This action might not be possible to undo. Are you sure you want to continue?

**Tapan K. Sarkar1, Zhong Ji1, Kyungjung Kim1, Abdellati' Medourf, and Magdalena Salazar-Palma3·· .
**

Department of Electrical Engineering and Computer Science Syracuse University 121 Link Hall, Syracuse, New York 13244-1240, USA Tel: 1 (315) 443-3775; Fax: 1 (315) 443-4441; E-mail: tksarkar@syr.edu

1

2Ecole Nationale des SCiences Appliquees Abdelrnalek Essadi University . Tangier, Morocco E-mail: amedouri@ieee.org 3Grupo de Microondas y Radar, Dpto. Senales, Sistemas y Radiocommunicacion, ETSI Telecommunication Universidad Politecnica de Madrid . Ciudad Universitaria, 28040 Madrid, Spain E-mail: Salazar@gmr.ssr.upm.es

Abstract

In order to estimate the signal parameters accurately for mobile systems, it is necessary to estimate a system's propagation characteristics through a medium. Propagation analysis provides a good initial estimate of the signal characteristics. The ability to accurately predict radio-propagation behavior for wireless personal communication systems, such as cellular mobile radio, is becoming crucial to system design. Since site measurements are costly, propagation models have been developed as a suitable, low-cost, and convenient alternative. Channel modeling is required to predict path loss and to characterize the impulse response of the propagating channel. The path loss is associated with the design of base stations, as this tells us how much a transmitter needs to radiate to service a given region. Channel characterization, on the other hand, deals with the fidelity of the received signals, and has to do with the nature of the waveform received a receiver. The objective here is to design a suitable receiver that will receive the transmitted signal, dlstorted-due to the multipath and dispersion effects of the channel, and that will decode the transmitted signal. An understanding of the various propagation models can actually address both problems. This paper begins with a review of the information available on the various propagation models for both indoor and outdoor environments. The existing models can be Classified into two major Classes: statistical models and site-specific models. The main characteristics of the radio channel - such as path loss, fading, and time-delay spread - are discussed. Currently, a third alternative, which includes many new numerical methods, is being introduced to propagation prediction. The advantages and disadvantages of some of these methods are summarized. In addition, an impulse-response characterization for the propagation path is also presented, including models for small-scale fading. Finally, it is shown that when two-way communication ports can be defined for a mobile system, it is possible to use reciprocity to focus the energy along the direction of an intended user without any explicit knowledge of the electromagnetic environment in which the system is operating, or knowledge of the spatial locations of the transmitter and the receiver.

at

Keywords; Land mobile radio cellular systems; land mobile radio propagation factors; communication channels; multipath channels; fading channels; transient response; reciprocity .

1. Introduction

he commercial success of cellular conununications, since its initial implementation in the early 1980s, has led to an intense interest among wireless engineers in understanding and predicting radio-propagation characteristics in various urban and suburban areas, and even within buildings .. As the explosive growth of mobile communications continues, it is very valuable to have the

IEEE ~ntennas and Propagation Magazine. Vol. 45, No.3, June 2003

T

capability of determining optimum base-station locations, obtaining suitable data rates, and estimating their coverage, without conducting a series of propagation measurements, which are very expensive and time consuming. It is therefore important to develop effective propagation models for mobile communications, in order to provide design guidelines for mobile systems. 51

**2. Definitions and Terminology used for Characterizing Various Parameters of a Propagation Channel
**

In order to understand the nature of the to be presented, several definitions and the narrowband and wideband wave propagation are first described, in order to familiarize the nology and parameters of the problem. models that are going terminology for both over a radio channel reader with the termi-

Multipath propagation causes severe dispersion of the transmitted signal. The expected degree of dispersion is determined through the measurement of the power-delay profile of the channel. The power-delay profile provides an indication of the dispersion or distribution of transmitted power over various paths in a multi path model for propagation. The power-delay profile of the channel is calculated by taking the spatial average of

Ih (t t over a

Ih(t)12

local area. By making several local-area measurements of

2.1 Path Loss

Path loss (PL) is a measure of the average RF attenuation suffered by a transmitted signal when it arrives at the receiver, after traversing a path of several wavelengths. It is defined by [lj (1)

for different locations, it is possible to build an ensemble of powerdelay profiles, each one representing a possible small-scale multipath channel state [3, 10]. A typical plot of the power-delay profile is shown in Figure I. Many multipath-channel parameters are derived from the power-delay profile. Power-delay profiles are measured using wideband channel-sounding techniques, and are presented in the form of plots of the received power as a function of an additional or excess delay with respect to a fixed time-delay reference. There is a delay between the time of transmission of the signal and when it is received, due to the finite velocity of propagatiori of the electromagnetic signal. However, additional delay may be introduced by the propagation medium, as well. A mobile channel exhibits a continuous multipath structure; hence, the power-delay profile can be thought of as a density function, of the form

where 1) and

p,.

are the transmitted and received power, respec-

tively. In free space, the power reaching the receiving antenna which is separated from the transmitting antenna by a distance dis given by the Friis free-space equation:

(4) (2)

where 01 and G; are the gain of the transmitting and the receiving antenna, respectively". L is the system loss factor, not related to propagation. A is the wavelength in meters. It is clear that Equation (2) does not hold for d '" 0 . Hence, many propagation models use a different representation for a close-in distance; do, known as the received-power reference point. This is typically chosen to be I m. In realistic mobile radio channels, free space is not the appropriate medium. A general PL model usesa parameter, r, to denote the power-law relationship between the separation distance and the received power. So, path loss (in decibels) can' be expressed as [2] PL(d) where '" PL(do)+lOr characterizes 10g(d/dO)+ Xu, (3)

**2.3 Time-Delay Spread
**

Time dispersion varies widely in a mobile radio channel, due to the fact that reflections and scattering occur at seemingly random locations, and the resulting multipath channel response appears random, as well. Because time dispersion is dependent on the geometrical position relationships among the transmitter, the receiver, and the surrounding physical environment, some parameters that can grossly quantify the multipath channel are used. They are described in the following subparagraphs.

r =2

free ~pace. However,

higher for wireless channels.

X;'

r

is generally

denotes a zero-mean Gaussian

random variable of standard deviation a, which reflects the variation, on average, of the received power that naturally occurs when a PL model of this type is used. Path loss is the main ingredient of a propagation model. It is related to the area of coverage of mobile systems.

OdB

**2.2 Power-Delay Profile
**

Random andcomplicated radio-propagation channels can be characterized using the impulse-response approach. For each point in the three-dimensional environment, the channel is a linear filter with impulse response h (t ). The impulse response provides a wideband characterization of the propagating channel, and contains all of the information necessary to simulate or analyze any type of radio transmission through that channel.

52

L-~~

~

~

~r

Euas DoIoy""

Figure 1. An illustration of a typical power-delay the definition of the delay parameters.

profile and .

IEEE Antennas and Propagation Magazine, Vol. 45, No.3, June 2003

2.3.1 First-Arrival

Delay

(T A)

Bc"'--~'

50rRMS

I

(7)

This is a time delay corresponding to the arrival of the first transmitted signal at the receiver. It is usually measured at the receiver. This delay is set by the minimum possible propagationpath delay from the transmitter to the receiver. It serves as a reference, and all delay measurements are made relative to it. How the origin is defined is seen in Figure 1. Any measured delay longer than this reference delay is called an excess delay.

It is important to note that an exact relationship between the coherence bandwidth and the RMS delay spread does not exist. The real coherence bandwidth depends on the actual impulse response of the channel.

2.5 Types of Fading 2.3.2 Mean Excess Delay

(Te)

This is the first moment of the power-delay profile (as shown in Figure I) with respect to the first delay. It is expressed as (5)

The type of fading experienced by a signal propagating through a mobile radio channel depends on the nature of the transmitted signal, as well as on the characteristics of the channel. Different transmitted signals will undergo different types of fading, according to the relationship among the signal parameters, such as the path loss, the bandwidth (BW), the symbol period, etc., and the channel parameters (such as the RMS delay spread and the Doppler spread). Figure 2 describes the different types of fading and the different relationships that exist among them [5). The phenomenon of large-scale fading is affected primarily by the presence of hills, forests, and buildings between the transmitter and the receiver. The statistics oflarge-scale fading provide a way of computing an estimate of the path loss as a function of distance and other factors. A channel is said to exhibit frequency-selective fading when the delay spread is greater than the symbol period. This condition occurs whenever the received multipath components of a symbol extend beyond the time duration of the symbols. Such multipath dispersion of the signal yields a kind of inter-symbol interference (ISI) called channel-induced lSI. When the delay spread is less than the symbol period, a channel is said to exhibit flat fading, and there is no channel-induced lSI distortion. But there can still be performance degradation, due to the irresolvable phasor components that add up destructively to yield a substantial reduction in signal-to-noise ratio (SNR) at the receiver. Fast fading and slow fading are classified on the basis of how rapidly the transmitted baseband signal changes, compared to the rate of the electrical-parameter changes of the channel. If the channel impulse response changes at a rate much faster than the transmitted signal, the channel may be assumed to be a fast-fading channel. Otherwise, it is assumed to be a slow-fading channel. It is important to note that the velocity of the mobile unit or the velocity of objects using the channel through a baseband signal determines whether a signal undergoes fast fading or slow fading.

**2.3.3 RMS Delay (rRMS)
**

This is the square root of the second central moment of a power-delay profile, as seen in Figure I. It is the standard deviation about the mean excess delay, and is expressed as (6) The RMS delay is a good measure of the multipath spread. It gives an indication of the nature of the inter-symbol interference (lSI). Strong echoes (relative to the shortest path) with long delays contribute significantly to 'RMS' The effects of dispersion on the performance of a digital receiver can be reliably related only to it, independently of the shape of the power-delay profile, so long as it is small compared to the symbol period (1) of the digital modulation. It is also used to give an estimate of the maximum data rate for transmission.

2.3.4 Maximum Excess Delay

[rm)

This is measured with respect to a specific power level, which is characterized as the threshold of the signal. When the signal level is lower than the threshold, it is processed as noise. For example, the maximum excess delay spread can be specified as the excess delay ('I'm) for which P(.) falls below -30 dB with respect to its peak value, as shown in Figure I.

2.6 Adaptive Antennas

An application of antenna arrays has been suggested in recent years for mobile communications systems, to overcome the problems of single-antenna systems. The use of adaptive antenna arrays helps improve the system's performance by increasing channel capacity and spectrum efficiency, extending range coverage, tailoring beam shape, steering multiple beams to track many mobiles, and electronically compensating for the aperture distortion. It also reduces multipath fading, co-channel interference, system complexity and cost, bit-error rate (BER), and outage probability [6]. A phased-array antenna uses an array of simple antennas, and combines the signal induced on the elements to form the output.

2.4 Coherence

Bandwidth

While the delay spread is a natural phenomenon caused by the reflection and scattering of the transmitted signal in a radio channel, the coherence bandwidth, Be, is defined in terms of the RMS delay spread. It is a statistical measure of the range of frequencies over which the channel can be considered "flat." It is defined as the bandwidth over which the variation of the signal is about 10%, and is approximated by [4, 10]

Vol. 45, No.3, June 2003

IEEE Antennes' and Propagation

Magazine,

53

1 Three Basic Propagation Mechanisms Reflection. (8) 54 IEEE Antennas and Propagation Magazine. When = L Ano(t n=1 N -'n) exp] . and from furniture. Multipath Propagation In a typical mobile-radio application. little can be done to eliminate it. respectively. to achieve a better quality of service. corresponding to path n.--------------. the base station is fixed in position.------------------------- Figure 2. If the input signal is a unit impulse. the output will be the channel impulse response. 3. and have sound knowledge of the propagation mechanisms and their influence on the system. 3. The channel impulse response can thus be characterized by N timedelayed impulses. June 2003 . Different types of fading. o(t) . diffraction. while the mobile unit is moving. 3.. and time delays.Large-scale fading due to Motion over la[ge areas Small-scale fading due to Small changes in position Mean signal Attellllation Vs distance Variations about the mean Time-delay domain Fourier transforms Frequency· domain Doppler-shift domain \. because of obstruction of the line-of-sight (LOS) path. No. phases. Details of a phased-array antenna can be obtained in [6.). An' '''' and rp" are the attenuation. which can be written as [9] h(t) 3. while simultaneously placing nulls along undesired directions.1.1 Reflection Reflection occurs when a propagating electromagnetic wave impinges upon an object that has very large dimensions compared to the wavelength of the propagating wave. resulting in a phenomenon known as multipath propagation. The propagation models used for an adaptive antenna are different from those for a single antenna. delay in time of arrival.jrp.7]. Although multipath interference seriously degrades the performance of communication systems. Vol. The radio channel is then obtained as the sum of the contributions from all of the paths. and scattering are the three basic propagation mechanisms [10] that impact propagation in mobile communication systems. However. from walls. the best design for the system can be selected to achieve good propagation performance and.• \ \ \ \ • \ Duals Duals \ . either by reflection or diffraction from buildings and terrain or objects within buildings. Here. Reflection occurs from the surface of the ground. Radio waves therefore arrive at the mobile receiver from different directions with different amplitudes. hence. This is usually subject to a condition such that the propagation between them is largely through scattering. each represented by an attenuated and phaseshifted version of the original transmitted impulse. 45. if we characterize the multipath medium well. They are briefly explained below. The term "adaptive antenna" is used for a phased array when the gains and the phases of the signals induced in the various elements are weighted before combining to adjust the gain of the array in a dynamic fashion along a particular look direction. and phase.

The presence of trees. the angle of incidence. and polarization of the incident wave at the point of diffraction. exhibition centers Open environment. it is also important to estimate the joint angle and delay for various signals.jkd is the propagation phase factor due to the path length d ( k = 2Jf I A. and other obstacles must also be considered. Vol. underground streets. and where the number of obstacles per unit volume is large. or by other irregularities in the channel. and then to picocell. with A representing the wavelength). etc. and are less sensitive to the environment's E receive = LEi' where (9) Ei and = EOPliPriLi(dHlr(¢ji)TI j T(¢pj _r-jkd. Li (d) is the path. r (¢ji) is the reflection coefficient for the jth reflec- Proposed Categories Category Number 1 tion of the ilh multipath component. June 2003 2 3 4 5 6 7 8 . Underground. phase. small objects. Propagation into and inside buildings has. e.depends on the geometry of the object. 3. due to radio-wave propagation. 55 coefficient for the pth transmission of the ith multipath component. diffraction . subways. there is also a great deal of interest in characterizing radio propagation inside buildings. airports. even when a line of sight (LOS) path does not exist between the transmitter and receiver. The terrain profile of a particular area also needs to be taken into account. buildings. reflections from the ground and buildings. require less computational effort. as well as on the amplitude. suburban. the wave may also be partially refracted.like reflection . The direct path.2 Diffraction Diffraction occurs when the radio path between the transmitter and receiver is obstructed by a surface that has sharp edges. each with its own angle of arrival (AOA). and the frequency of the propagating wave. e. 45. When the communication system uses an adaptive-processing methodology. With the advent of personal communication systems (peS).2 Propagation in Outdoor and Indoor Environments With the growth in the capacity of mobile communications. the scattering can be neglected [11]. 3. The indoor radio channel differs from the traditional outdoor mobile radio channel in two aspects: the distance covered is much smaller. giving rise to the bending of waves around the obstacle.3. In most cases. and diffraction from the corners and roofs of buildings are the main contributions to the total field generated at a receiver.3 Summary of Propagation Models In mobile communications. Scattered waves are produced by rough surfaces. For diffraction.and receiving-antenna field radiation patterns along the direction of interest for the ith multi path component. and stairs within buildings can induce scattering in mobile-communication systems. moving cars. and attenuation. a more complex multipath structure than an outdoor propagation environment. street signs. Pi. path delay. and rural areas. e. Classification of buildings. This is mainly because of the nature of the structures used for the buildings. curved Earth to a highly mountainous region. Table 1. is the field strength of the ith multipath component.loss characterization for the ith multipath component. At high frequencies. 3. etc. When propagation is considered in an outdoor environment. d is the path length (m). sports halls. railway stations. most importantly. and Pri are the transmitting.1. The coefficients of reflection and refraction are functions of the material properties of the medium. to some extent. There are two main models for characterizing path loss: empirical (or statistical) models. The terrain profile may vary from a simple. p (10) EO is the amplitude of the reference incident field (Vim). the size of a cell is becoming smaller and smaller: from rnacrocell to microcell.1. the layout of rooms and. and site-specific (or deterministic) models.3 Scattering Scattering occurs when the medium through which the wave propagates consists of objects with dimensions that are small compared to the wavelength. Table 1 represents the categories of buildings where propagation measurements have been made [13].z.reflection occurs. The former are based on the statistical characterization of the received signal. and the variability of the environment is much greater for a much smaller range of transmitter and receiver separation distances [10].z. IEEE Antennas and Propagation Magazine. the type of construction materials used. foliage. T( ¢Pi) is the transmission Description Residential houses in suburban areas Residential houses in urban areas Office buildings in suburban areas Office buildings in urban areas Factorv buildings with heavv machinerv Other factory buildings. one is primarily interested in three types of areas: urban. In practice. A sound knowledge of the physical details of the objects can be used to accurately predict the scattered signal strength. and the complex field received from the various incident paths is given by [12] --> --> 3. the product of the complex reflection and transmission coefficients is replaced by the complex diffraction coefficient. No. The service environments include both outdoor and indoor areas. and generally depend on the wave polarization. and E. lampposts. They are easier to implement. The waves produced by the obstructing surface are present throughout space and even behind the obstacle. signals from the mobile unit arrive at a base station via multiple paths.

respectively.7)hre - -a (hre) + (44." Lmsd is the "multi-screen Lo diffraction loss. and is traveling at 150 km!hr. respectively. There is another parameter that is sometimes seen in the various models: this is the effect of the Doppler frequency on the channel characterization. No.to medium-sized cities.110gfc -0. Lrls is the "roof-top.1. are generally derived from a lot of measurements. and require a vast amount of data regarding geometry. of building and of furniture in buildings.56 log fc -O. d is the distance from the base station to the mobile antenna." The free-space loss is given by IEEE Antennas and Propagation Magazine. it is given by To obtain the path loss in a suburban area. Amu is the median attenuation in the medium relative to free space at frequency J.5510ghte)logd J: doppler = 2 x 150000 X 900 x 10 =250 Hz 3600 X 3 X 108 6 (11) .d) and GAREA can be found from empirical curves.3. (14) 4.9 . The major disadvantage of this model is its slow response to rapid changes in terrain profile. Models of time-delay spread. Okumura et al.3 COST -231-Walfisch-lkegami Model This model utilizes the theoretical Walfisch-Bertoni [16]. respectively. The latter have a certain physical basis. then we know that in this case the Doppler shift will be given by among the simplest and best in terms of accuracy in predicting path loss for early cellular systems. the Doppler shift is relevant to the model. terrain profile.55+ 26. 900 MHz.1610gfc -13. For small.1. new models.1. June 2003 .1.2 Hata Model The Hata model is an empirical formulation [15] of the graphical path-loss data provided by Okumura's model. It would be useful to study why such a small value of the Doppler shift will have an impact on the processing of a modulated 900 MHz signal.1. (17) This model is quite suitable for large-cell mobile systems.1.1 Empirical or Statistical Models for Path Loss 4. where L50 is the median value of the propagation path loss. and here we have included additional. and is composed of three terms [17]: model and hre· are the effective heights where (IS) of the base-station and the mobile antennas (in meters). If a mobile unit is operating at. and has become a standard for system planning in Japan. and are more accurate. Most models of fading apply stochastic processes to describe the distribution of the received signal. (13) where fe is the frequency (in MHz). Both Amu (f. which is a function of the size of the area of coverage.1 Outdoor Case There are a number of empirical or statistical models suitable for both macrocell and microcell scenarios for the outdoor environment. The model can be expressed as LSO (dB) = Lso( urban) -4.8210ghte . GAREA is the gain generated by the environment in which the system is operating. ~e and hre are the effective heights of the basestation and the mobile antennas (in meters).geometry.tostreet diffraction and scattering loss.1. and so on.331og!c -40. say. It is useful to use these models to simulate propagation channels and to estimate the perfonnance of the system in a homogeneous environment.6. 4. but not for personal communications systems that cover a circular area of approximately I krn in radius. a (hre) is the correction factor for the effective antenna height of the mobile unit. the mobile-antenna correction factor is given by a (hre) = (1. both for outdoor and indoor environments. Lp is the free-space propagation loss. The thinking is that since the mobile wireless receiver is moving.98. Model This is one of the most widely used models for propagation in urban areas [141. 45.'s model is considered to be 56 represents the free-space loss.7S(logfcl-1S. For a large city. hie 4.station antenna and the mobile antenna. These deterministic models require more computations. G (hte) and G ( hre) are the gain factors for the base. It is vel)' practical. locations .1 Okumura et al. which varies from 150 MHz to 1500 MHz. Models for Path Loss 4. Some of the propagation models have been discussed in [142]. The formula for the median path loss in urban areas is given by LSO (urban)( dB) = 69.S)dB . and d corresponds to the distance between the base and the mobile unit. the standard Hata formula is modified as follows: (I6) The path loss in open rural areas is expressed through 4. Vol. (1. Some of these are described below.

675(¢-35) 4.7. Lbsh and ka represent the increase of path loss due to a reduced base-station antenna height. windows.2000 (IMT2000) standards activities. ?cO. Vol. Using the abbreviation (24) where hEase is the base-station antenna height.4 + 2010gd + 20 log! . due to 'reflection. which is used commonly when the transmitting antenna is several wavelengths or more above the horizontal ground plane. No. 21]. have been discussed in recent times [27. . we observe that Lbsh and ka are given through _ {-IS log (1 + "'hEase) hEase > hRoo/ . respectively.(L_I) .8 .1. 28. and by kf =-4+1. but it can be affected by the layout in a building. h <h ' o Base . 25].2 Indoor Case Indoor radio propagation is not influenced by the terrain profile. 1 (27a) and kl =-4+0. and doors inside a building. (22) This model is based on a two-ray model [24. the distance between the base station and the receiver.5km and h8ase (26) 54 -1.1. They are given by 18 kd = 18_15MBa. hMobtle' Lori is given by 0° for 35° Lori = I 4.5 Other Models Other models.hBase d e dbrk ' (28) where 11 =:' PL (do).354¢ -55) 2. It is given by [26] I<d <dbrk -10+O. This can be represented as a function of d.1. These models have been developed from measurements. and (21) is the difference between the height of the building on which the base-station antenna is located. The transmitted signal often reaches the receiver through more than one path.4 Dual-Slope Model s ¢ < 35° s ¢ < 55°.6MBased s hRaaf The terms kd and k f control the dependence of the multi-screen diffraction loss as a function of distance. Lmsd is given by where b is the distance between the buildings along the signal path.5km and hEase S hRaaf < 0. and n] and n2 represent the Lbsh - (25) ka'" j 54-0. Recently. June 2003 . the path I~ss in dB at the reference point.9-1010g w+ l Glog j" + 20 log MMohiie + Lori . as is outdoor propagation.114(¢ 55° s¢ s 90° where ¢ is the angle of incidence relative to the direction of the street. and diffraction of the radio wave by objects such as walls. 19J. and use different parameters for different situations. refraction.<e hRaol 4. with values of n2 ranging from two to seven. L(d) .Roof 54 I!J. based on the Uniform Theory of Diffraction (UTD). Some improved solutions for diffraction by multiple absorbing half-planes have also been developed [18.0-0. and the radio frequency of operation. wis the street width (in m). where d is the radio-path length (in km).3. hRoa/' and the height of the mobile antenna.5+0.1. This model is being considered for use by the International Telecommunication Union Radiocommunication Sector (ITU-R) in the International Mobile Telecommunications . 141].fis MHz). If the transmitter and receiver antenna heights are known. 925 ' (27b) 4. slopes of the best-fit line before and after the breakpoint. (19) the radio frequency (in for metropolitan centers. The "point" where this transition occurs is often. It is suitable for the line-of-sight (LOS) propagation regions. then the path loss can be computed. and Lrts = -16. The performance of the different methods in estimating the multiple-diffraction loss term and the final diffraction loss term is given in[22]. 57 for medium-sized cities and suburban centers with' moderate tree densities. 45. do. Lb is a basic transmission-loss parameter that depends on frequency and the antenna heights. The propagation loss. based on the two parameters n] and n2' It is very reasonable to let n] = 2 for the region prior to the Fresnel breakpoint There is much more variability in the path loss and the exponent for the region beyond the Fresnel breakpoint. Other solutions. dbrk represents the breakpoint or the turning-point distance. (20) Here.1.5(L_1) 925 ' (27c) IEEE Antennas and Propagation Magazine. a correction to the COST-231-Walfisch-Ikegami Model has also been reported [23]. especially ifthere are various building materials.Lo = 32. including the use of wideband measurements for different situations. in this case is described by' a dual-slope model. hBase> d d hRaaf . along with the distance between them. called the Fresnel breakpoint. are also available [20.

When electromagnetic radiation is incident on a wall or a floor in an oblique 58 . also varies with frequency [33].7 Through 3 floors 24. hallway and are in sight of each other.8 1. respectively.2 9. this method has been applied to.2 8. and is dependent on the building materials used in a particular environment [34]. p=1 (29) where FAF(q) and WAF{p) are the floor and the wall attenua- 4. When the base station is out of the building. Unfortunately.2 The distance/power model is the main propagation model for path loss. as described in Equation (29). Reference [41] modifies the WAF (p ) term to WAF(p)jcos¢p where WAF(p) and the FAF(q) term to FAF(q)jcoS¢q. the propagation characteristics of electromagnetic waves could be computed exactly by solving Maxwell's equations.1 2. provided that the relative refractive index of the medium is constant. No. this approach requires very complex mathematical operations and requires considerable computing power.4-3.0 9.6 Location PL (dB) 7. buildings [29. 4.6 14. and are widely used when the accuracy of the data is not a critical requirement.4 2. Unlike statistical models.Table 2.9 Through 2 floors 18.1 Ray-Tracing Technique Ray tracing is a technique based on Geometrical Optics (GO) that can easily be applied as an approximate method for estimating the levels of high-frequency electromagnetic fields. resulting in [35. site-specific propagation models do not rely on extensive measurements. 45. As shown in Equation (3). r r In order to take into account the attenuation due to walls and floors. In reference [44].41] PL(d)=PL(do)+IOnlog(d/do)+ I.4 Through 4 floors 27.2 Through 2 floors 27. the path loss in the building was given in reference [43]. GO assumes that energy can be considered to be radiated through infinitesimally small tubes. Table 2 [29] shows the' parameters of Equation (3).1 2. They lie along the direction of propagation and travel in straight lines.0 2.4 7. and they provide accurate predictions of the signal propagation.7 7. many researchers estimate the rate of decay of a transmitter signal through this relation [29-34]. These rays are norma) to the surface of equal signal power.3. often called rays.7 - 914 914 1300 910 1300 1300 1300 1300 915 1900 915 1900 850 Table 3.5 2. Table 3 lists the FAF values for two It is has been observed that the propagation path loss as a function of the distance also has two distinct regions for indoor environments [42J. By using the IEEE Antennas and Propagation Magazine. simplified environments. In an enclosed environment. [35J [35] [36] [371 [381 [381 [381 [38] [39] [391 [391 [39] [401 3.FAF(q)+ q=l £WAF(p). respectively. The empirical or statistical models described in this section are simple to implement.5 to 1.8 2.2 Site-Specific Models for Path Loss Site-specific propagation models.8 when the transmitter and receiver are placed in the same . ranges from three to four [30-33].2 12.5 Through 3 floors 31. In theory. .2.0 2. two additional terms are added to Equation (3).9 5.2 1. A diffraction term was also added to the formula in [41]. and ¢p and ¢q are the angles of incidence of the signal on the walls and floors. signal propagation can be modeled via ray propagation. June 2003 tion factors. but on knowledge of greater detail of the environment. the value of r in Equation (3) may be 1.6 2. also called deterministic models.8 3. less power will be transmitted through the wall than would occur at normal incidence. Building Grocery store Retail store Open-plan Factories Open-plan Factories Open-plan Factories B Open-plan Factories B Open-plan Factories C Open-plan Factories C Suburban office building open plan Suburban office building ooen plan Suburban office building soft partition Suburban office building soft partition Suburban office building hard oartition r 1. 1. 30J. Vol.2 2.3 2.1 14. References [5] and [32] provide reviews of early work on these models.8 3 (dB) 5.7 factor (FAF). The path loss measured in different buildings. are based on the theory of electromagnetic-wave propagation. and FAF(q) are the values of the attenuation factors at normal incidence. The average floor attenuation FAF (dB) Office Building 1 Through I floor 12.0 Office Building 2 Through 1 floor 16.7 9. determined by measurements for different buildings [35-40].4 2.9 PL Frequency (MHz) Ref. fashion. Therefore.7 4. When the receiver is located within a room off the hallway.

No. In references [53-55]. only reflections from walls. Two types of ray-tracing methods .. are taken into account. The image method is efficient. corresponding to two reflections [48]. Ground reflections and rays over rooftops are neglected [58]. and diffraction from comers in buildings. one-reflection) images of a source. and the conventional image method is not adequate. . An image is formed behind the cube representing a structure. then a family of diffracting rays is generated. 55. 58] and the brute-force ray-tracing method are generally used. It is most accurate when the point of observation is many wavelengths away from the nearest scatterer. The concept of a lit region has been introduced in [58] and is illustrated in Figure 4. especially in the zero-field regions predicted by GO. SBA • ~ . refraction. Thus. known as the diffracted rays. 46-72. and 59 . If a ray hits a wedge. has been used to deal with radio-wave propagation in furnished rooms. In GO. If a ray hits an object. the accuracy of the model. June 2003 A finite sample of the possible directions of the propagation from the transmitter is chosen.2. then a reflecting ray and a refracting ray are generated. building building Figure 4. only direct. abrupt transition areas may occur. If the radius is too small. A reception sphere with the correct radius can describe a region that will receive exactly one ray. N (N -1) (N -1) three-reflection images. The Geometrical Theory of Diffraction (GTD) and its uniform extension. The reflection by a wall in 3 building is modeled by an image source placed behind it and a lit region in front of lt. This method requires more computing power than the image method.1 Image Method This method generates the images of a source at all planes. the Uniform GTD (UTD) [45]. These are now explained. IEEE Antennas and Propagation Magazine. the energy reaches the destination through multiple reflections and contributes to the received power. N (N -1) 4. I ~"II SB . A and B. The Fermat principle and the principle of the local field are two basic concepts extensively used by the ray models. The ray-tracing method is widely used in propagation-model and system design [12. The purpose of these rays is to remove the field discontinuities and to introduce proper field corrections. and the same specular ray might be counted twice. a modified shooting-and-bouncing-ray technique.1. This region where the image is formed is termed the unlit domain. and diffraction when hitting a surface.2. The effects of diffraction have also been considered..2 Brute-Force Ray·Tracing Method This method considers a bundle of transmitted rays that may or may not reach the receiver. 45. If the radius is too large. 3. and refracted rays are considered. while the principle of the local field states that the high-frequency rays produce reflection. " . 48. combined with the image method. If there are N reflecting planes. The number of rays considered and the distance from the transmitter to the receiver location determines the available spatial resolution and. two-reflection images. The Fermat principle states that a ray follows the shortest path from a source point to a field point. Consequently.the image method [12. and the interaction of the rays can be described using the well-known theory of refraction and reflection and interactions with the neighboring environment. Vol. Once a ray has been traced through all its reflections to the source. Images due to a source placed between two mirrors. two rays could be received. 42. To determine whether an image of the source is visible at the destination is to trace the intersection of the reflected ray at all the necessary planes of interest. All scatterers are assumed to be large when compared to a wavelength. This depends only on the electrical and geometrical properties of the scatterer in the immediate neighborhood of the point of interaction. These images then serve as secondary sources for the subsequent points of reflections. the attenuations associated with all the reflection terms are calculated. .concept of ray tracing. rays can be launched from a transmitter location. Figure 3 shows a source and its images. I J I I I \' SAB Figure 3.e. Many environments with which we are concerned in our daily life are complicated. A ray is launched for each such direction. reflected.1. 145]. it is possible that none of the rays will reach the reception sphere. then there are N first-order (i. . and so on [48]. corresponding to the boundaries of tbe regions where these rays exist. complement the GO theory by introducing a new type of rays. 4. hence. For the twodimensional case. but it can only handle simple environments. A threshold must be set with respect to the number and order of reflection and diffraction rays that can be considered.

from the receiver to the ray is computed. the ray is considered to contribute to the received signal. Three-dimensional ray-tracing model [12. Otherwise. To keep all the ray-manipulation routines general. Antenna patterns are incorporated to include the effects of antenna beamwidth in hoth azimuth and elevation. it will provide high accuracy and it will take a lot of time to compute. d. the ground reflections appeared to be less important.-. if the angle if = 1 then there will be 360 rays ~obe traced. and the diffraction point is processed as a source. 64-72]. the other is a three-dimensional (3D) method. the inputs are: a) the two-dimensional geometry described by means of vectors specifying the location of the building walls. No rays due to a single ground reflection from the transmitter to the receiver existed in the shadow regions.. the two-dimensional ray-tracing algorithm is quite accurate when the transmitting and receiving antenna heights are significantly below the rooftops of the surrounding buildings. Rays are launched from the transmitter at an elevation angle e and with an azimuth angle if. the ray is treated as not having reached the receiver location. \ \ \ ray~to:.-. and that each wavefront have an identical shape and size at a distance r from the transmitter. This model of propagation between a transmitter and receiver located close to the ground is usually called the canyon model. The transmitter and receiver are modeled as point sources when using this ray-tracing technique.n a . and a large number of rays must be launched. b) the estimated electrical characteristics of the building walls (the permittivity and conductivity. or until the ray is received. c) the base-station location. for LOS regions. The incident ray is decomposed into an object-reflected ray arid an object-penetrated ray. a Two~dih1ensional ray-tracing model [42]. However. VOl. When using a two-dimensional model. Rays generated from a source in two dimensions. At the' source. There is no reception sphere associated with the usage of the image method. as defined in the usual coordinate system. This procedure is continued until the rays are weaker than an assumed threshold. An intersection with a surface of an object is represented by a node in the tree. 3. A two-dimensional diffraction model was introduced in [62. Here. Reception Radius $112 TX Reception Radius too large Figure 5. the received power could have a range (R) dependence given by power law of 1/ R2 . There are two kinds of methods to obtain the rays at the source point.~-.. If d is greater than or equal to (¢L )/2 for the two-dimensional case.-. it is desirable that each ray tube occupy the same solid angle. all the rays or ray tubes are ray sectors. ¢ is the angle between two rays. or the scalar reflection coefficient). How the angle if is chosen depends on the accuracy required and the computation time. Additionally. and can be traced through a binary tree. dn. As reported in [64].--. Figure 5 shows the proper size of the reception sphere that may receive a ray. Both rays then propagate to the next intersection. and that the ray that penetrates the object keeps the original direction of the incident ray. June 2003 the specular ray will be excluded [42. 'One is two-dimensional (2D) approach. In order to determine all possible rays that may leave the 'transmitter and arrive at the receiver in three dimensions.-. ' . The reception tracing. rays are launched along different direc60 . An intersection with a wedge is also represented by a node. 59]. the perpendicular distance.-. sphere Reception Radius too small for a two-dimensional ray Each ray is launched from the source.-. For ranges less than I krn from the transmitter (primary source).-. it is necessary to consider all possible angles of departure and arrival at the transmitter and receiver.-. 45. together with the total (unfolded) ray-path length. \\ ray n+l \ \. 0 . The key part of the ray-tracing method is the generation and description of the rays. The strength of the field at the receiver is then calculated according to Equation (9). as shown in Figure 6. if. the weak contribution from the signals from the over-rooftop rays could be neglected.-. For example. ray n-l Figure 6..-. The authors considered the various buildings to be vertical knife edges.-. neglecting the over-rooftop diffraction and ground reflection. d) the antenna patterns. or until the ray leaves a predefined propagation area. these wavefronts must be such that they can be subdivided. or (ifL)/ J3 for the three-dimensional case. and e) the 'frequency of operation.--. In two dimensions. No. \\ \ \ \ \ \ rayn \ . It is assumed that the reflected ray propagates along the specular direction (the incidence angle equals the aspect angle). in a plane. For each receiver location. from the source to the perpendicular projection point. 64]. Because the buildings were much higher than the basestation (BS) and mobile-station eMS) antennas in an urban microcellular environment. If the angle is small. L. The decomposition process is repeated as a recursion process.tions with the same sector angle. so that an increased ray resolution can IEEE Antennas and Propagation Magazine.

As a numerical method. To simplify this problem. It is well known that for outdoor propagation prediction. especially in non-LOS regions. A hybrid technique [76]. The generation of a large number of diffracted rays limits the number of diffractions that can be considered. they cover the surface of interest without gaps. where the building walls are nearly always vertical planar polygons. A combination of these two models is also possible. diffraction from edges must be accounted for in addition to specular reflections. metal. concrete. In particular. or even glass. a ray-tracing algorithm provides a relatively simple solution for radio propagation. the Finite-Difference Time-Domain (FDTD) method is an alternative. let r = I. Neglecting the differences among the reflectivities of the various materials will degrade the prediction accuracy of the propagation model. A simple approach for introducing the correct spherical-wave spreading has been developed. For complex lossy structures with finite dimensions. Such problems can be solved by numerical solution of Maxwell's equations. where the object database was held in two dimensions. The three-dimensional rays were produced by combining the results of two two-dimensional ray tracers. Some sectors of the walls in a corridor can be made of different materials. patches of different dielectric constants and physical sizes were introduced in [65]. Hexagonal [12] and triangular [53. a two-and-one-half dimension (2. or where there are paths that cannot be taken into IEEE Antennas and Propagation Magazine. for example. where raybased solutions are not sufficiently accurate. 4. we choose at most two. since it depends on experimental data as well as on the propagation model. many researchers have developed a large number of methods [68. 4. A comparison between a two-dimensional model and a threedimensional model was made in [64]. and approximates diffraction along horizontal edges by restricting the diffracted rays to lie in the plane of incidence or in the plane of reflection. the permittivity of this effective material is not easy to determine. since a single incident ray encountering an edge will generate a whole family of new rays. A comparison with FDTD predictions could be used to evaluate and refine the GTD-based methods. and the results were compared with simulated data. In order to find the contributing rays in an urban environment. To improve the efficiency of ray-tracing models. 54] ray wavefronts have also been used. transmitting and receiving antennas are often inevitably installed close to structures with complex material properties for which no asymptotic solutions are available. a two-dimensional FDTD is generally applied [73]. but a ray-tracing engine operated in three dimensions. The basic idea was to use ray tracing to analyze wide areas. In a simple outdoor environment. which can give signal-coverage information throughout a given area. Vol. It can also be used for rooftop antennas. and that it simultaneously provides a complete solution for all the points in the map. and for areas where buildings of different heights exist. it is well known that GO provides good results for electrically large objects. ray-tracing fails to correctly predict the scattered fields." so that all are of the same size and shape and so that.he easily handled. The advantages of the FDTD method are its accuracy. In a complicated communication environment. by significantly enhancing the concept of "illumination zones. June 2003 61 . The vertical-plane-launch approach can treat many multiple forward diffractions at horizontal edges. in [l2] the concept of an "effective building material" was proposed to represent the complicated physical constitutive materials used in the walls of a building. However. diffraction is very time consuming to model. In order to introduce arbitrarily shaped antennas into the simulations. No. Unfortunately. 45. and let the total wavefront be the surface of a unit sphere.72].3. However. for more accurate modeling of radio-wave propagation was also suggested. the FDTD requires large amounts of computer memory to keep track of the solution at all locations. For any given path. hut more computational time is needed. The key to a propagation model based on ray-tracing is to find a computationally fast way to determine the dominant ray paths so as to provide accurate path-loss predictions. Moreover. while the Method of Moments (MoM) model is better when higher precision is required and when the size of the buildings is smaller. and UTD is rigorous only for perfectly conducting wedges. A reduced formulation for the standard FDTD technique [74] . one on a honzontal plane and the other on a vertical plane.requiring four scalar field components instead of the usual six was used to predict channel statistics inside a residential building.5D) or a multi-mode FDTD method was established for indoor radio-propagation calculations [75].3 Moment-Method Models Ray-tracing models can be used with sufficient precision to predict radio coverage for large buildings having a large number of walls between the transmitter and the receiver. collectively. and extensive calculations to update the solution at successive instants of time. Therefore. The problem then becomes one of subdividing the sphere's surface into areas of equal "patches. based on combining a ray-tracing method with an FDTD method. The application of an accurate numerical-analysis method to model an entire area is neither practical (because of the computational resources required) nor is it necessary for open areas without many objects.2. It is noted that the size and dielectric constant of each patch were chosen according to the physical dimensions and the material of which it was made. In [68]. a vertical-plane-launch (VPL) method has been developed [66]. For cases where a lot of small but dominant obstacles are present. using the advantages of each of them.2 FOTO Models Based on Geometrical Optics (GO) and usually supplemented by UTD. Measurements were also conducted." the performance of the algorithm could be dramatically improved. unless an approximation can be made to find the important contributing rays. For example.2. which may have different reflectivities for the incident wave. and the FDTD to study areas close to a structure with complex material properties. a hybrid technique was presented. The procedure of ray tracing in three dimensions is similar to a two-dimensional model. The vertical-plane-launch approach accounts for specular reflections from vertical surfaces and for diffraction at the vertical edges. wood.

1 Vector Parabollc-Equatlon Model This model has been applied to the modeling of radiowave propagation in an urban environment. The use of artificial neural networks (ANNs) has shown very good performance in solving problems with mild nonlinearities on a set of noisy data. 45.(IE-) based techniques. the MoM can be used to check and verify a ray-tracing program. corridor. was presented in [79]. It has also been successfully applied to more hilly terrain.account by a ray-tracing model.2. based on the Moment Method. No distinction was made between the concrete and the brick parts of the walls.corridor. Those excited ray tubes were continuously traced to determine their contributions to the receiving antennas. The ANN model [81] that had the form of a multilayer perception was generally used with 12 inputs and one output. Some of these are described in the following subparagraphs. and also scattering due to periodic structures in buildings [80]. As a parabolic version of Maxwell's equations. where good agreement was achieved. Such an approximation introduces a best-nonlinear-approximation capability into the prediction model. based on adaptive learning from measurement data. In this case. 4. No. the output nodes of which form a combination of radial activation functions computed by the hidden-layer nodes. and the aluminum frame was modeled as a perfectly conducting material. The PMM was applied to evaluate the specular and grating transmission and reflection coefficients of the periodic structures. combining the ray-tracing method and the Periodic Moment Method (PMM) for material objects. Due to limitations of computer memory and CPU time. The glass plates were also assumed to be homogeneous. In addition. it however suffers from the drawbacks of slow convergence and unpredictable solutions during learning. A novel simulation approach. penetrations. there was an input for each defined environmental category. many new methods have been introduced to predict propagation for mobile communications. Another key feature of the neural network is the intrinsic parallelism. based on topographical and morphographical data.3. and scattering by a group of buildings with sloping roofs. the w~lls were modeled as two long dielectric slabs. The REF neural network is a two-layer localized receptive field network. such as wall. 62 4. The network used was similar to that described in [81J. The process of learning may have lasted for a couple of hours. and so on) of that category along the straight line drawn from the transmitter to the receiver. laboratory. 138. To overcome these difficulties.2. An improved version of the "shifting function" has been introduced. as the data obtained from measurements is always noisy. by choosing structures with dimensions around a few wavelengths.5. 139] of three-dimensional electromagnetic scattering. Okumura's data are often included to demonstrate the effectiveness of the RBF neuralnetwork approach. up to an arbitrary resolution. The accuracy of such a prediction model depends significantly on the accuracy of the environmental databases. June 2003 . long enough so that any diffraction or reflection at the outer edges would not influence the results. which can improve the IEEE Antennas and Propagation Magazine. theoretical investigations into the suitability of a neural-network simulator for the prediction of field strength. and compared with published experimental results in the 900 MHz band. All particular locations were classified into 11 distinct categories. The technique is improved by incorporating the Green's-function perturbation method. In [82]. a two-dimensional floor plan was used for a database with a resolution of 10 x 10 ern. The transmission of a UHF wave through a window in a wall is critical when integrating systems that include both indoor and outdoor areas. 4. and the distances between them. Examples include scattering by a building with a hemispherical roof.the real-time learning capabilities and to achieve rapid convergence [83].2. 78]. Effective input and output data processing was developed using deterministic and heuristic formulas for the training of a neural-network simulator. It is particularly useful for accurate modeling of scattering by a single building or by a group of buildings at microwave frequencies. The inputs were the frequency.2 Fast Far-Field Approximation Model This model is substantially faster than conventional integralequation. In the simulations. in order to accurately estimate the propagation loss over an arbitrary environment.5. was developed to study indoor wave propagation. and to surfaces with added buildings [85]. Val. A hybrid approach. outdoor area. 4. This case corresponds to a problem of field-level prediction. the heights of the antenna for the base and mobile stations.5 Other Models Recently. the MoM model can be used [77. which is not possible with scalar versions of the algorithm. The method has been applied to gently undulating terrains. However. allowing for fast evaluation of the solutions. Although the multilayer neural network [82] is a useful method for approximating the propagation loss. while the site-specific models lack computational efficiency. but the process for field-level prediction was fast. Those data were then used in a ray-tracing program to find the specular and grating rays for each ray tube illuminating one of the periodic structures. The solutions determined by the MoM are numerically exact as long as the spatial segmentation used for the objects is small enough. Appropriate centers and connection weights in the REF network lead to a network that is capable of forming the best approximation to any continuous nonlinear mapping. radial-basis function (REF) neural networks that have a "linear-in-theparameters" representation were proposed to enhance. or an appropriate percentage (wall. The resulting simulations were then compared to a set of measurements. It was developed to predict the propagation in an indoor environment. respectively. it allows a full treatment [84. the MoM is usually applied for analyzing objects that are tens of wavelengths in size.2. The output was the field strength. A two-dimensional problem that included stair-shaped walls above a lossy ground was simulated by using both the MoM and the ray-tracing method [77]. One input of the network represented the normalized distance from the transmitter to the receiver. Other inputs represented either a normalized number of occurrences (doors and windows). and so on.4 Artificial Neural-Network Models The main problem with the statistical models is usually the accuracy. were presented.

the FDTD method is used to calculate the effects of walls in an indoor wavepropagation environment. First. where some of the parameters used were determined empirically from measurements. Vol. The second approach used site-specific methods. Details of Environment No No No No Accuracy Good Good Good Good Time Little Little Little Little Other Graphical path-loss data Experimental Data Based on experiments No No No RayTracing FDTD Complex Complex No No Yes Every detail Very Good Best Very Much Very Much Very Much Little Often combined with ray tracing MoM ANN Indoor (small) Outdoor and indoor .5. an improved version of a ray-tracing method is presented. Two-dimensional simulations were performed. 4.4 Boltzmann Model This model was initially developed for simulated fluid flows.3 Summary of Models for Path Loss A summary of the various propagation models dealing with path loss has been given. A comparison Model Name Okumura Model HataModel COST-231 Dual-Slope Suitable Environment Macrocell Macrocell (early cellular) Microcell (outdoor) Microcell and picocell (LOS region) Outdoor and indoor Indoor (small) Complexity Simple Simple Simple Simple of various models for path loss. pie and easy to implement. The issues of profile truncation and small-scale-roughness effects have been addressed. such as scattering from a wedge. starting from a city map. Assuming the screens and slits are distributed by a Poisson law. June 2003 . but the predictions are not very accurate. Algorithms for the path-loss prediction were presented and compared with experimental data in the references cited above. 4.4 Efficient Computational Methods For Propagation Prediction for Indoor Wireless Communication This section presents two different deterministic methods for efficient characterization of an indoor channel.2. Complex Complex No Yes Every detail Detail Best Very Good Takes time to learn from experimental data 63 IEEE Antennas and Propagation Magazine. there is a need to develop efficient algorithms for the computation and mapping of the field distribution in such structures. Ray tracing was the main method. Theoretical analysis of propagation in a city street modeled as a three-dimensional multislit waveguide was proposed. a simple empirical or statistical model of the path loss was considered. In this case. No. On the other hand. other than in the choice of the parameters). and a renormalization scheme was proposed. Each of these two kinds of approaches makes a very different trade-off of accuracy versus complexity. First. Propagation models dealing with path loss for mobile communication have been emphasized. Some other numerical methods used in electromagnetic-field computation have also been applied. and the numerical results presented showed excellent agreement with published measured data. 140]. the locations or all the objects. The method. 4. the transmitting and receiving antennas are both located below the rooftops. site-specific models are considerably more accurate than the empirical models. but they require a great deal of specific information about the area of interest (at a minimum. Table 4. A brief comparison of some of the main models is presented in Table 4. 143]. which is sirn. 87. provided good path-loss predictions when compared with site measurements. 4. Next. The empirical (statistical) models are extremely simple (no environmental information is used. using two very different approaches. and possibly the locations oflarge objects).3 Waveguide Model In large metropolitan areas that have tall buildings. the statistical propagation characteristics in such a waveguide are expressed in terms of multiple ray fields approaching the observer. This technique can take into account complicated boundary conditions. and the city streets act as a type of wave-guiding structure for the propagating signal [86.5.2.performance of the technique for more-challenging problems. It describes a physical system in terms of the motion of fictitious microscopic particles on a lattice [88. 45.3.

All the propagation phenomena. The air hole in the block forms a periodicity in the blocks.. and diffraction.4. and it took only 10 minutes to compute the path loss for 250 sample points. 65.. June 2003 . 126].2 Analysis of the Effects of Walls on Indoor Wave Propagation Using FDTD The ray-tracing technique does not take the effects of the inner structure of walls into account for indoor propagation prediction. A numerical approach to treat this problem. transmission. A typical periodic inner structure of walls.1 Efficient Ray-Tracing Methods The application of several ray-tracing techniques in combination with the Uniform Theory of Diffraction has been an efficient method for the prediction of propagation in the UHF (communication) band in an indoor environment. Simulation results using three different methods: . 4. IEEE Antennes and Propagation Magazine. The periodic structure of walls was also considered. 135].FDTD (periodic). This indicates that the effects of the inner structure of the walls cannot be neglected for propagation prediction.Ray tracing (uniform). As a full-wave analysis tool. A two-dimensional ray-tracing method was used to analyze the uniform structure. reflection and refraction by layered materials and diffraction from the comers of walls was considered. a new three-dimensional propagation-prediction model was developed that can save 99% of the computational time of a traditional three-dimensional model. In addition.4. When using a ray-tracing method.1 Numerical Results The fields of the environment were calculated using a twodimensional FDTD method when the wall was first assumed to be a periodic structure.2. using the FDTD method. It has been found that non-specular reflection occurs when the operating frequency is above 1.4. material properties of walls. but few have paid attention to the inner structure of the walls [134. but they did not consider non-specular reflection. Numerical results for the path loss were calculated by an FDTD method. -30 -35 4. FDTD Vol.3. 4. FDTD (uniform). The maximum difference in the field strength when the wave was assumed to propagate through a periodic structure and through a uniform structure was more than 10 dB when using the FDTD method. . by making use of the two-dimensional ray-tracing results. This theoretical model had a good correlation with experimental data. and were compared to those obtained by the ray-tracing method.method of homogenization can be used to determine the effective _. The FDTD method was used to predict propagation properties for indoor environments [137]. 45. and that the FDTD method can give more accurate results. Numerical results for the path loss calculated by an FDTD method were compared with those obtained by the ray-tracing technique. In this approach.2 GHz [134J. Figure 8 shows the simulation results of the path loss for the sample points. This is discussed in detail [136J. the reflection and transmission characteristics of the walls can be derived by solving for higher-order Floquet modes. In this model. is now described. where the computational efficiency of the two-dimensional ray-tracing method was improved by reorganizing the objects in an indoor environment into irregular cells. and then uniform in nature. For the FDTD method.. with an effective dielectric constant. The application of ray-tracing techniques in propagation prediction has been under development for more than a decade. the FDTD is more accurate than the ray-tracing method. This new hybrid model is more accurate than two-dimensional models and more efficient than traditional three-dimensional models in computing the path loss to any point in a building. No. There have been a number of recent investigations into indoor radio-propagation modeling using the ray-tracing method [57. are included. such as reflection. A raytracing method considers only reflection and transmission. 5000 time steps were calculated. Also. It was proved that the inner structure of the walls had a considerable influence on the path loss when predicting propagation. it is assumed that reflection from walls has a substantial specular component Typical concrete blocks used in wall construction are shown in Figure 7. The maximum difference in the field strength when propagating Figure 7. They are powerful and easy to use. and it took more than six hours of CPU time on a Pc. the 64 -40 -45 -50 ·55 ~~~-L~~~~~~~~~~-L~ 20 40 60 80 100 120 140 160 180 200 220 240 o SampJe Points Figure 8.

Models For Small-Scale Fading Small-scale fading refers to the dramatic changes in the signal amplitude and phase that can be experienced as a result of small changes (as small as ~.half wavelength) in the spatial separation between a receiver and a transmitter. we have to be extremely careful when using a statistical model. p(r) j 0' r exp [r2 -~ + A2 J (Ar) 10 .~-30 r/. A typical Rayleigh fading envelope for a movi~g mobile at 900 MHz is shown in Figure 9 [90].2 20' J2) ' (34) K(dB)=JOlog-2• (31) K is known as the Ricean factor. or of an individual multipath component.° 0'2 A:?:O.. the Rayleigh distribution is widely used to describe the statistical timevarying nafure of the received envelope of a flat-fading signal. For mobile-radio channels. and 10 ( . The log-normal PDF can be expressed as the modified Bessel function of the first kind and zero order. (32) o The probability that the envelope of the received signal does not exceed a specified value R is given by the corresponding cumulative distribution function (CDF): . The Ricean distribution is often described in terms of a parameter. the composite signal resembles a noise signal that has an envelope of a Rayleigh distribution. and 20'2 is the predicted mean power of the multi path signal. A few possible choices of the statistical distributions for modeling the envelope are explained below. It is possible to estimate the Ricean K factor of a signal from measurements of received power versus time. j _::_exp [0'2 LJ 2a2 o 0 s r $ 00 r<O . the fading distribution is Ricean. . the method described in [89] yields an explicit and quite simple expression for K in terms of the measured moments.3 Log-Normal Fading Model This model is used to quantify the distribution of rays that experience multiple reflections and diffractions between a transmitter and a receiver. However. The Ricean distribution is given by where r is the envelope amplitude of the received signal. A typicai ~yleigh fading envelope at 900 MHz. June 2003 .\. A simple and rapid method has been developed. The median value is often used rather than the mean values. One approach is to compute the distributions of the measured data and to then compare the result to a set of hypothesis distributions. These dramatic changes in the envelope of the received signal are described statistically by a stochastic process.r:?:O (30) r<O where r is the amplitude of the envelope of the received signal. The Rayleigh distribution has a probability density function (PDF) given by 5. R !p(r)dr=l-exp ( --2 R2 P(R)"'?. No. It is given by [10] A2 20' p(r)= r ~exp 2rr0'2 I 1 [In(r)-m . The method is exact when perfect moment estimates of the Ricean envelopes are available.J In 11 " V I J ~. 5. VOl. By contrast.. it is important to study the distribution of the envelope of the received signal. and completely specifies the distribution. As explained and emphasized in [160]. using suitable goodness-of-fit test. 65 1£££ Antennas and Propagation Magazine. When there is a dominant stationary (non-fading) signal component present. and it is easy to compare different fading distributions. received by a mobile unit traveling at 120 kmlhr. 10 5 ~0 -5 ~ g 1\. which may have widely varying means. as a complete lack of knowledge of the system cannot be supplemented by a stochastic distribution. a ~-1O '-' -15 ] -20 til -25 .(r<R)= 2a J ' (33) 5.l Y I' n 'I 11 IINI . 45. which is defined as the ratio between the deterministic signal power and the variance of the multipath. and 20'2 is the predicted mean power of the multipath signal. Since the fading data are usually measured in terms of the fields. based on calculating the first and second moments of the timeseries data. V' ~N' I ~ n/I. K.2 Rayleigh Distribution As the dominant signal in a Rician distribution becomes weaker. quantities for a particular distribution cannot be assumed. In that case. Another is to compute a maXimum-likelihood estimate using an expectation/maximization (EM) algorithm. both of these approaches are relatively cumbersome and time consuming.) is 5. A denotes the peak amplitude of the dominant signal.3. such as a line-of-sight propagation path.1 Ricean Distribution This is also called a Rice distribution or a Rician distribution.through a periodic structure using the FDTD method and the raytracing method was around 18 dB. -35 -40 o 50 100 150 Elapsed Time (ms) 200 250 Figure 9. p(r)= In order to get a sound understanding of the channel. the factor K can only be obtained implicitly by equating a ratio of the measured moments to a complicated function of K.

called "POCA.(t) is substituted for by the product 1](t) (38) = '(t). n = (r2)iS 66 r is the envelope the time-averaged amplitude of the received signal. but rather small (e. It where a is the standard deviation and r is the amplitude. 5. m = (r2rj ((r2 -{r)Zf) is the inverse of the normalized v~ance of r2. was shown that the Weibull distribution characterizes indoor propagation-path losses quite well in [146]. a new theoretical model for the prediction of fast fading in an indoor environment was developed. the process '. Here. averaged over a period of few seconds. 5. but it lacks a theoretical basis. and . especially for indoor environments. This assumption led to the development of a new theoretical model. This is a widely accepted statistical model for the receivedsignal envelope in macrocellular mobile radio channels where there is no direct Los path.5 Nakagami Model The Nakagami model was developed in the early 1940s by Nakagami.7 Other Fading Models Many other models for fading have been developed. Usually. and . and a is the standard deviation of the corresponding normal distribution. This model made the assumption that the number of dominant propagation paths that contribute to the signal in the receiver in a multipath environment is not infinite. and the Kdistribution. m. Techniques such as the Monte Carlo method and the Schwartz and Yeh method have been developed to simulate the power sum for the log-normal components [92]. IEEE Antennas and Propagation Magazine. where the log-normal process '(I) = exp[. The assumption of statistical independence does not always meet the real conditions in multipathwave propagation." that is more general and more efficient than the Rayleigh Model. The received power. Some of these are the Rice Log-Normal Model [98J. It provides a moreaccurate approximation of the sum of correlated complex lognormals for a wider variety of channel behaviors. These are mixtures of two kinds of distributions.6 Weibull Model This model arises when results from mobile-radio propagation measurements are plotted on graph paper that is scaled such that a Rayleigh distribution appears as a straight line with a slope of -1. The requirement of statistical independence between fll (I) and flz (t) is identical to the demand for a symmetrical power spectrum for A(t). No. according to the relation (35) where 5. propagation models that include both the effects of shadowing and multipath fading were developed and used in studying terrestrial and satellite channels. 15).4 Suzuki Model The Suzuki model combines two distributions. For the special m. which is a substitute for the Rayleigh log-normal distribution [101]. The product process with this particular amplitude-density distribution is called the Suzuki process. Details of the implementations of these models can be obtained from the various references cited. The corresponding probability density function is written as [96] (37) Here. A simulation of the Nakagami model has been presented in [97]. power of the received signal. due to various shadowing effects. June 2003 5.3. some of the statistical models for fading have been introduced. with variance and mean b=[(2!a)r(2!a)t2 is a normalization factor. given appropriate bounds on the parameters.. In order to adapt the model to this behavior.) is the Gamma function. 45. and was simulated in [95]. r(. In [102J. so it has been modified [94J.(1) can be regarded as the envelope of one complex- valued normal random process. A (t ). comprised of the log-normal and Rayleigh distributions. The Rayleigh probability density function is obtained for m = I. the Weibull PDF becomes a Rayleigh PDF.'(r). The Weibull distribution provides flexibility to model any shape offered by a Nakagami distribution. ro is the RMS value for r.u3(t)J s2 where a is a shape parameter that is chosen so as to yield a best fit to the measured results. and are now widely used. the Nakagami Log-Normal Model [100]. It is generally fit for environments in which there exists a small number of dominant propagation paths. In [103] and [104].where m is the median value. the Rayleigh distribution is obtained from two statistically independerit normal processes. 2s (36) case in which a = 1/2. The Nakagami PDF may be shown to be a more-general expression of other well-known density functions. and is given by [93] Pll(r)= l~exp(oa 20- r 2J 2 . Only their PDF distributions have been described. parameterized by the fading severity parameter. fll (t) and fl2 (t) . obtained by using the transformation y=ln(r) [91]. with zero means and identical autocorrelation functions..g. The Weibull PDF can be written as [91J '.27rsa ~ exp[ (lna~m)2lda. can val)' considerably. A Nakagami model. In this section. is defined by a normal process fl3(t) . the Nakagam-Rice Model [99J. Vol. The Nakagami PDF can also be approximated by both Rice and lognormal distributions over certain domains. has been shown to fit well to some urban rnultipath propagation data.

the RMS delay spread was analyzed as a function of the path loss and Resolution=Pulsc Width Digital Storage Oscilloscope 6. the RlvIS delay spread becomes constant for all ranges [105]. because of the presence of a larger number of scatters in the obstructed line-of-sight environment than in the LOS environment.lS. the RMS delay spreads do not exceed 13. the RMS time delay in most cases was significantly smaller than at 2. with high data rates and small cell sizes. Vol. sight (aLaS) charinels are subject to higher variations in R1vi:s delay spread.3. In [l08]. In digital wireless communications. 45. both for outdoor and indoor environments. In Table 5. A large number of measurements have been made for the impulse response.75 GHz. The results of measurements. Resolution. Direct pulse measurements. A block diagram of the spread-spectrum impulse measurement system. corresponding to different environments and frequencies. There are many factors affecting the impulse response. for which the values were. Spread-spectrum sliding correlator measurements. as well as modeling fading. channel X(m) Port I Vector NetwOtk Analyzer with Swept Frequeocy Oscillalor s·parameter test set Y(m) Port 2 Figure Iflc. a number of wideband channel-sounding techniques have been developed. B. "Med" indicates median .6. and Gaussian noise is the dominant factor that causes bit errors. In modern mobile-communication systems. such as the frequency. 144. along with some typical models. are described here. it is found that the RMS delay spread varies from several nanoseconds to several microseconds. or indoor furniture). the RMS delay spreads do not exceed 7 or 8 J.5 us.5 GHz. and C. the timedelay spread cannot be neglected. These techniques may be classified as Figure lOa. and that in urban areas with surrounding hills. fixed objects (mountains. A multipath propagation channel is modeled as a linear filter. and "Ave" implies the mean average values. since the output of the system can be computed through convolution of the input with the impulse response. 147-151]. If the symbol rate is much lower than the coherent bandwidth. It was shown in [121] that in many cities where the terrain is flat. Impulse-Response Models The narrowband or continuous-wave (CW) path loss is a parameter that can predict the power level of the system>and the space coverage of a base station. followed by a description of some deterministic models. it was found that at I 1. and the calculated RlvIS time delays have been summarized in Table 5. Some measurements for both indoor and outdoor environments. buildings. if the system is linear. If the symbol rate is relatively high. it is necessary to model the effects of multipath delay. but a different conclusion was obtained in [107]. and has a complex baseband impulse response [105-125. one of the main reasons for the occurrence of bit errors is inter-symbol interference (lSI) caused by multi path propagation. A block diagram of the direct RF channel impulseresponse measurement system. In [26. Some models have been derived based on the measured data.II R. 123]. When there is furniture and mazes of semi-permanent partitions. values. (nns poise width) A. A block diagram impulse measurement. the time-delay spread can be neglected. caused by movement of the terminal around a small area. Obstructed line-ofIEEE Antennas and Propagation Magazine. Swept-frequency measurements. shown in Figure 10 [10]. June 2003 Digital Storage Oscilloscope Widebaod Filter Narrowband Filter Figure lOb. The block diagram for each of these techniques is. and moving humans or objects close to the transmitter and/or the receiver. On the other hand. In this case.1 Models Based on Measurement Results Because of the importance of determining the time-delay spread for wireless communications. The reason that two different conclusions were reached is that all of these various conjectures have been derived from specific measurements. ''T-R'' indicates the distance between the transmitter and receiver. multipath propagation only causes fading of the signal level. In this table. It is concluded that in an empty indoor environment. No.4 GHz and 4. of the frequency-domain channel 67 . The impulse response is a useful characterization of the system. It is shown in Table 5 that the RMS time delay varies little for different frequencies for indoor environments [106]. the height of the transmitting and receiving antennas. the RlvIS delay spread is a function of the transmitter-receiver separation. Almost the same results can be obtained from the last two methods [105]. about the same. the short-time variations in the RMS delay spread in obstructed line-of-sight environments were found to be less than those in LOS environments [117]. in general.

Retail store Laboratories at same floor Outdoor Site 1 Site Z Site 3 and RMS delay spreads. 60-250.) 98-1270 590 (Ave.5d +40 (dis T-R) 12.3.6-325.1700 910 [109] D) E) 1300 4000 [110] A B C B A - B C C In two buildings Four types of indoor locations Hamburg Dusseldorf Frankfurt(bank) Frankfurt( apart.5 m 68 Standard Deviation of RMSDelay (ns) Location Ref. The RMS delay spread as a function of the height of an antenna. 'RMS (ns) Frequency (MHz) A) B) C) Ref.6 173. 45.120 (Med) 10-50 with mean 20-30 10-20 (Med) 10-20 (Med) 5-15 (Med) 1300 (typical) 3100 (typical) 8100 (largest) 19600 (largest) 2500-7500 2000-7000 2500-13500 1000-25500 140.0 127.5 [1221 Table 6.) 14. 136.4 37200 910 [111] [112] 40000 2600 2400 910 [ll7] [l13J [114J [115] [118] 900-1300 2400 4750 1150 942.70 (Med) 60-250.7m 8.60 20.225 (119] [108] [120] 892 [121] 2197. r1061 A A 30-100 15-29 (Area 31-62 (Area 48-90 (Area 52-57 (Area 19-37 (Area 7-120 40-130 15-125 5-40 850.3 m 3.1 O[OSI Med.74-102. Measurements Technique B Location Office building Manufacturing floors Indoor sports arena Open-plan factory Textile plant Office building Within aroom (4 rooms) Urban area I Urban area II Building I Building 2 Sidewalk of LOS street Engineering bldg.) Washington Greenbelt Oakland San Francisco LOS (100-400 m) A A - 7-16 10 ns (Med. Vol.3 LOS (Med.130 (Med) 40-130.7m 8.85-84.) About 12-72 About 4-25 0.0 147.44 8. Mean RMS Delay (ns) Antenna Height 3. June 2003 .1 352.3 OLOS2 Med.8 275.4 156.8 176.1 138.Table 5. 22.9 134.) 61-2940 480 (Ave.8 San Francisco Ottawa (26] (123] IEEE Antennas and Propagation Magazine. No.5 m 13.

5 ~ _ K Model 6. 6.the height of the antenna. rural. 21C]. The measurements indicated that the statistics of the channel impulse response were independent of the polarizations of the transmitting and the receiving antennas if there was no line-ofsight path between them. andy is the log-normal variable. where PL is the path loss in decibels. They may not be suitable for prediction in other environments.2 Statistical Models of Time-Delay Spread 6. 6. Initially. Vol. June 2003 69 .2.) 110 ns < TK < 200 360 (41) 6. and their density functions are imifonnly distributed over [0.08 + 0.8 ns. This model [9. TK. the piecewise functions of excess delay for the probabilities of multipath arrivals were derived as [3] 1TK 367 TIS. and the third conjecture was that 'RMS tended to increase with shadow fading. and mountainous areas.62e 200 ns < TK < SOO ns TK<lOOils • P (T R K. scattering the electromagnetic radiation at greater distances. The impulse response of the model is represented by 0.2 Saleh and Valenzuela Model Saleh and Valenzuela [30] reported the results of indoor measurements between two vertically polarized omnidirectional antennas.3 Deterministic Models of Time-Delay Spread The above statistical models are based on measurements made in a specific environment. It indicates that the delay spread increases as the antenna height increases.2.0.1 Two-Ray Rayleigh Fading Model A commonly used multipath model is an independent Rayleigh fading two-ray model [10]. 119] takes into account the clustering property of paths caused by the grouping property of scatters. 0." Each bin is assumed to contain either one multipath component or no multipath component. In this model [119.3.2. If no further paths arrive in the interval [t. 45.K_-_I_lO-<. Table 6 shows the measurements of 'RMS with relation to the height of the antennas. called "bins. The clusters and components within a cluster formed a Poisson arrival process. TRMS can be given by (40) where 1] is the median value of 'RMS at d = 1km. The process is described by transitions between two states. representing different mean arrival rates.3 Log-Normal at Any Distance In [124]. If a path arrives at time t. in units ofns.2. 6.2. This was derived from measurement data that were classified into urban. A reasonable bin size is the resolution of the specific measurement. a transition is made back to State 1 at the end of the interval. located on the same floor of a medium-sized office building. IEEE Antennas and Propagation Magazine. Cardoso and Mouliness presented three conjectures about delay spread. with different rates. and takes on values that are integer multiples of7. 8 is an exponent that lies between O. the process starts in State 1. because as the antenna is raised._( T"". with a mean arrival rate K Ao (t ). 6.65 _ _. respectively.6 Discrete-Time Model 6. it becomes visible to more objects.SS+TK 667 0. The model assumed that the multipath components arrived in clusters.S and l. No.t+~]. 125]. t is the time delay between the two rays. suburban. a transition is made to State 2. ~ and B2 are also two independent random variables. The first was that 'RMS was log-normal at any distance. Rayleigh PDF.4 SIRCIM Model Based on measurements at 1300 MHz in five factory and other types of buildings. with a mean arrival rate Ao (t ). the time axis is divided into small time intervals. The predicted results can be bounded by an exponential model of the form 'RMS(ns)= exp(0. The possibility of more than one path in a bin is excluded.06SPL). S) 2 - r (TK-lOO) 75 (42) 100 ns < TK < SOO ns where S] and S2 correspond to LOS and obstructed LOS environments. is the excess delay at which where al and a2 are independent random variables and have a a multipath component will arrive at the receiver.2. The second conjecture was that the median 'RMS increased with distance.22_(TK-200} 1360 O.

Second. it is computationally prohibitive. and can deal with the number of resolvable paths that are determined by the number of samples available. in fact. however. the angle corresponding to each delay . 7.3. it will be impossible to achieve agreement with either the instantaneous measured shape or the value of the RMS delay spread for a single individual measurement. There are other techniques in addition to the JADE models. but in an outdoor environment. then the angles and delays can be estimated jointly. this algorithm transforms the data using a discrete Fourier transform (DFT). Since the phase of each ray arriving at a receiver varies significantly with distance and cannot be accurately predicted. such as the subspace-based methods [158]. Hence. can lead to better system service. However.156] 7. and vice-versa [8. Third. 152J.approach is based on the MUSIC algorithm. 7.2 VRP Model This is similar to a seven-ray model.is definitely not optimum.2 JADE-MUSIC Model [154] A faster . The joint angle. the ray-tracing technique (12. It assumed some virtual reflection points (VRPs). in conjunction with the DOA information. exploiting the spatial structure by using adaptive antenna systems and properly estimating the delays can improve the characterization of the channel and. where L is the number of muItipath components. respectively.3. the relative delay estimates between two or more synchronized base stations can be used. The RMS delay spread was predicted. to estimate the location of the mobile unit. IEEE Antennas and Propagation Magazine. 127] can determine almost all multi path components. the multi path channel is characterized not only by the various time delays of the different propagation paths. consequently.and delay-estimation technique is. The predicted results for the RMS delay spreads were verified by measurements. The MUSIC method estimates the angles of arrival and the delay parameters by finding the L minima of the pseudo-spectrum.joint estimation has a number of advantages. 105. It did not consider the effects of traffic and moving humans. Hence.4 JADE-ESPRIT Model [157] If the array is a uniform linear antenna array.1 JADE Maximum-Likelihood Model [8. The JADE-ESPRIT algorithm requires stricter identifiably conditions.which first estimate delays and. The algorithm can thus be written in closed form and is computationally attractive. Another advantage of ray-tracing models over other propagation models are their ability to incorporate antenna radiation patterns and. to consider the effect of the radiation pattern individually on each ray. and the velocity oflight. The Cramer-Rao bound (CRB) has been derived for each of these models. the most-probable portable direction is the one that is associated with the smallest delay. In [153]. 6. In the unconditional-maximum method. and a Bayesian-based model [159]. including their amplitudes. when several angles are detected.structure. Unfortunately.6. The comparison between measured and modeled results should be for average values in a small area around the actual mobile position [127].3 SI-JADE Model [155. 123. time delays. or if it has the ESPRIT doublet. A detailed knowledge of the environment is required when applying a ray-tracing method. the problem with this model even though the formulation is quite esoteric from a signal-processing point of view . the envelope of the fading path is assumed to be a stationary Gaussian random process.3. First. because the direct path has the shortest propagation time.153] In the conditional-maximum method. and agreed well with that calculated from measurements. No. in particular. only acts on a single-channel estimate. and a different cost function is obtained. It then stacks the result into a Hankel matrix. There are several variations of this technique. and phases. In comparison to disjoint techniques .is that the directions of arrival and the time delays of the signals are not independent quantities. but also by their directions of arrival (DOA) or angles of arrival (AOA). Estimating the time delay by ignoring the spatial characteristics of the multipath channel . which is often abbreviated as JADE. and were used to simulate propagation in a room [105]. often employed in radars using space-time concepts. parameters are intimately related through the distance between the source and the field point. subsequently. A seven-ray and a 25-ray model consider only single or double reflections.as most systems now do . even for moderate values of L. Under this assumption. Vol. and performs a deconvolution with the known pulse-shape function. The two 70 By assuming a uniform time sampling and a uniform linear array.1 Ray Tracing In theory. the maximum-likelihood-method estimates of the angle of arrival and the time delay are obtained by minimizing the cost function. the vector representing the complex envelope of the fading path is assumed to be non-random. It involves a 2L-dimensional optimization over both the angles of arrivals and the time delays. 118. 7. Joint Angle-Delay Estimation (JADE) In mobile communications systems. and the technique is effective in predicting the time-delay spread. it is theoretically possible to exploit the difference in path delays to improve angle-estimation accuracy. a basic physical understanding of the problem is absolutely essential in trying to obtain a meaningful solution. This model is quite similar to the SI-JADE algorithm.though suboptimal . 45. SI-JADE. This reduces the problem to one that can be solved using a twodimensional ESPRIT algorithm. and may also introduce an undesired coupling between angle and delay estimates in the case where the model is not perfectly valid. 7. Indeed. and involves only a two-dimensional search. June 2003 . located at the intersection points along the LOS on streets and at building walls [122]. only simulation results have been used to verify all these models. an efficient method in which 9 and T were estimated in parallel was developed.

and integrates or samples the signal on reception. however. we have a voltage source VT1 transmitting at the same frequency. consider the system represented in Figure J I.To conclude this section.. that the transmitter is sending signals using spatial diversity. V~ and n. unless the impulse responses of the transmitting and receiving antennas are included in a broadband channel characterization. it is difficult to see how one can obtain a meaningful physical model of the wireless propagation channel! knowledge about the spatial locations of the transmitters and the receivers. there are two receivers. June 2003 transmit/receive scenario. A multiple-user IEEE Antennas and Propagation Magazine. which is the domain of wireless. The smallest antenna is a Hertzian dipole. we simultaneously employ the concept of reciprocity and use the principle of adaptivity on transmission. On reception. No. However. the open-circuit voltages are proportional to the single derivative of the incident fields. it is not necessary to know these path propagation models in order to direct the energy from the transmitter to a pre-specified mobile. Vol. It is not even necessary to know the particular electromagnetic environment in which they are transmitting. we want to find another set of voltages. for finite-length antennas. using a number of transmitting antennas spatially separated from each other.e..3. A brief description ofthis novel technique is provided here. and even that has a directional pattern. The internal impedance of this voltage source is RZ. it is possible to direct the energy transmitted from a base station to a pre-selected mobile station. In this novel procedure. At the second transmitting antenna. the radiated fields are proportional to the double derivative of the current induced on the structure. it is important to point out that all these models treat each antenna as an isotropic radiator. i. vi" that will excite the two trans- lrJ Trnnsmittex I Transrnitter2 Receiverl Figure 11. Let us assume that transmitting antenna #1 is transmitting a voltage VlJ at a frequency 10. represented by RJ and R2• The goal of this transmitting system is that we want to maximize the received power at receiver I by exciting transmitting systems I and 2 with voltages V~ and Our goal is to minimize the received power at receiver 2 while directing the energy to receiver I. Hence. An admirer of James Clark Maxwell will immediately realize that such an antenna does not exist in the electro magnetics world. It is not even necessary to have any an internal impedance of RT2 • In addition. without worrying about either the presence of other near-field scatterers or the existence of a multipath environment. Epilogue However. For example. for example. Here. we point out that an electrically small antenna differentiates the pulse on transmission. but does exist in the acoustic arena. in a mobile communication environment where there is an established two-way communication. The details are available in [160]. 45. In this way. let us assume. This can be accomplished using the principle of reciprocity. with 8. 71 . Similarly. Finally. it may be omnidirectional in certain planes.

we consider the currents that are induced at the transmitting and receiving antennas to be at the loads that are associated with the feed point of each antenna. I1j at transmitter #1. However. where each receiver carries a unique code. such that VT2 = VR. one obtains (45) It is important to observe that reciprocity only links the two transmitting and receiving ports under consideration. The induced current at receiver 2 under these conditions will be W-/I~ +Wi1f2. so that (44) Similarly. Under the same environment. so that Vlj = 1 and Vr 2 = VR ~ I. The characteristics of these two voltage sets with a superscript of 2 are such that when they are used to excite the transmitting antennas. Reciprocity tells us that if we excite transmitting antenna # I with a voltage V1j and with the other voltages set to zero. if we excite receiver 1 with a voltage VR. This will induce currents and in transmitting antennas 1 and 2. It will also induce currents at receiver #2 and transmitter #2. like trees or buildings that are in the immediate neighborhood of the transmitting or the receiving antennas.without violating the principle of reciprocity . Under this situation. However. IR1.mitring antennas so that they will induce zero fields on receiver #1 (designated by R1J and maximum fields at receiver #2 (designated by R2). The principle of reciprocity can help one to treat such a complex situation without any knowledge of either the electromagnetic environment in which the antennas are communicating. we have the result that if we excite transmitter 2 with I V at the same frequency 10' then VTz = 1 and VIi = VR1 = VR. 45. In this way. and it is not necessary to know the spatial locations of objects that influence the electromagnetic coupling mechanisms.: VR2 "" O. let us also ignore these for the moment. Based on the principle of reciprocity. Then. this is under the assumption that all the loads terminating the two transmitting and the receiving antennas are all equal and numerically equal to Rr. In this way. = 0. June 2003 . then it will induce a current -let us say. as in a CDMA environment. what the reciprocity principle tells us is that if we apply I V at the feed of the transmitting antenna #1. we can spatially direct the energy to a pre-specified receiver only. = VR2 then it is going to induce currents I~ in transmitting antenna #1. as illustrated in Figure II. respectively. No. then the transmitted signal will induce a From the principles of reciprocity in the frequency domain. we now implement the system of adaptivity on transmission. from the principle of reciprocity. That is why it is possible . where a unique code is assigned to each receiver. = I V with VR2 = Vlj = VT. The relationship provided by Equation (43) is valid . respectively. Now. we know that (43) Equation (43) represents the reciprocity theorem in the frequency domain. then this will induce currents = VR 2 =0 . at transmitters I and 2. both spatially and in the temporal domain. R]). = O. Step V1j = 11 with Ih 1 V. It will also induce currents at receiver #2 and transmitter #2. = 0. whereas a receiving antenna integrates the incident signal. and If. as long as it remains the same when we switch the voltages on the transmitters and receivers to satisfy the relationships of Equations (44) and (45). with 2: Excite receiver #2 with I V. . Let us assume that we can measure these two currents. This is ideally suited in a CDMA (code-division multiple access) environment.. The measurement is easy to carry out in a CDMA environment. = O. 72 VIi = VTz . and simultaneously minimize the received signals at the other receivers for which that particular transmission was not intended. Let us assume that each receiver is given a particular code. for example. Let us assume that we can measure these currents. namely. Let us assume that we transmit the code with amplitude I V. then the current of the following induced in two terms: 1 will be the 2 sum JtJ II}. This implies that the propagation that resulted in the induced currents described by the variables in Equations (44) and (45) can occur even when one is communicating in non-isotropic and heterogeneous media. and Ih in transmitting antenna #2. those related to the transmitter and the receiver. The relationship is much more involved in the time domain.3. respectively. since we know the impedance values of the loads. If the signal is to be IEEE Antennas and Propagation Magazine. When we are going to transmit that particularly weighted signal. Of course.irrespective of the shape of the transmitting or receiving antennas.for a transmitting antenna to radiate essentially a differentiated version of the temporal input waveform applied to it. let us ignore that for the moment. how to carry this out? We propose to achieve this objective using the principle of reciprocity. Nor is reciprocity dictated by the material medium influencing the propagation path. Vol. From the above observations. with VT2 = VR. This will induce currents If. wi. we know the voltage induced across the loads of the transmitting antennas. if we excite receiver 2 only with a voltage of 1 V. broadcast simultaneously over all the transmitting antennas. The problem now is. the electromagnetic signal will be directed to receiver 2 and will induce practically no energy at receiver 1. there will be an induced current d 2 in receiver 2. From these induced currents. It also does not depend on the presence of various near-field scatterers. so that when we receive any signal we know from which receiver it originated. we can essentially do adaptivity on transmission. So. From reciprocity. at the feed point of the receiving antenna marked #1 (i. One does not even require any knowledge about the spatial locations of the various antennas. = RTz = RRI = RTz = I Q. "" VR2 = 0. it is possible to direct the energy to a pre-selected receiver. Step 3: If we now excite transmitter I with a voltage Wj! and transmitter 2 with a voltage receiver .e. we can carry out the following procedure: Step 1: Excite receiver #1 current. Here. with VlI = VT. = 0. where the reciprocity theorem performs an integration over time. + wi!}. then it will induce a current IRI1 into receiver If. so that VRz = I V and Vlj = VTz= VR. The other portions of the electromagnetic network do not come into the picture. when VR. and iil in trans- mitting antennas 1 and 2.

comprising the base station. depending on the values of the weights chosen. If the operating frequency is I GHz. (47) respectively. by appropriate choice of the voltages exciting the two transmitting antennas. it is possible to select another set of weights that will result in receiver As developing a large signal at its terminals. and A3. due to an excitation voltage of 1 V at receiver A4. can be applied only at the terminals of the transmitting and receiving antennas. it is possible to select a set of weights based on reciprocity. producing a large value for the induced current. We now illustrate these statements through examples. if we apply the voltages W. then the received electromagnetic energy will be vectorially destructive at the terminals of receiver # 1. then the voltages received at the three base-station antennas will be v]'2.2 . simultaneously directing the received energy to a pre-selected receiver and having practically zero energy induced on the rest of the receivers). then it will generate the following voltages at the three transmitters: and the electromagnetic energy received from these two transmitting antennas will add vectorially at the terminals of receiver # I. and A3. Now. an example. by knowing the voltages that are induced at each of the transmitting antennas due to each one of the receivers. and so on. if we excite the antenna at the mobile. and will add vectorially at the terminals of receiver #2. This relationship. They are spaced with a half wavelength between themselves. Let us call the voltages received at transmitters AI' A2. Vol. and also in transmitters AI. A5. In contrast. Hence.2 on transmitting antenna #1 and on transmitting antenna #2. In receiver 2. 3.and that the voltages received from these three transmitting antennas will combine vectorially in such a way that they will produce practically zero currents at the terminals of the remaining two receivers. June 2003 V]" V]. v] . resulting in practically zero current flowing through its load. and they are separated from the transmitters bya distance of 2 m. and their lengths are 15 em. marked as A4. the first sum should be zero and the second sum should be unity. based on reciprocity. if there are more receivers than transmitting antennas. but some intermediate solution will still be possible. we can choose a set of complex wl Figure 12. and A) it is possible to direct the electromagnetic energy in such a way that it will add vectoiially at only one receiver (say. it is possible to direct the signal to a receiver chosen a priori. A4 ). 73 . 45. In a similar fashion. Equivalently. then. AI' A2 . and will be destructive at the terminals of receiver #2. and vl. A six-antenna system. Finally.1 Numerical Simulations should be such that the first sum should be unity and the ·1 second sum should be zero. and that will induce a current in both receivers A5 and A6. located at a base station. IEEE Antennas and Propagation Magazine. A2. irrespective of the nearfield electromagnetic environment in which they may be operating. as shown in Figure 12. Similarly. Next. then it may not be possible to achieve a perfect match (namely. not to receiver 1. with I V. they are of no consequence. These induced currents will generate voltages across the loads of the remaining five centrally loaded dipole antennas. the induced current wili be Wj21~ + Wllf2 . Step 4: Solution of Equation (46) provides the set of weights that. we consider three receiving antennas. we want to demonstrate that by choosing a proper set of weighted excitations on the transmitters AI' A2. if we excite receiver A6 with I V. VJ" V~2' and VJ). In this simulation. then the weights and wi Wi' As 8. ) Even though voltages are The caveat here is that if we apply a voltage equal to Wjl on transmitting antenna #1 and wi on transmitting antenna #2 then also induced in the other two receivers. The three antennas are con- then the current induced in receiver 1 will be the sum of these two terms: 1fj21~ + Wllj. the four equations described above that can direct the transmitted energy to a pre-selected re~eiver can be written in a compact matrix form as (46) No loss of generality will occur with the choice of a value different that 1 n for the impedance terminating both the transmitting and the receiving antennas. while the signals received at the terminals of the other two receivers. if we were to direct the signal to receiver 2 and sidered to be three dipoles of radius 5 mm. As. since it will only affect the scale factor for all four of the weights.directed to receiver I and not to receiver 2. producing zero current at its loads. We now illustrate how this can be accomplished based on the principles of reciprocity. consider three transmitting antennas. when used in conjunction with the signals received from the two receivers producing currents at the transmitters. Now. No. A5 and A6. A4 and A6. the three antennas are a half-wavelength long. In summary. the claim is that. will direct the transmitted energy to receiver 1 or to receiver 2. if we now excite transmitter 1 with a voltage Wj2 and transmitter 2 with a voltage wi. and A6. we excite receiver I with I V. and A3. It is independent of the size and shapes of the receiving antennas. First. ' V]l . nor does it depend on the near-field environment. based on the available information" from the above three experiments. In summary. Therefore. wiil be practically zero.

wl = -0.446 . and the received voltages will cancel each other resulting in practically zero currents at the terminals of the other . 74 This clearly demonstrates that by appropriately choosing the three complex amplitudes of the excitations at different transmitting antennas simultaneously it is possible to create an electromagnetic environment such that the induced electromagnetic energy will be vectorially additive at anyone of the terminals of the pre-selected receivers. From the ·principle of reciprocity.Oll.007. No. wi.062.. when all of the three transmitting antennas AI' A2• and A3 are simultaneously excited with the voltages W]l. 45. I and 3.003 . Based on these three voltages. June 2003 .0003 (all the currents are in rnA). + jO.366 . and wl =0. the electromagnetic environment will be such that it will induce a large current at the load of receiver I. if the three transmitting antennas are excited with the following three complex voltages. ITl =-0. ITJ = -0.OOI.446 . will result in an vectorial additive combination of the electromagnetic energy at receiver 1.415 + jO. IT2 = -0.679 . It is not. W/ = -0. This clearly shows that for this particular choice of the excitation voltages. obtained from the solution of Equation (48).jO.OI.0. . and that the received voltages will try to cancel each other at the terminals of the other receiving antennas. wi. It is practically zero only at the feed terminals and nowhere else! This is because the principle ofreciprocity only holds at the pre-specified terminals or ports of a network.OOI. Vol.679 . resulting in practically zero current. then the currents induced at the terminals of the three receivers will be given by I'Ii W)I.jO. Similarly. This is enforced through the following equation: = -0. jO. then one can show that the currents induced at the terminals of the three receivers will be I'Ii = -0.003+ jO.jO.0.845 + jO.679 .0003 . we are going to evaluate the induced currents at the loads located at the feed points of the three receivers. two receivers. when fed to the three transmitting antennas.jO.j0. Such a procedure can be carried out without knowing precisely the near-field environment in which the transmitter and the receiver are operating. w? = 0. IT) = 1. The receiver currents at the three locations will then be At Receiver I: At Receiver 2: At Receiver 3: The objective is now to select the weights W in such a fashion that the voltage received will be finite at receiver I. = RTJ = 50 Q ): 111 =1.the terminals of the other two receivers. while simultaneously inducing practically no currents in the other two receivers. .001- (all the currents are in rnA).679 . In summary. and W:Jlthat. Finally.007 (all the currents are in rnA). one can show that if the excitation voltages are chosen for the three transmitting antennas as follows: W]2= -0.0025 .366.j0. we carried out an electromagnetic simulation using WIPL-D [161]. This in no way implies that the current distribution on the receiving antenna is zero. while producing no appreciable currents at the terminals of the other two receivers.0012 . the steps involved in this new method are as follows: 1£££ Antennas and Propagation Magazine. voltages W]I. and obtained the following three induced currents in the three receivers (with R1I = Rr. using the above procedure.415 + jO.003 +jO.137 + jO. and practically zero currents at.0.jO. IT2 = 1.062 • then the electromagnetic energy will add vectorially at the terminals of receiver I. if the excitation voltages at the three transmitting antennas are chosen to be (48) Now.002 IT3 = 0.jO. and without knowing the precise spatial locations of the transmitting or the receiving antennas.845 wi = -0.jO. wi = -0.446 wj = 0. which clearly shows that the induced energy can be vectorially additive at only receiver 2.446 wi = -0. and zero at receivers 2 and 3.3.003 + jO.

pp. Y.3. McGraw Hill Publications. Part I. Lee. 7. July 1991. 936-942. J. Okumura. Despite the enormous efforts to date. Rappaport. S. and practically zero induced currents at the terminals of the other receivers. "A Statistical Model for Urban Multipath Propagation. "Statistical Channel Impulse Response Models for Factory and Open Plan IEEE Antennas and Propagation Magazine. Upper Saddle River. "Joint Space-Time Parameter Estimation for Wireless Communication Channels. June 2000. S. Acknowledgement Part of this work was supported by a grant from the National Science Foundation. 1. "Comparison of Urban Propagation Models with CWMeasurements. 2. Saunders and F. References 1. G. when the ports of a communication system are well defined. Bonar. pp. 9-10. From Pioneers to the 21st Century. R. and by the AFRL Sensors Directorate under grant F30602-95-1-0014. R. Gutieerez. Frontiers of Technology. "Explicit Multiple Building Diffraction Attenuation Function for Mobile Radio Wave Propagation. 14. Perez. Andersen. pp.136-146. Feb. 5. and o. "Site-Specific Propagation Prediction for Wireless In-Building Personal Communication System Design. B. B. pp. By following the above procedure. 42-49 . 16. 12. and K.. Fukuda. This will produce the maximum voltage at that pre-determined receiver. 5." IEEE Antennas and Propagation Magazine. 15. Vol. "Review on Radio Propagation into and within Buildings. Prentice Hall PTR. NJ. 15-27. Ohmori. 1. C. T." Electronics Letters. L." lEE Proceedings-H. Takamizawa. 317-325. 1992. S. 61-73. 10.T. 1968. 825873. G. Radio Propagation for Modern Wireless Systems. H. 4. "Empirical Formula for Propagation Loss in Land Mobile Radio Service. 46." IEEE Communications Magazine. From our numerical simulations [160]. May 1991. 9. 1. 42. Bertoni. 1276-1277. Conclusions Propagation models are needed not only for installation guidelines. VT-43. Performance Improvement. Mobile Cellular Telecommunications New York. 1788-1796. Prentice Hall PTR. Step 2: Using these induced currents. 1988. and S. K. 90-92. Youshida. Saez de Adana. Seidel. Y. T. 16. 1997. "Smart Antennas. G. the inversion of which will contain the information necessary to excite the transmitting antennas." IEEE Transactions on Communications. and System Considerations." IEEE Antennas Propagation Magazine. Walfisch and H. maximizing the induced currents. No. but they also playa key part in any analysis or design that strives to mitigate interference. I. R. pp. 1333-1343. ECS-9901361. S. 794-807. depending on the choice of which of the mobile receiving antennas the signal is meant for. R. 45. June 2003 . 1994. 2." IEEE Transactions on Vehicular Technology. AP-36. we observed that this principle can be applied over a relatively wide band of frequencies. Bonar. Wireless Communications: Principles Practice. An FDTD method has been described to analyze wave propagation through the walls in a building. "Efficient Ray-Tracing Techniques for Three Dimensional Analyses of Propagation in Mobile Communications: Application to Picocell and Microcell Scenarios. 4. L. VT-21. Seidel and T. "Prediction of Mobile Radio Wave Propagation over Buildings of Irregular Heights and Spac75 11. S. W. pp. Godara. we have surveyed some of the typical propagation models that provide good estimates for both large-scale and small-scale fading channels. L. generate a matrix. 33. Turin. S. "Propagation Measurement and Models for Wireless 'Communications Channels. a set of weights can be solved for that will result in a vectorial addition of all the incident electromagnetic energy at the terminals. "Rayleigh Fading Channels in Mobile Digital Communication Systems. 5. and K. Building Radio Communicate System Design. M. 1989. In addition." Proceedings of the IEEE. one can carry out adaptivity on transmission based on the principle of reciprocity. 1972. 3." IEEE Transactions on Signal Processing.1980. "Field Strength and its Variability in VHF and UHF Land Mobile Service. 40. Saunders and F. F. Hata. In this paper. pp. Sklar. . S. Raleigh and T. pp." IEEE Vehicular Technology Society 42nd VTS Conference. 2.1031-1060. pp. Bertoni. T. much work remains in understanding and predicting the characteristics of mobile communications channels. Step 3: Then. Part 1. pp. 10. & 9. Rappaport. July 1997." IEEE Transactions on Antennas and Propagation. 1-9. 12. 3. This methodology works even when there are near-field scatterers. Chryssomallis." Review Electrical Communication Laboratory.pp. C. an efficient ray-tracing method has been presented for tracing rays in an indoor propagation system. pp. 1991. 19. Finally.Step 1: Excite each of the mobile receiving antennas and measure the currents that are induced in the base-station transmitting antennas. 18. it has been seen that the principle of reciprocity can be applied to direct the transmitted energy to a pre-specified mobile unit without having any a priori knowledge of the electromagnetic environment. 3. "A Theoretical Model of UHF Propagation in Urban Environments. 17. 7. ' 6. Catedra." IEEE Transactions on Vehicular Technology. 1996. 27. 138. Rappaport. 11. April 1998. pp. 879-891. "Applications of Antenna Arrays to Mobile Communications. January 1995. L. 14. Rappaport. 35. VT-29. M. Y. Upper Saddle River. COM-39.pp. pp. F. Hence. Systems. 13. D. Low. M." IEEE Transactions on Vehicular Technology. i29-136: and 8. at least in this case. pp. Feasibility." IEEE Communications Magazine. 85. May 1998. Molkdar. NJ. 2000. S. J. Boros. E. 9. the propagation-path modeling can be avoided through the use of the principle of reciprocity.

and N. Seidel. C." IEEE Transactions on Antennas and Propagation. "Characterization of UHF Multipath Radio Channels in Factory Buildings. pp. AP-41. 31. Cardona. C. August 1995.1999. "914 MHz Path Loss Prediction Models for Indoor Wireless Communications in Multifloored Buildings. 27. 36. 1992: pp. Samper. Takeuchi. "Characterising Buildings for Propagation at 900 MHz. Bultitude. and S. Cardona. L Blackard'. J. 299-302. pp. 2. pp. 618-620. and J. A. Sau. October 1994. Ramos. and E. pp. 38. 32. pp. 1058-1Q69. September 1983. AP-39." IEEE Transactions on Vehicular Technology. "Finite Conductivity Uniform GTD versus Knife Edge Diffraction in Prediction of Propagation Path Loss. 28. Ruiz. "Indoor Wide Band Radio Wave Propagation and Models at 1. Rappaport. 9961001. pp. Niver. 1. T. Valenzuela. 20. P. "Measurement Characterization and Modeling of Indoor 800/900 MHz Radio Channels for Digital Communications. VT-48. 137·144. "Radio-Wave Propagation for Emerging Wireless Personal-Communication Systems. Har. W." Electronics Letters. "Effect of Radio Propagation Path Loss on DS-CDMA Cellular Frequency Reuse Efficiency for the Reverse Channel. 860. Kawasaki. 1230-1232. Juan-Llacer. A A. 24. pp. 210Q2101. Ikegarni. June 2003 . pp. M. D. pp. Pavlasek. 2. "Propagation Modeling for Indoor Wireless Communication. ICAP' 89. Taga. K. 29. AP-47. .S. 1999. H. W. 5-12. D. pp. 7.pp. 41. R. 10. Valch. and Outage Models as Functions of Antenna Height for Macrocellular System Design. 14391447. August 1989. pp. VT-41. H. 37. T. pp. 1. "Prediction of Cellular Characteristics for Various Urban Environments. 45. March 1987." IEEE Transactions on Vehicular Technology." IEEE Transactions on Vehicular Technology. Report to FCC. K. pp. B." IEEE Transactions on Vehicular Technology. AP41. Feuerstein. M. August 1998." IEEE Transactions'on Antennas and Propagation. 26." Electronics & Communication Engineering Journal. Whitman. January 1984. W. AP. and R. "Application of Some Theoretical Models for Coverage Prediction in Macrocell Urban Environments. 22. Rowe. pp. Chadney. J. 319-324. 3. and T. 24. 1463-1468. 34. Ganesh. belay Spread. 39.0 GHz. 36. Olmos. 8. T. 221-228. N. 25. Ichitsubo. Kim. Pahlavan. D." IEEE Journal on Selected Areas in Communications. . 1. S. "A Comparison of Time Delay Spread and Signal Level Measurements within Two Dissimilar Office Buildings. p. S. Rappaport. N. 6.621-629.42. 3. Cheung. pp. "Comment on Diffraction Loss of Rooftop-to-Street in Cost 231-Walfich-Ikegami Model." Electronics Letters." Electronics Letters. S. 43.. Davasirvatham. H. Beaunstein. 1992. October 1995." IEEE Transactions on Vehicular Technology.4. "Assessment of GTD for Mobile Radio Propagation Prediction. S. "Multipath Propagation Model for Line-of-Sight Street Microcells in Urban Area. and A. Furuno. 1994. Xia. R. G. 135-145." Electronics Letters. pp. S. F. Tam and V. Conf. S. and T. F. "UTD Solution for the Multiple Building Diffraction Attenuation Function for Mobile Radiowave Propagation.20. August 1994. "A New Empirical Model for Indoor Propagation Prediction.5. J. 1. 1. AP35. Murch. AP-32. 1451-1452.3. B." IEEE Transactions on Antennas and Propagation. and N. "Cross-Floor Signal Propagation Inside a Contemporary Ferro-Concrete Building at 434. and R." IEEE Transactions on Vehicular Technology. M. 70-76. Rappaport and S. "Mechanisms Governing UHF Propagation on Single Floors in Modern Office Buildings. VT-43 . K. 9293. No. VT~4. "Path Loss. 1421. J. Y. Alexander. 207-217. Telesis Technologies Laboratory. 35. Neve and G. Pundey. October 1998. 26.2. 200Q. 40. S. R Milstein. H. 1. 30. D." August 1991. Part 2: Propagation. and 1705 MHz. S. Tran. A. 25.496-504. 7. Saleh and R. and R. J. 42. Rappaport. 21." IEEE Antennas and Propagation Magazine.. R. L. 1.3 GHz and 4. Agusti. October 1990.3. T. . Luebbers." Electronics Letters." IEEE Transactions on Antennas and Propagation. L R Maciel. Perez.864. Xia. 1989. 1. T. 487-498. and 1. T. 33. 41. 276-281. pp. 231-242. Rowe. K. 29. 5. H. 3. March 1991. S. 5.pp. 46. Watson. Rappaport and 1. M. Hotaling. 1999. 19. 1987.1987. on Antennas and Propagation. T. AP. J. 5. A. 34. "Experimental License 40." IEEE Communications Magazine. Honcharenko. "Radio Propagation Characteristics tor Line-of-sight Microcellular and Personal Communications. April 1997. Haand. K. Lindsay-Stewart. Sandhu. October 1993. "A Statistical Model for Indoor Multipath Propagation. 3. 45. August 1992. 1800-1802." IEEE Transactions on Antennas and Propagation. Juan-Llacer 33." IEEE Transactions on Antennas and Propagation. M J. pp. VT-47. "Software Tool for Optimising Indoor/Outdoor Coverage in a Construction Site. pp. pp. S. Segal." IEEE Transactions on Antennas and Propagation. Seidel and T. 22. and H. VT-48. 23. Vol. "Theoretical Prediction of Mean Field Strength for Urban Mobile Radio. AP-37. 44." IEEE Transactions on Vehicular Technology. Lebel. 128-137. Dailirig. "In-Building Propagation at 9Qo MHz and 1650 MHz for Digital Cordless Telephone. "A Theoretical Model for Radio Signal Attenuation Inside Buildings." IEEE Transactions on Antennas and Propagation. Bertoni. February 1989. 47." 6th Int. VT-49." IEEE Transactions on Antennas and Propagation. R. S. M. 6." IEEE Antennas and Propagation Magazine. D. E. Y.422-427.ings. 1999. 3. 76 IEEE Antennas and Propagation Magazine. Owen and C." Electronics Letters. pp. Bertoni. pp. 3. pp. Y. G. pp. "Multipath Propagation Measurements on Manufacturing Floors at 910 MHz. Yoshida. April 1993. 225-227. M.

W. Hsu. Ontario. S. 782-789. E. J. March 2000. T. VT-48. 3.pp. Pino. January 1997. Willis III. Tan and H. VT-43. S." IEEE Transactions on Vehicular Technology. April 1998." IEEE Transactions on Vehicular Technology. August 1995. 66. 50. VT-48. Ling. August 1995. N. 4." IEEE Transactions on Vehicular Technology. Chen and S. Tan and H. 15-28. M. May 1997. 72. 322335. 49. March 2000. 2. Whitman. VT-46. A." IEEE Transactions on Electromagnetic Compatibility. 55. 73." IEEE Transactions on Vehicular Technology. Lawton and J. pp. 621-629. "Exhaustive Ray Tracing Algorithm for Microcellular Propagation Prediction Models. Sanchez. pp. VT-46. "SBR Image Approach for Radio Wave Propagation in Tunnels with and without Traffic.3. Gambi. H. 57. 508-518. J. 931·944. 1804-1805. 18. AP-46. V. "The Application of a Deterministic Ray Launching Algorithm for the Prediction of Radio Channel Characteristics in Small-Cell Environments. 4. 3. 7. "Propagation Mechanisms in Microcell and Indoor Environments. K. Kim. R. 68. 54.. 16. Wagen. pp. August 1996. K. pp. G. March 1996. Tan. J. Mckown and R. 65. 52. VT44. U. W. pp. Y. 56. D. 21. 3. Zhang. "Efficient Ray-Tracing Techniques for Three-Dimensional Analyses of Propagation in Mobile Communications: Application to Picocell and Microcell Scenarios. pp. 58. P. 570-578. A. VT-48. 45. and F. G. W. Pottie. S. Tarng and T. pp. 3. Evans. M. and B.9 GHz. and R. pp. and E. pp. "Comparison of GTD and FDTD Predictions for UHF Radio Wave Propagation in a Simple Outdoor Urban Environment. "Propagation Model for Microcellular Communications Applied to Path Loss -Measurements in Ottawa City Streets. Tamg. J." IEEE Transactions on Antennas and Propagation. "A Microcellular Ray-tracing Propagation Model and Evaluation of Its Narrow-Band and Wide-Band Predictions. 32. and M. and Y. pp. pp." IEEE Transactions on Vehicular Technology. V. "A New Approach to 3-D Ray Tracing for Propagation Prediction in Cities. D. 2. Gutierrez. "Indoor Channel Characterization for Wireless Communi- ! IEEE Antennas and Propagation Magazine. and P. G. F. Erceg. Scanlon. May 1997. N. "Effective Models in Evaluating Radio Coverage on Single Floors of Multifloor Buildings. 333-342. November 1994. J. pp." IEEE Transactions on Vehicular Technology. VT -45. B. A. A. L. 74. S. A. J. Bertoni. Moore. pp. Dersch and E. RahmatSamii. S. J. Chen and S. de Haro. F. H. K. "Radio Propagation Measurements and Prediction Using Three-Dimensional Ray Tracing in Urban Environments at 908 MHz and 1. pp. O'Brien.2. Rosen. Gardiol. J. July 1997. W. 4. November 1997. "A Ray Launching Method for Radio-mobile Propagation in Urban Area. 60. and T. 313-317. VT-44.1991. R. McGeehan. May 1995. No. Valenzuela. G." IEEE International Symposium on Antennas and Propagation Digest. C. 5. Jr. J. Catedra. June 2003 77 . J. J. A. "Hybrid Image/Ray-Shooting UHF Radio Propagation Predictor for Populated Indoor Environments. Jeng. F. Levy." IEEE Transactions on Vehicular Technology. Tan. pp. 51. J. pp." IEEE International Symposium on Antennas and Propagation Digest. W. and E. W. Y. M. Zollinger. C." IEEE Transactions on Vehicular Technology.. 3. 1997. 955-969. 622630. S." IEEE Transactions on Vehicular Technology. DeFlaviis. pp. 64. 2. 53. Tan and H. H. Mokhtari and P. 17. M. J. 255-261. Calvo. "An SBRlImage Approach for Radio Wave Propagation in Indoor Environments with Metallic Furniture. J. S. K. H. "Three-Dimensional Modeling of 900-MHz and 2. W. G. S. November 1994. P. Ling. 62. VT-43. Y. October 1999. Canada. "Two-Dimensional Raytracing Modeling for Propagation Prediction in Microcellular Environments. M. and M. J. S. pp. 428-436. Valenzuela. Luebbers. Jeng. "Comparisons of a Computer-Based Propagation Prediction Tool with Experimental Data Collected in Urban Microcellular Environments. Perez-Saez. 1058-1066. Hamilton. E. Niver. March 2000. "A Novel Propagation Modeling for Microcellular Communications in Urban Environments." IEEE Transactions an Vehicular Technology. "A Theory for Propagation Path-Loss Characteristics in a City-Street Grid. Guarino Jr. pp. Kim." IEEE Journal on Selected Areas in Communications. K. Rappaport. de Adana. 1021-1026..pp. Cullen. C. Athanasiadou.48. L. S. Patwari. 519-527. "Theoretical Model for Radio Signal Attenuation Inside Buildings." Electronics Letters. Liang and H. January 1999. Kondylis. Lazaridis. M. Thomas.Y." IEEE Journal on Selected Areas in Communications. Rustako Jr." IEEE Transactions on Vehicular Technology. 59. "An Efficient Implementation of a Three-Dimensional Microcell Propagation Tool for Indoor and Outdoor Urban Environments. F. 624-625. pp. J." IEEE Transactions on Antennas and Propagation. Tan. VT -49.44-GHz Radio Propagation in Corridors. R.3l. McGeehan. Jan and S. "Improved Three-Dimension Ray Tracing Technique for Microcellular Propagation. 677-684.1999. J. 67. pp. Bic. . G. AP45. 27-30. 2. H. Liu. Erceg. G. Vol. Nov. VT-49. S. EMC-37. E. 70." Electronics Letters 35. Durgin. C. A. pp." Electronics Letters. 71. L. May 1997. 33. P. 853-863. 3." IEEE Antennas and Propagation Magazine. and J. J. 63. 1412-1413. 2. 3. Schuster and R. and O.1. 6. "Fast Two-Dimensional Diffraction Modeling for Site-Specific Propagation Prediction in Urban Microcellular Environments. Gabillet. "Improved 3D Ray Launching Method for Wireless Propagation Prediction." IEEE Transactions on Vehicular Technology. 98-106." Electronics Letters. 2022-2025. Rizk.S. Villanese. Nix. M. "Comparative Study of Lateral Profile Knife-Edge Diffraction and Ray Tracing Technique Using GTD in Urban Environment. May 1999. London. G. pp. 3. S. Fortune. S. Kenny.1503-1505. 40. L. Jeng. VT-46. and J. June 1998. Fortune and R. Chang. 1540-1543. G. Rossi. 61. 1.4. F. 69. 1. 1991." IEEE Network Magazine. "Ray Tracing as a Design Tool for Radio Networks." IEEE Transactions on Vehicular Technology. August 1995. 15. pp.

Ko. Blaunstein. and D. Chaudhuri. Zaporozhets. W. 673680. VT-45. N.Antennas and Propagation. 12. Suzuki. COM-47. Ying." IEEE Communications." Electronics Letters. Csaba. "Indoor Electric Field Level Prediction Model Based on the Artificial Neural Networks. pp. U. 1782-1789. F.2000. pp." IEEE Communications Letters. C. Safak. 881-890. K. 251-255. 22502253. 1." IEEE Transactions on Communications. Neskovic and D. 1999. pp. Inc. 88. 6. 9." lEE Proceedings . 77. S." IEEE Transactions on Antennas and Propagation. C. 31. A. Bahr. C. June 1994. B. pp. February 1997. "3D Ray Launching and Moment Method for Indoor Propagation Purposes. Killat. Landstorfer.8 GHz. P. 4." IEEE Transactions on Antennas and Propagation. 84. Hess. 91. and F. 454-458." Electronics Letters. IEEE Transactions on Communications. G. "Statistical Analysis of the Power Sum of Multiple Correlated Lognormal Components. K. 58-61. 4. 6. "Generalized Rice-Lognormal Channel Model for Wireless Communications. Kaveh. J. 94. April 1998. "K Distribution: An Appropriate Substitute for Rayleigh-lognormal Distribution in Fading-Shadowing Wireless Channels. I. pp. Stocker. R. 97." IEEE Transactions on Antennas and Propagation." IEEE Vehicular Technology Conference. Satellite Channels and its Application to Nongeostationary 78 IEEE Antennas and Propagation Magazine. 82. "MomentMethod Estimation of the Ricean Kvfactor. Safavi-Naeini. February 2000. 214-221." IEEE Transactions on Vehicular Technology. Laue. pp. and S. April 1993. Tjhung and C. pp. pp. 95. January 1997. 190-192. Rappaport. J. B. Paumovic." IEEE Transactions on Vehicular Technology. Sandor. Thomas. C. Artech House. and C." 1989 International Conference on Acoustics. Olyslager. I. 93. and Signal Processing. June 2003 . "Bit Error SimulaRadio Communications Using TwoImpulse Response Models." IEEE Transactions on Antennas and Propagation. A." IEEE Transactions on Vehicular Technology. 12. 130-134. Gschwendtner. 2. and J. "Application of Vector Parabolic Equation Method to Urban Radiowave Propagation Problems. "Lattice Boltzmann Method for Wave Propagation in Urban Microcells. pp. MA. "A Statistical Model for Urban Radio Propagation. Luthi. pp. B. Michelson. 79. 4. August 1997. 80. VT-3. C. Vatalaro. 155-160. AP-46. June 1998. 91-92. COM-42. IL. Mazar and A. Patzold. H. F. 78. COM-48. "Average Field Attenuation in the Street Waveguide. pp. V. 144. Antennas and Propagation. De Zutter. VT-37. pp. pp. Speech. "Modeling of Signal Envelope Correlation of Line-of-sight Fading with Applications to Frequency Correlation Analysis. M. August 1993. VT-48. "Coverage Prediction for Mobile Radio Systems Operating in the 800/900 MHz Frequency Range. pp. ~'A Hybrid Technique Based on Combining Ray Tracing and FDTD Methods for Site-Specific Modeling of Indoor Radio Wave Propagation. Wu." Electronics Letters. E. Brennan and P. VT-46. S. Yip and T. P. Chang and W. Abdi and M. G. 165-169. and T. Greenstein. "A Simulation Model for Nakagami-m Fading Channels. Lauer. F. D.cations Using Reduced Finite Difference Time Domain. "Application of the Fast Far-field Approximation to the Computation of UHF Path Loss Over Irregular Terrain. Vol. De Backer. 33. 1402-1406. 89. 11. 100." IEEE Communications Letters. pp. 98. Yang. Antennas and Propagation. Chicago. Pamp. "The Study of Wave-Propagation through a Windowed Wall at 1. Iwai. 907919. 75. Kunisch. 1899-1900. "Multi-Mode FDTD Simulations of Indoor Propagation Including Antenna Properties.3. "A Statistical Model for Land Mobile. Z. L. August 1999. 4. 175-176. Handbook of Land-Mobile Radio System Coverage. pp. Nagy. 393-405. pp. 34. Krantzik and D." IEEE International Symposium on Antennas and Propagation Digest." IEEE Transactions on Communications. Yang. Corazza and F. A. PIMRC '97. pp. m> I".6. and V. 22. Cullen. and F. "Environment-Adaptation Mobile Radio Propagation Prediction Using Radial Basis Function Neural Networks. A. A. Y. Erceg. "A Deterministic Digital Simulation Model for Suzuki Processes with Application to a Shadowed Rayleigh Land Mobile Radio Channel. Wu. T. 102. 6. 3. 1989 ICASSP-89. E. Norwood." IEEE Transactions on Vehicular Technology. 81." lEE Proceedings H Microwaves. 253-256. J. Borjeson. 86. pp. T. B." lEE Proceedings . Szabo. 1977. pp. pp. 1769-1772. Karasawa and H. A. 743-754. June 1999. "Fade Statistics in NakagamiLognormal Channels. T. May 1996.Microwaves. A." IEEE Transactions on Communications. tion for pi 14 DQPSK Mobile ray and Measurement-based Journal on Selected Areas in pp. Ng. 5.4. "Simulations and Measurements for Indoor Wave Propagation Through Periodic Structures. 99. 101. February 1993. Woff. Yang and B. October 1995. pp. 76. pp. 96. VT-46. 90. 1.45." in Proceedings of the 1995 IEEE 45th Vehicular Technology Conference. C. Wolf. pp. "Analysis of a Modified Suzuki Fading Channel Model. 1. Fung. I. "Neural Network Approach to Prediction of Terrestrial Wave Propagation for Mobile Radio. Indoor and Mobile Radio Communications. 140. Vatalaro. COM-25. 87. AP-46. A. Wagen. 1996. 85. pp. I. 83. J. "A Ray-Tracing Method for Modeling Indoor Wave Propagation and Penetration. 2201-2203. February 1988. 92. VT-42. 1997. 2. 851-852. L. 318-331. "Propagation Model of a City Street for Personal and Microcellular Communications. M. 3. E. O. 6. Chai. H. 384-387. No. and B. 315-320. IEEE Vehicular Technology Society Committee on Radio Propagation. K. 1. June 1998. December 1998. Bronshtein. pp. 146. May 2000." IEEE 46th Vehicular Technology Conference. 1999. Z. December 1999. 3-72." The 8th IEEE International Symposium on Personal. Chopard.

7. H. Hashemi. H. 26. L. I.3. T. Babalis and C. Morrison. T. Bertoni." lEE Proceedings I Communications. 18. F. T. "An Improved Ray-Tracing Propagation Model for Predicting Path Loss on Single Floors. L. lEE Proceedings-H Antennas and Propagation. Speech and Vision. Rappaport and S. 20772079. VT-43. 3. pp. and R. "Modeling and Measurement of 2. "Propagation Characteristics on Microcellular Urban Mobile Radio Channels at 910 MHz.22. Tarng. 5. No. "Pulse Propagation Characteristics at 2. Pahlavan. Liu and 1. Fattouche. September 1993. 738-741. VT-48. and M. Bultitude. October 1996. and 11. 119. 1272-1288. 322335. R. 117. 3. 107. "A Microcellular Ray-tracing Propagation Model and Evaluation of its Narrow-Band and Wide-Band predictions. Ichitsubo. 25. T. 127. pp. pp. Hahn. Y. Lawton and J. I." IEEE Journal on Selected Areas in Communications." IEEE Transactions on Vehicular Technology. 14. Vol. VT46. 138. "Channel Modeling for 2GHz-band Urban Line-of-sight Street Microcells. Jain. Prokki. VT -45. February 1989. S. February 1989. 653-657." IEEE Journal on Selected Areas in Communications. pp. 140. Devasirvatham. "Indoor Wideband Radiowave Propagation Measurements at 1. pp. 11. 4. Prasad. 967-978. G. "A Deterministic Ray Tube Method for Microcellular Wave Propagation Prediction Model. A. L." IEEE Transactions on Vehicular Technology. and M. pp. 116. I. 122. October 1990. 194-196. R. 43 [-440. Rappaport. A.7 GHz Inside a Metropolitan Office Building. S. 579-592. 115. 109. "Impact of the Combined Slow and Fast Fading Channel Characteristics on the Symbol Error Probability for Mu1tipath Dispersionless Channel Characterized by a Small Number of Dominant Paths. 225-227. 956-958. H. "The Application of a Deterministic Ray Launching Algorithm for the Prediction of Radio Channel Characteristics in Small-Cell Environments. 1344-1350. Yeh. 4. 477-485. "Estimating Local Mean Signal Strength of Indoor Multipath Propagation. Janssen. C. J. "The Dependence of Indoor Radio Channel Multipath Characteristics of Transmit/Receiver Ranges.3 GHz and 4. M.56-59. 114. Son and N. pp. 125. Tholl. Y. J. G. 27. 955-969. B.5 GHz. pp. Furuno. 11. "Multipath Propagation Model for Line-of-sight Street Microcells in Urban Area. Furuno. "A New Path-GainJDelay-Spread Propagation Model for Digital Cellular Channels. 104. November 1994. Polydorou and C. "Propagation Considerations for the Design of an Indoor Broad-Band Communications System at EHF. Singh. and T. Hotaling. June 1991. Seidel. 25. and R. Nix. COM-44. "900 MHz Multipath Propagation Measurements in Four States Cities. 121. April 1996. 1. Hawbaker and T. March 2000 pp. G. R. pp. September 1993. P. 108. Seidel. X. T. 7. R. 106. Pahlavan. 126. 422-427. and Y. August 1996. Erceg. and D.3. III. Kim. 721-730. Kawasaki. P. July 1989. P. August 1994." IEEE Transactions on Communications. Bedal. 203-212. May 1999. pp. Y. March 2000. 4. 979-990. 21. "Comparison of Indoor Propagation Channel Characteristics at Different Frequencies.44 GHz Radio Out-Of-Sight Propagation on Single Floors"." IEEE Journal on Selected Areas in Communications. pp. Ganesh. and K. Davies. V. VT-46.2. J. I. D. "Experimental Characterization of EHF Multipath Indoor Radio Channels. VT-47. Melancon. February 1997. 103. "Path Loss. N. 120. pp. R. S. Capsalis. Chen. and C. and D. and J. Capsalis. C. 113. and R. G." IEEE Transactions on Vehicular Technology. Lord. L. November 1991. R. pp. 1999. M. Zaghloul. T. Athanasiadou. "Multipath Propagation Measurements on Manufacturing Floors at 910 MHz." Electronics Letters. VT46. 748-755. 14. 112. R. Zaghbul. pp. 45. McGeehan. Microwave and Optical Technology Letters. COM-47. N. January 1997. D. 7. S. 8. 3. A. 39-41. Suwa. 262272. G." IEEE Transactions on Vehicular Technology. R." IEEE Transactions on Vehicular Technology." IEEE Transactions on Antennas and Propagation." IEEE Transactions on Vehicular Technology. "Time Delay Spread Measurements at 850 MHz and 1. Taibi and G.4 GHz Inside Buildings. H. L. pp." Electronics Letters. M. H. 4. 1. S. J. C. Rappaport. Myung. pp. R. pp. 15. H.75. Scattering and Multipath Delay Statistics in Four European Cities for Digital Cellular and Microcellular Radiotelephone. and R. Wang. 235-245. H. C. Stem. 1800-1802. S. May 1997." Electronics Letters. R. C. 123. August 1997. VT-40. McGeehan. Banerjee. Li. pp. Valenzula.Orbit Systems." Electronics Letters. R.0 GHz. IEEE Antennas and Propagation Magazine. P. Taga. Jacobs. Greenstein." Microwave and Optical Technology Letters. Z. VT49. "Wideband Indoor Channel Measurements and BER Analysis of Frequency Selective Multipath Channels at 2. 118. AP-47. Landron. Ji. 105. October 1991. Zhou. 3. P. Murray." IEEE Transactions on Vehicular Technology. August 1993. "Statistical Modelling and Computer Simulation of Indoor Radio Channel. "Statistics of Short Time and Spatial Variations Measured in Wideband Indoor Radio Channels". 110. 22. Morrison. pp." IEEE Transactions on Communications. 31-39. Y. 297-302. 1. V.4. K. Y. S. S. 153161. VT43. Bultitude. pp. pp. G. January 1999. 3. pp." IEEE Journal on Selected Areas in Communications. "Impulse Response Modeling of Indoor Radio Propagation Channels." IEEE Transactions on Vehicular Technology. pp." IEEE Transactions on Vehicular Technology. 10. D. and M. . K." IEEE Journal on Selected Areas in Communications. M." Electronics Letters. H. Pahlavan. February 1998. pp." IEEE Transactions on Vehicular Technology. Ganesh and K. T. June 2003 79 . pp. Clark. A. 124. Bultitude and G. S. E. August 1999. R. H. M. Taga. "A New Theoretical Model for the Prediction of Rapid Fading Variations in Indoor Environment. 25. 128. 1. Ganesh and K. Stigter. O. January 1989. pp. 1. Delisle.

2. 6. 1. pp. S. 146. 1993. 10. 143. "A Theoretical Formulation for the Distribution Density of Multipath Delay Spread in a Land Mobile Radio Environment. 150. 142. "A Three-Dimensional Parabolic Equation Applied to VHFIUHF Propagation over Irregular Terrain. "Wireless Propagation in Buildings: A Statistical Scattering Approach. and S. M. 12 -14. L." IEEE Transactions on Vehicular Technology. 10. SP-45. 55-57. H. 2. 2477-2484. M. Hashemi and D.1999. Shepherd. Cotton." IEEE Transactions on Vehicular Technology." IEEE Transactions on Vehicular Technology. pp. J. M." IEEE Communications Letters. Y. February 1994. 149. VT-48.3. "A Body-Shadowing Model for Indoor Radio Communication Environments. P. Bertoni. C. Vanderveen. 145. 1999. Vanderveen. A. H. Pennock. 145. Paulraj.129. Y. 405-418. Ji. Jenn. J. 134. G. K. Antennas." IEEE Antennas and Propagation Magazine. 151." IEEE Transactions on Vehicular Technology. 139.E. MTT-50. "2D RayTracing Model for Indoor Radio Propagation at Millimeter Frequencies. 130. Wang. M. 1999. van der Veen. H. "Improvement of the Computation Efficiency for a Ray-Launching model. pp. April 2001. "Joint Estimation of Time Delays and Directions of Arrival of Multiple Reflections of a Known Signal. R. pp. L. Zhang. Zelley and C. pp. pp. I. "Analysis of Composite Walls and Their Effects on Short-Path Propagation Modeling. pp. 1. Donohue and 1. Li. Kozono and A. and A. H." IEEE Transactions on Vehicular Technology. Bertoni. "Multi-Ray Propagation Model for Indoor Wireless Communications." IEEE Transactions on Antennas and Propagation." IEEE Transactions on Communications. and B. Papadias. Chang and H. 110-120." IEEE Communications Letters. 1. L. 1. Vanderveen. No. pp. Holloway and P. Tholl. 1. 154. 349-353. Honcharenko and H." Microwave and Optical Technology Letters. "Prediction of Multipath Delay Profiles in Mountainous Terrain. pp. G. C. Loew. 18. "Joint Angle and Delay Estimation (JADE) for Multipath Signals Arriving at an Antenna Array. VT-43. Paulraj. "Mobile Propagation Loss and Delay Spread Characteristics with a Low Base Station Antenna on an Urban Road. "Analysis of Impulse Response Measurements for pes Channel Modeling Applications. McKenna. and the Study of Diversity Techniques". S. A. VT-48. 336-346. van der Veen. Driessen. 135. May 1997. "Efficient Ray-Tracing Methods for Propagation Prediction for Indoor Wireless Communications. C. Sarkar. W. L. 136. B. 147. M. M. Tran. pp. L. Y. pp. R. 4. "UHF Propagation Prediction for Wireless Personal Communications." lEE Proceedings Microwaves. 249-254. 3. C. J. J. 4.4. l40. 1995. pp. 1998. 9. Kuttler. 920-927. 5. 2. "A Model for Predicting the Power Delay Profile Characteristics Inside a Room. January 1997. Kim. and A. "Joint Angle and Delay Estimation Using Shift-Invariance Techniques." IEEE Transactions on Antennas and Propagation. 142145. 82. Bertoni. Wax and A. June 2003 . B. AP-47. Tam and V. pp. and Propagation. 12 -14. Iskander and Z. C. Parsons. VT-46." Proceedings of the IEEE. 379-388. "Transmission and Reflection Characteristics at Concrete Block Walls in the UHF Bands Proposal for Future PCs. AP-48. August 1997. Taguchi. 1586-1596. pp. T. 1999. Papadias. 1. and P. and J. 1453-1462. "Statistical Analysis and Characterization of the Indoor Propagation Channel. Hoffman." IEEE Transactions on Signal Processing. Babich and G. A. 2000. P. 1997. 2000. VT-48. "Path-Loss Prediction Model for Microcells. February 1998. 43. Obayashi and J. Perini. R. 103-109." Electronics Letters. 2. Ullmo and H. H. K Sarkar. pp. Hwang. pp. 29. "Propagation Prediction Models for Wireless Communication Systems." IEEE Transactions on Antenna and Propagation. C. and H. Honcharenko.Microwaves. pp. Gibson and D. "Prediction and Measurement of Wall Insertion Loss." IEEE Transactions on Vehicular Technology. pp. Holloway. 138. 662673. 132. L. Leshem. X. M. and A. H. U. 1. Xia. P. and Propagation. 135-137. COM-48. D. Baranger. H. W. pp. A. 3. "Joint Angle and Delay Estimation (JADE) for Multipath Signals Arriving at an Antenna Array. W. 141." IEEE Transactions on Vehicular Technology. 1996. Paulraj. Antennas. Chen. C. and A." IEEE Transactions on Microwave Theory and Techniques. 2001. pp. H. 1999. Zander. 1921. lEE Proceedings . N. Lombardi. B. 156. H. and L. R. R. "Analysis of the Effects of Walls on Indoor Wave Propagation Using the FDTD Method. 730-738. K. Y. 144. pp. 152. Li. 3. T." IEEE Transactions on Signal Processing. 45. and T. 153. VT-42. pp. "Joint Angle and Delay Estimation Using Shift-invariance Properties. 137." IEEE Signal Processing Letters. 1993. AP-46. SP-46. Li. 2000. J." IEEE Transactions on Vehicular Technology. Ghobadi. 3. "Statistical Modeling and Simulation of the RMS Delay Spread of Indoor Radio Propagation Channels. D. 1998. 5. Vol. 1333-1359. l. October 1997. C. Paulraj. VT-43. 1110-1120. Wepman. T.1999." IEEE Transactions on Antennas and Propagation. pp. 32. AP-42. 2002. 2. C. Constantinou.1. 232-239. 131. 2. January 1997. 133. 455-464. Z. Ji. K. Macel. t." IEEE Transactions on Vehicular Technology. pp. VT44. 3. Z. 303308. 41-49. "Propagation Modeling over Terrain Using the Parabolic Wave Equation. Yun. Xia. H. e. 155. F. 148." IEEE Journal on Selected Areas in Communications. L. 145. pp. pp. and H." IEEE Transactions on Antennas and Propagation.pp. 1994. R.4. pp. D. C. 947-955. AP-47. 1994.3. Har. 80 IEEE Antennas and Propagation Magazine. B. 260-277. 613-620. "UHF Radio Propagation Characteristics in Straight Open-groove Structures. F. VT-48. Vanderveen. D.

. Fredericton. Kannan andW. C. 1999. 159. NY. J. and time-domain analysis of electromagnetic fields. June 2003 . 2003. in 1989. Syracuse. Papadias. He received the Best Paper Award of the IEEE IEEE Antennas and Propagation Magazine.3. He is on the editorial boards of Journal of Electromagnetic Waves and Applications and Microwave and Optical Technology Letters. K. B. in 1988 and 1991. He is currently a professor at the National School of Applied Physics." Conference Record of the Thirtieth Asilomar Conference on Signals. Dr. and for' the 1997 National Radar Conference. pp. He was a teacher of Shandong University from 1991 to 1997. pp. 333-336. Kyungjung Kim was born in Seoul. His research interests are in the areas of propagation models for wireless communications. Korea. and the MS degree from Syracuse University. C." IEEE Transactions on Signal Processing. SP-46. Tetouan. Syracuse. in 1998. He was with the Rochester Institute of Technology. in 1995. Morocco. he was with the TACO Division of General Instruments Corporation. B. Harvard University. M. in2000. He is now a Professor in the Department of Electrical and Computer Engineering. France. He was elected Fellow of the IEEE in 1991. and the PhD degree from the University of Granada. Morocco. 158. C. Rochester. March 2003. His current research interests deal with numerical solutions of operator equations arising in eIectromagnetics and signal processing with application to system design. Vanderveen. and the Chancellor's Citation for Excellence in Research in 1998 at Syracuse University. Vanderveen. S. Dr. He obtained one of the "best solution" awards in May.157. "Joint Angle and Delay Estimation (JADE) for Signals in Multipath Environments. MA. Van der Veen. China. 1977. Syracuse.. 3. 161. He has been a US National Committee representative to many URSI General Assemblies. Sarkar. He was a research student from 1989 to 1995 in the Department of Applied Physics. March 1998. from 1977 to 1978. al. Fitzgerald. and A. He received the title Docteur Honoris Causa from Universite Blaise Pascal. No. India. including most recently Iterative and Self Adaptive Finite-Elements in Electromagnetic Modeling and Wavelet Applications in Engineering Electromagnetics. Currently. 45. He was a Research Assistant at Syracuse University from 1998 to 2001. 2501254. M. Kolundzija. B. M. eNg. He is currently working towards the PhD degree in the Department of Electrical Engineering at Syracuse University. University of Granada. respectively." in Proceedings of Signal Processing Advances in Wireless Communications. 2. He has published more than 210 journal articles and numerous conference papers. He has authored or coauthored ten books. Korea. He was a Research Fellow at the Gordon McKay Laboratory. A. 1996. and has been a Graduate Fellow since 2001. Tangier. in 1971. John Wiley and Sons. Sarkar et. New York in 1975. from 1976 to 1985. From 1975 to 1976. and the Friends of Clermont Ferrand award in 2000. in 1969. J. Smart Antennas. and T. T. 2000. J. Systems and Computers. He was a Research Scientist at the National Research Center of Morocco from 1996 to 2001. K. He received the BS degree in Physics from University Abdelmalek Essaadi. His research interests include sensor 81 Vol. WIPLD: Electromagnetic Modeling of Composite Metallic and Dielectric Structures. Norwood. He was the Chair of the URSI Inter-Commission Working Group on TimeDomain Metrology (1990-1996). He received the College of Engineering Research Award in 1996. "Joint Angle and Delay Estimation' (iADE) for Fading Multipath Signals in a -stable noise. NY. Spain. His book on Smart Antennas will be published by John Wiley in March. he is a Distinguished Lecturer for the IEEE Antennas and Propagation Society. Sarkar is a member of Sigma Xi and USNCfURSI Commissions A and B. respectively by Artech House. Transactions on Electromagnetic Compatibility in 1979. and has written chapters in 28 books. Paulraj. He received the PhD degree from Shanghai Jiao Tong University. Abdellatif Medouri was born in Tetouan. China. Syracuse. He received the BS degree from Inha University. "Estimation of Multipath Parameters in Wireless Communications. 1965. He is currently a Research Associate with Syracuse University. Ognjanovic. Clermont Ferrand. Spain. 160. Cambridge. Sarkar is a registered Professional Engineer in the State of New York. Morocco. the MScE degree from the University of New Brunswick. NY. New York. 682-690. and the MS and PhD degrees from Syracuse University. NY. pp. Canada. on December 11. Introducing the Feature Article Authors Zhong Ji received both the BS and MS degrees in Electronic Engineering from Shandong University.' Kharagpur. at the Rome Air Development Center (RADC) Spectral Estimation Workshop. Artech House. B. Tapan Kumar Sarkar received the BTecb degree from the Indian Institute of Technology. Paulraj. which were published in 1998 and 2002. MA. He was an Associate Editor for Feature Articles of the IEEE Antennas and Propagation Society Newsletter. Syracuse University. and A. and he was the Technical Program Chairman for the 1988 IEEE International Symposium on Antennas and Propagation and USNCIURSI National Radio Scieilbe Meeting. His current research interests include digital signal processing related to adaptive antenna problems and wavelet transforms.

For now. I hope I never have to. I'm going to take the liberty of repeating most of the comments I made in receiving the award. through the years and now. I plan not to. along with information about the other AP-S and IEEE awards presented there to AP-S members. She has served as Vice Chair and Chair of the IEEE MTT/AP-S Spanish Joint Chapter.. Thank you for the award . in books. Editor's Comments Continued from page 40 implementation and further discussion with IEEE IT staff. in part. I was very honored and grateful to receive the AP-S Outstanding Service Award for being Editor of the Newsletter and Editor-in-Chief of the Magazine. She has authored a total of 10 contributions for chapters and articles. the wording of the citation was very nice and somewhat more detailed.array signal processing. She has received two individual research awards from national institutions. E_~~ 1llIIlllImlrmlllllllllUItIiIlUnlillUlllllllltllll~mllllUlmll1l1l1 Thank You! At the Columbus Symposium. European. Actually. June 2003 . Systems and Radiocommunications Department) at the Escuela Tecnica Superior de Ingenieros de Telecomunicacion of the same university. some of them in the framework of European Community Programs. on their behalf. One of the most rewarding aspects of editing the Magazine is that I get to work with a large number of very wonderful people. Thank you! It also includes our Staff. a dedicated group of professionals who have given selflessly of their time and talents. She has been a member of the Technical Program Committee of several international symposiums. plus 'a number of national publications and reports. to all of you. This is beginning to be as serious a problem as the spam. Without them there also would be no Magazine. and measurements of hybrid and monolithic microwave integrated circuits. You'll have to decide for yourself whether or not you want the IEEE to implement such filtering on your e-mail. and computational and numerical methods for microwave structures. and they must share this award: I accepted it. Dan Senese told me that the use of this spam-Iabeling software would be offered to IEEE alias-account users only on an "opt-in" basis. now you '1/ be able to make the decision for yourself. because it's important that all of those who have been such an important part of creating and sustaining these publications be recognized. She received the PhD degree in Ingenierode Telecomunicaci6n from the Universidad Politecnica de Madrid (Spain). circuit networks and filter theory. Vol. 15 papers in international journals. She has taught courses on electromagnetic field theory. Without them. but also of principle. and numerical methods for electromagnetic field problems. 45. Thanks. My thanks to them . implementation. if you use an IEEE e-mail alias. and national institutions and companies. Finally. and antenna analysis. She has participated in 19 projects and contracts. At least. I would never turn the final decision about the disposition of an e-mail message so labeled over to software: the potential consequence of a "false positive" is too great. support. and 75 papers in international symposiums and workshops. Her research within the Grupo de Microondas y Radar (Microwave and Radar Group) is in the areas of electromagnetic field theory. at least. This includes the many AP-S members who contribute what we publish. Your support is what has made the Magazine possible. and for making the decision they did. One of the best things about this award is that I get to keep on editing the Magazine. our readers. optimization. even though I've tried sending them via two e-mail accounts with totally different ISPs. She is a member of the editorial board of two scientific journals. analog and digital communication systems theory.and to all of you. Because my opposition to this is not just a matter of practicality. financed by international. identification of radar targets. She has assisted the Comision Interministerial de Ciencia y Tecnologia (National Board of Research) in the evaluation of projects. itself Magdalena Salazar-Palma was born in Granada. in Allan Schell's report of the Awards and Fellow Committee in this issue. She has acted in the past as topical Editor for the disk of references of the Review of Radio Science. too. we have no Magazine. She has lectured in several short courses. She has also served in several evaluation panels of the Commission of the European Communities. as well as related laboratories. . simulation. A final note: As this is going to press. where she is a Professor Titular of the Departamento de Sefiales. design. If I were ever to do so. passive components. and highorder statistics. microwave and antenna theory.and for the wonderful support of our AdComs and Officers and members. 3. No. I salute and thank Dan Senese and the IEEE staff for their willingness to listen. 'and has acted as reviewer for international scientific journals. I think that's an excellent solution. I wouldn't be able to edit the Magazine without the constant love. and giving up of time I would otherwise spend with them by my wife and daughter. and is currently serving as Chair of the Spain Section of the IEEE. in letting me continue to do this job! 82 IEEE Antennas and Propagation Magazine. again. Sistemas y Radiocomunicaciones (Signals. and network and filter theory and design. you can read it. I have yet another situation in which my e-mails to an author are being rejected as spam. Spain.

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.

scribd