does not seem to be useful in
increasing the information capacity. However,
multi-antenna arrays or smart antennas have
been introduced to increase the information
capacity.
The simplest multi-antenna array—the “steered
beam” or “phased” array—consists of many
individual antennas that each transmit the same
signal using different phase shifts, which are
arranged so that different signals interfere
constructively in one direction and destructively
in every other direction. Transmitting and
receiving antenna phase shifts can be steered
electronically.
Beam steering increases the signal directed
toward an intended receiver and reduces the
reception of stray signals intended for other
targets. The receiver perceives these stray signals
as noise. Overall, the relative power gain
obtained using an array with m antennas is
roughly a factor of m [2].

© 2006 Bechtel Corporation. All rights reserved. 81

 antennas on the roof of BP5. BP2, a typical three-
ABBREVIATIONS, ACRONYMS, AND TERMS story office building approximately 70 by 210 feet
AES Advanced Encryption Standard and 45 feet high, is positioned directly in front of
AGL above ground level and 600 feet from the first antenna, which is
installed on top of BP5 at a height of
BSS basic service set
approximately 55 feet above ground level (AGL)
EIRP effective isotropic radiated power with no downtilt. See Figure 2 for a photograph
FCC Federal Communications depicting a view of BP2 from Antenna 1 on BP5.
Commission
FIPS Federal Information Processing
Standard
IBSS independent BSS
Beam steering
LOS line of sight
increases the signal
PC personal computer
directed toward an
RF radio frequency
intended receiver
WEP wired equivalent privacy
and reduces the TM
Wi-Fi wireless fidelity
reception of stray
WLAN wireless local area network
signals intended for
other targets.
Figure 2. View of BP2 from Antenna 1 on BP5
TEST SETUP AND PROCEDURES
he phased array antenna used in this
T evaluation is a three-channel unlicensed (FCC
Part 15) wireless device. It allows point-to-point
The tools used to measure the radio frequency
(RF) environment and signal quality included
AirMagnet™ for iPaq, AirMagnet laptop, Cisco
packet communication to client devices through
Aironet™ 350 personal computer (PC) card, and
an integrated high gain, electronically steered
Antenex® YE240015 antenna.
transmitting and receiving antenna. The antenna
is configured in a 100-degree pattern horizontally The antenna uses a Wi-Fi switch that supports an
and a 12-degree pattern vertically. The antenna infrastructure basic service set (BSS) mode. The
pattern is divided into 13 focused areas. Channel network consists of one access point connected to
assignments and other settings for each area can the wired network infrastructure and a set of
be changed to optimize overall Wi-Fi operation wireless end stations. The independent basic
within the full antenna pattern. The antenna is service set (IBSS), in which wireless stations
approximately 43 by 36 by 7 inches and weighs intercommunicate directly without using an
80 pounds. Figure 1 shows the placement of two access point or connection to a wired network, is
not supported.

Wireless wall software [3] was used to provide

Figure 1. Placement of Antennas on BP5 Roof on
Bechtel Campus in Frederick, Maryland
system security, since wired equivalent privacy
(WEP) is susceptible to wireless interception
and fraud [4, 5]. This software manages the
authentication process using an IEEE 802.11
framework and software-specific extensions to
prevent session hijacking or denial of service
attacks at any point. The software adheres
to Advanced Encryption Standard (AES) [6]
to protect sessions and networks. AES is a
Federal Information Processing Standard (FIPS
Publication 197) that specifies a cryptographic
algorithm to be used by US government
organizations to protect sensitive information.
The wireless wall software operates at Layer 2 of
the networking stack, providing the highest level
of protection against radio-based attacks.

82 Bechtel Telecommunications Technical Journal

The test setup goal was to map antenna range and
coverage in open outdoor environments and
inside BP2. The effective isotropic radiated power
(EIRP) level was 41 dBm.

EXPERIMENTAL RESULTS AND ANALYSIS

ignal, noise, speed, and throughput were
S measured at various locations inside, just
outside of, and on the roof of BP2. Similar
measurements were also taken at locations off
Figure 3. Signal Level Measurements
of the Bechtel campus.
The noise
measurements
Outdoor-to-Indoor Measurements
Data was collected from 11 locations on the third show an
and second floors of BP2 and from 10 locations on approximately
the first floor of BP2. Figure 3 shows the variation flat noise level
of the signal level as a function of the relative
that averages
position of the measurement points with respect
to the antenna. For each building floor, the –95 dBm.
measurement points were divided into three
groups:

• Group A: Points in the areas directly facing

the antenna Figure 4. Speed Measurements
• Group B: Points in the middle of the
building
• Group C: Points deeper inside the building

The measurements for each group were averaged

over the nodes in that group and then compared
to the averages of the other groups for each floor.
The noise measurements show an approximately
flat noise level that averages –95 dBm. Speed and
throughput measurements are summarized in
Figures 4 and 5, respectively.

The results show that good indoor coverage exists

at locations near the windows facing Antenna 1
Figure 5. Throughput Measurements
on the first, second, and third floors; however, the
signal degrades significantly away from the
windows toward the inside of the building.

Outdoor-to-Outdoor Measurements
Figure 6 shows the outdoor measurements for
Antenna 1 around BP2. Signal strength was
measured at several points in the front and back
and on the roof of BP2. The measurements show
that there is a good signal just outside of BP2 on
the ground facing Antenna 1 and on the roof of
BP2. However, the signal did not completely
penetrate through the building, resulting in little
or no coverage immediately behind BP2 at ground
level. At roof level, if the signal was blocked by
Figure 6. Outdoor Measurements for Antenna 1 Around BP2
the metal skirt around the air conditioning
system, the signal degraded noticeably.
Otherwise, the signal level was the same or better.

January 2006 • Volume 4, Number 1 83

 Figure 7. Test Locations B, C, D, E, and F from Antenna 1 at Location A
Figure 8. Test Locations G, H, I, J, K, L, and M from Antenna 2 at Location A
Outdoor measurements for Antennas 1 and 2
were taken at several locations off of the Bechtel
campus. Figure 7 shows the outdoor-to-outdoor
measurement points from Antenna 1, and
Figure 8 shows the outdoor-to-outdoor points for
Antenna 2.
The test locations were chosen so that line of
sight (LOS) or near LOS could be maintained.
To ensure that LOS existed, point-to-point
profiles were evaluated and examined for all
measurement points. Figure 9 is an example of
one of these profiles.
Figure 10 shows the measurement results
from points B, C, and G. The following are
worth noting:
84
• At other locations away from Antennas 1
and 2, the signal was very weak. At several
locations, the signal level was adequate,
but connectivity to the test laptop could
not be established using just the wireless
local area network (WLAN) adapter. When
an Antenex antenna was used, connectivity
was established.
• Many IEEE 802.11 applications are already
being used around the test locations.
• The Antenex antenna greatly enhanced the
performance in low signal strength areas.
• In marginal areas, the AES algorithm
degraded performance beyond use.
Bechtel Telecommunications Technical Journal
 Phased array
Figure 9. Point-to-Point Profile from Antenna 2 at Location A to Test Location J
antennas are a
good choice for
TRADEMARKS
outdoor-to-outdoor
AirMagnet is a trademark of AirMagnet, Inc.
scenarios where
Aironet is a trademark of Aironet Wireless LOS can be
Communications, Inc.
maintained.
Antenex is a registered trademark of
Antenex, Inc.

Wi-Fi is a trademark of the Wireless Ethernet

Compatibility Alliance, Inc.

REFERENCES
[1] C.E. Shannon, “A Mathematical Theory of
Figure 10. Measurement Results from Points B, C, and G
Communication,” Bell System Technical Journal,
Vol. 27, pp. 379–423 and pp. 623–656, 1948.
[2] S.H. Simon, A.L. Moustakas, M. Stoytchev, and
CONCLUSIONS H. Safar, “Communication in a Disordered
World,” Physics Today online, 2001.
his paper examined the use of phased array
T antennas for outdoor-to-indoor and outdoor-
to-outdoor applications using extensive
[3] Cranite Systems, Inc., “Best Practices: Wireless
LAN Design, Implementation and Management,”
White Paper, September 2003.
measurements at several indoor and outdoor [4] W.A. Arbaugh, N. Shankar, Y.C.J. Wan, and
points. As the experimental results reveal, Z. Kan, “Your 802.11 Wireless Network Has
No Clothes,” IEEE Transactions on Wireless
using the system has both advantages and Communications, December 2002, Vol. 9, No. 6,
disadvantages; these are mainly application pp. 44–51.
specific. The system is conveniently easy and fast [5] T. Karygiannis and L. Owens, “Wireless Network
to install, and it appears to be the only choice for Security: 802.11, Bluetooth and Handheld
Devices,” NIST Special Publication 800-48,
applications such as a marina. For the outdoor-to- November 2002.
indoor scenarios, it is a good choice for narrow [6] D.J. Welch and S.D. Lathrop, “A Survey of
multistory buildings. The signal degrades toward 802.11a Wireless Security Threats and Security
the inside of wide buildings (such as BP2). Use Mechanisms,” Army G6 Technical Report
ITOC-TR-2003-101, 2003.
of repeaters is recommended to enhance the
signal in deep building areas that are not covered.
Phased array antennas are also a good choice for
outdoor-to-outdoor scenarios where LOS can be
maintained; however, since the frequency is
unlicensed, interference is an issue over an
extended range. When security is enhanced using
the AES algorithm, there is also a tradeoff
between the level of security in the system
and throughput.

January 2006 • Volume 4, Number 1 85

 BIOGRAPHY
Glenn Torshizi joined Bechtel
in 2001 and is currently a staff
scientist/engineer at Bechtel
Telecommunications’ Training,
Demonstration, and Research
(TDR) Laboratory in Frederick,
Maryland. Before relocating to
Frederick, he spent more than
3 years working as an RF design
engineer on the Bechtel AWS
GSM, GPRS, and UMTS Program in Philadelphia,
Pennsylvania, and as the market RF lead in Harrisburg,
Pennsylvania, and Hackensack, New Jersey.

Before joining Bechtel, Glenn was involved in planning,

optimizing, and integrating the Triton PCS TDMA
system in Norfolk, Virginia, and the Cricket
Communications CDMA system in Pittsburgh,
Pennsylvania. As a technical expert witness on
numerous planning and zoning boards, he was very
successful in obtaining final site approvals.

Glenn has a BS in Physics from Southwestern

Oklahoma State University and an MS in Physics from
the University of Tennessee, Knoxville. He has done
research in relativistic heavy ion physics at Oak Ridge
National Laboratory and Brookhaven National Laboratory.

86 Bechtel Telecommunications Technical Journal

BECHTEL
TELECOMMUNICATIONS
echtel Telecommunications, a unit of Bechtel
B Corporation, provides turnkey deployment services
that include network planning, RF design, engineering,
Telecommunications Leadership
and project and construction management for the
deployment of wireless, wireline, and other
TIMOTHY D. STATTON
telecommunication facilities worldwide. We are
Executive Vice President,
headquartered near Washington, DC, and have major Bechtel Corporation and
offices in London and Sydney, and numerous project President, Bechtel Telecommunications
offices worldwide.

Bechtel has performed telecommunications work for ROBERT CASAMENTO

over 35 years. We have successfully completed more Principal Vice President and
than 140 major telecommunications projects worldwide, General Manager
including more than 85,000 wireless sites; 23,000
kilometers of wireline fiber; and communication centers
such as POPs, NOCs, and data centers.
LARRY ALBEE
Bechtel Telecommunications is a global leader in Principal Vice President and
applying proven project and construction management Manager of Functional Operations
tools to geographically and technically complex
network builds. We deliver the essential elements of
project success while setting records for speed, JAMES A. IVANY
reliability, quality, and cost control. Principal Vice President and
Chief Financial Officer

J. S. (JAKE) MACLEOD
Principal Vice President and
Chief Technology Officer

LEE LUSHBAUGH
Principal Vice President and
General Manager, Americas

RICK ASTLEFORD
Principal Vice President and
General Manager, Europe, Africa,
Middle East, Asia, and Asia-Pacific

January 2006 • Volume 4, Number 1 87

 Bechtel Corporation

echtel Corporation is a privately held

B company headquartered in San Francisco,
California, with approximately 40,000 employees,
more than 40 offices worldwide, and 2004
revenues of approximately $17.4 billion. Founded
in 1898, Bechtel has been under the leadership of
its founding family for four generations. Riley P.
TECHNOLOGY EVALUATION Bechtel, great-grandson of the founder, is the
current chairman and chief executive officer.
ey to Bechtel's deployment success is our
K technology expertise. To provide our
clients with complete turnkey deployment
Bechtel is one of the world's premier engineering,
construction, and project management companies.
services, we offer expert technology research
• The Financial Times recently listed Bechtel
and testing services. We have assembled a
as the “World’s Most Respected Company”
full-service Technology Group of leading
in the property and construction industries.
professionals that provides our customers
with neutral technology services and • Global Finance magazine named Bechtel
oversees our in-house telecommunications the “World’s Best Company” in the
laboratories: the Bechtel Wireless Test Bed construction sector for 5 years in a row.
(BWTB) and the Training, Demonstration
• Engineering News-Record ranked Bechtel as
and Research (TDR) Laboratory.
Number 1 on the list of the “Top 400
The BWTB, an over-the-air wireless test bed, Contractors” for 7 consecutive years.
enables us to analyze network products and
• Engineering News-Record also ranked
technologies for our clients in a real-world
Bechtel Number 1 Telecommunications
environment. It operates 24/7 and can be
Contractor from 2002 to 2004.
dynamically configured to meet specific
testing requirements without sacrificing the Because of our unparalleled industry recognition
client’s network. and financial stability, we are the partner of choice
The TDR Laboratory offers our clients for network deployment.
the ability to integrate equipment from
multiple vendors to identify interoperability
issues, monitor product performance, and
ensure compatibility with legacy systems,
thereby reducing the risk of imple-
menting new technologies and paving the
way for investment in network upgrades
or expansion.

88
 Bechtel Offices

Abu Dhabi Houston Moscow Shanghai

Tel: (971-2) 631-8999 Tel: (1-713) 235-2000 Tel: (7-095) 961-3020 Tel: (86-21) 3305-4567
Fax: (971-2) 631-0091 Fax: (1-713) 960-9031 Fax: (7-095) 961-3015 Fax: (86-21) 6391-2999

Al Khobar, Saudi Arabia* Istanbul New Delhi Singapore

Tel: (966-3) 882-5288 Tel: (90-212) 278-2900 Tel: (94-124) 509-8000 Tel: (65) 6332-1220
Fax: (966-3) 882-5306 Fax: (90-212) 278-2913 Fax: (94-124) 234-3110 Fax: (65) 6332-1335

TURNKEY Bangkok Jakarta New York Taipei*

DEPLOYMENT Tel: (66-2) 545-7100 Tel: (62-21) 5291-6000 Tel: (1-212) 689-0100 Tel: (886-2) 2703-2880
SERVICES Fax: (66-2) 545-7777 Fax: (62-21) 5296-2278 Fax: (1-212) 689-4829 Fax: (886-2) 2703-2885

Network Planning Beijing Kuala Lumpur Oak Ridge, Tennessee Tokyo

and RF Tel: (86-10) 8453-9998 Tel: (603) 2078-2050 Tel: (1-865) 220-2000 Tel: (81-3) 3214-4481
Fax: (86-10) 8453-8303 Fax: (603) 2078-1950 Fax: (1-865) 220-2100 Fax: (81-3) 3214-2596
Engineering
Brisbane Lima Paris Toronto
Tel: (61-7) 3295-1400 Tel: (51-1) 445-8746 Tel: (33-1) 4720-5304 Tel: (1-905) 362-6600
Site Acquisition and
Fax: (61-7) 3295-1450 Fax: (51-1) 446-1255 Fax: (33-1) 4720-5506 Fax: (1-905) 362-6767
Right of Way

Buenos Aires London (Hammersmith Rd.) Rio de Janeiro Tysons Corner, Virginia
Procurement
Tel: (54-11) 4315-8000 Tel: (44-208) 846-5111 Tel: (55-21) 2579-2400 Tel: (1-703) 748-9300
and Contracting Fax: (54-11) 4315-0081 Fax: (44-208) 846-6940 Fax: (55-21) 2579-2413 Fax: (1-703) 748-9442

Construction and Cairo London (Pilgrim St.) Riyadh Warrington, England

Equipment Installation Tel: (20-2) 754-6763 Tel: (44-207) 651-7777 Tel: (966-1) 482-9051 Tel: (44-1925) 857-000
Fax: (20-2) 754-6804 Fax: (44-207) 651-7972 Fax: (966-1) 482-9563 Fax: (44-1925) 857-001
Integration and
Optimization Calgary Louisville San Diego Washington, D.C.
Tel: (1-403) 290-5000 Tel: (1-502) 244-2574 Tel: (1-619) 687-8700 Tel: (1-202) 828-5200
Fax: (1-403) 290-5050 Fax: (1-502) 244-2211 Fax: (1-619) 687-8787 Fax: (1-202) 785-2645

Frederick, Maryland Manila San Francisco *Joint-venture office

Tel: (1-301) 228-6000 Tel: (63-2) 751-1000 Tel: (1-415) 768-1234

Fax: (1-301) 228-2200 Fax: (63-2) 751-1005 Fax: (1-415) 768-9038

Glendale, Arizona Mexico City Santiago

Tel: (1-602) 368-1500 Tel: (52-55) 5201-3500 Tel: (56-2) 675-1000
Fax: (1-602) 368-1615 Fax: (52-55) 5202-1785 Fax: (56-2) 232-5208

Hong Kong Montreal Seoul

Tel: (852) 2970-7000 Tel: (1-514) 871-1711 Tel: (82-2) 512-2921
Fax: (852) 2840-1272 Fax: (1-514) 871-1392 Fax: (82-2) 512-2931
 Bechtel Telecommunications
Technical Journal
An Independent Analysis of Current

Telecommunications Operational Issues January 2006

Global Headquarters
Bechtel Telecommunications Authors Contents
5295 Westview Drive
Frederick, Maryland 21703-8306

Bechtel Telecommunications Technical Journal

Tel: +1 301-228-7500 (Main) or v Foreword
Tel: +1 800-946-3632 (Toll Free)
vii Editorial
Fax: +1 301-694-6971
Timothy D. Statton and Jake MacLeod 1 Telecommunications Research and Development in the United States:
London A State of Crisis?
Bechtel Telecommunications
S. Rasoul Safavian, PhD 13 IP Multimedia Subsystems (IMS): A Standardized Approach to
245 Hammersmith Road
All-IP Converged Networks
Hammersmith
London W6 8DP Brian Coombe 37 A Survey of MEMS-Enabled Optical Devices – Applications and
United Kingdom Drivers for Deployment
Tel: +44 208-846-5093
Fax: +44 208-846-4938
Jungnam Yun, PhD 45 PHY/MAC Cross-Layer Issues in Mobile WiMAX
(POSDATA America R&D Center) and (Invited Paper)
Prof. Mohsen Kavehrad, PhD
(The Pennsylvania State University [CITCTR])
www.bechteltelecoms.com
Peter Moskal and Krishnamurthy Raghu 57 ANSI/TIA Standard 222 – Structural Standard for
Antenna Supporting Structures and Antennas:
A Comparison of Revisions F and G

Brian Perkins 65 Reducing the Amount of Fiber in Fiber-to-the-Home Networks

Esmael Dinan, PhD, and Aleksey A. Kurochkin 73 The Impacts of Antenna Azimuth and Tilt Installation Accuracy
on UMTS Network Performance

Glenn A. Torshizi 81 2.4 GHz Wi-FiTM Phased Array Antenna Evaluation

Volume 4, No. 1
2006

T ELECOMMUNICATIONS Volume 4 No. 1

T ELECOMMUNICATIONS
7211