Professional Documents
Culture Documents
The vendor data in this Technical Brief derive from public documents available from the
Internet. Motorola provides this vendor data for Canopy users’ information only and does
not provide any recommendations.
www.motorola.com/canopy
Table of Contents
1.0 Abstract...................................................................................................................................... 5
2.4 Using Canopy for Site Planning to avoid DFS Events ......................................................... 12
3.1.2 Canopy 5.4 GHz Band OFDM 10 MHz Channel Plan ...................................................... 19
Appendix C References................................................................................................................. 32
Figure 11 ABCD channel plan for 5.4 OFDM; avoids weather notch. 20
In particular, Motorola 5.2 GHz and 5.4 GHz Canopy, a Broadband Wireless Access
(BWA) technology intended for the fixed and nomadic (“portable”) market, are impacted
by all of these DFS requirements. Fortunately the 5.4 GHz band is very large, namely
255 MHz, wider than 5.2 and 5.7 GHz bands together. The ability to move to alternate
channels within the 5.4 GHz band helps greatly to mitigate DFS issues that may occur.
In addition, nations outside of North America have DFS requirements in the 5.7 GHz
band. Canopy BWA products intended for those nations are DFS capable.
The DFS rules of the FCC and of the EU are not identical, but are close and are intended
to meet the same purpose, namely to avoid interference with certain radar signals.
However, in general only those radars profiled (pulse repetition rate and pulse width) with
a received signal strength above -62 dBm are a concern.
Note that this Deployment Guide concerns the general topics of 5.4 GHz deployment and
DFS and is not intended to be a User Guide supplement; please consult the Canopy
System User Guide for explicit information on how to configure Canopy DFS.
The FCC U-NII band devices are unlicensed intentional radiators that operate in the
frequency bands 5.15-5.35 GHz and 5.47-5.825 GHz. Specifically, the new FCC rules
require that U-NII devices operating in the 5.25-5.35 GHz and 5.47-5.725 GHz bands
employ DFS.
Note that 5.15-5.25 GHz and 5.725-5.825 GHz are not subject to DFS in the USA and
Canada.
Specifically the FCC issued “orders” for DFS by revising Part 2 and Part 15 of its
regulatory code. Since the initial issuance the rules for DFS have been changed slightly
and clarified. Consequently we reference this document:
The relevant Canadian DFS requirement is stated in document RSS-210 Issue 7, June
2007.
DFS is a feature that monitors the spectrum and selects for operation a frequency
channel that is not already in use. Prior to the start of any transmission, a U-NII device
equipped with DFS capability must continually monitor the radio environment for
presence of radar signals. If the U-NII device determines that a radar is present it must
either select another channel or enter a “lockout mode” if alternate channels are not
available; at that time the device then enters a one minute “listen mode”.
For equipment operating in the 5.25-5.35 GHz band, the FCC requires all manufactures
who apply for equipment certification of U-NII devices filed on or after July 20, 2006 to
comply with the U-NII rules for DFS. Operators using equipment operating in the 5.25-
5,35 GHz band must comply with the DFS requirements for that equipment imported or
marketed on or after July 20, 2007.
All U-NII equipment operating in the 5.47-5.725 GHz band must comply with the FCC
regulations for DFS.
In general only those radars profiled by the FCC (pulse repetition rate and pulse width)
with a received signal strength above -62 dBm are a concern.
The FCC DFS rules are applicable only to the Canopy Access Point (AP) and Backhaul
Master (BHM) whereas other nations require direct involvement of the Canopy SM and
BHS devices. In effect, the FCC DFS rules are somewhat simpler to implement.
TPC applies only to U-NII devices having an EIRP exceeding 500 mW (27 dBm) and is
intended maintain the EIRP even when a higher gain antenna is configured.
It is useful to know that Motorola Canopy products are also TPC compliant; they have the
capability of reducing Tx power by manual configuration.
2. Remember that in the “FCC states” only the AP and BHM devices are active in
DFS detection whereas the “ETSI states” additionally require DFS detection for
the SM and BHS. This may influence your deployment plan.
3. Run the tests in several directions and in as many channel centers as possible;
each test period should last several days. Note “problem sectors” and “problem
channel centers” over a week’s time.
6. Alternate channel frequencies should be at least one channel bandwidth from the
primary frequency and from each other, if possible. Remember that there is a
“Uniform Channel Spreading” requirement. This simply means to use channels
throughout the band.
7. Test and note that DFS “hits” can be direction-specific. DFS “hits” can affect one
AP (with the same frequency) without affecting its opposite mate on the cluster.
In this case, you may want to change the frequency of one AP, but not the other.
8. Even after careful site surveying and channel planning, it is possible to encounter
“random” DFS detections. If so, this means that service in the impacted AP
section is down for the required DFS “listen time” and then each SM (in the
sector) must scan and re-register with the AP on the chosen alternate channel.
In total, a particular subscriber may experience a service outage longer than the
one minute listen time. Experience suggests about 1.5 minutes; this depends
upon the number of SMs in the impacted sector.
Choosing alternate channels at least one channel bandwidth distance from your
primary channel increases your chances to resume transmission more quickly.
9. In those countries (“ETSI states”) requiring that the SM/BHS also actively listen
for DFS detection prior to operation, the service downtime will be 2 minutes at
minimum, with 3 minutes likely in SM-dense sectors.
10. For any FSK Canopy units, enable “schedule whitening”* to avoid self-
interference and reduce the possibility of “false positive” DFS detections.
Schedule whitening is not applicable to Canopy’s OFDM products.
11. In general site Canopy base stations away from the Terminal Doppler Weather
Radar (TDWR) sites hosted by major U.S. airports. Being located below the
horizon (10 or 20 miles from the radar site) is usually sufficient. As a rule, the
TDWR sites are located within 10 miles of the host airport. (See the list of U.S.
airports hosting TDWR sites in the Appendix.)
13. Many Canopy customers state that in their experience the 5.4 GHz band is
relatively quiet and “clean” compared to the 5.7 GHz band, and that they plan
further 5.4 GHz band deployments.
14. Consider using unique color codes to maintain the original AP to SM relationship
should a DFS channel change occur if that is important in your network.
15. Always update Canopy software to the latest version as it is further optimized for
DFS and additional fixes and features.
16. If experiencing DFS events try downtilting or re-orienting the AP to avoid the
radar’s beam..
A significant number of radar devices in the C-band do not fall within the FCC’s radar
profile. Therefore, they should not trigger a DFS event. For example, the FAA’s air traffic
control radar operates in a different band and does not fit within the FCC’s profile and
should not trigger a DFS event.
In the USA, the national government operates about 510 radar sites for weather and air
traffic control (excludes military) which can be categorized into seven types: ASR-9, ASR-
11, ARSR-1/2, ARSR-3, ARSR-4, NEXRAD, and TDWR. Of these seven types only the
FAA’s Terminal Doppler Weather Radar (TDWR) is in the C-band, with about 45 sites
deployed near the largest airports (representing 1.5% of all US airports)..
The Federal Aviation Administration (FAA) indicates that all TDWR use is in the 5.6-5.65
GHz range. The FAA site referenced in the Appendix indicates that no other FAA
aviation frequencies impact Canopy’s unlicensed bands. All the remaining government
radars (excludes military) operate in the S-band and so do not produce DFS events.
However, privately owned weather radars (e.g. television stations), do exist in the US
using the C-band. It is difficult to identify these sites or specifically characterize their
radar signal. From all indications, these also primarily operate in the 5.6-5.65 GHz “sub-
band” as this band appears to be particularly suited for weather detection.
Weather radar like TDWR has a very narrow “pencil” beam, typically less that 1° in both
horizontal and vertical. It typically rotates in about 5 to 10 seconds and “paints” a vertical
helix pattern, covering the desired “volume” over several minutes. Consequently, given
the earth’s curvature, it is likely that TDWR is not a 5.4 GHz band issue when the Canopy
equipment is at least 20 miles from the TDWR site (which is typically ~ 10 miles from the
airport such that the site is outside of the airport’s “ground clutter”).
1
Recall that the electromagnetic spectrum is divided into letter-coded bands, with the C-band being 4 cm to 8
cm wavelength (approximately 3.9 to 6.2 GHz). Source: http://www.jneuhaus.com/fccindex/letter.html).
Figure 2 shows a map of TDWR sites (circa 2001) and a picture of a TDWR tower
complex. A list of 44 US host airports is included in the Appendix E. Note that this list
could change over time should TDWR grow in popularity.
All government and private weather radar operation is licensed by the FCC.
Military Radar
The US military also operates a number of C-band radars which could trigger a DFS
event. Most of these military sites are fixed but some mobile units do exist including
airborne radar. This radar is not used frequently and may be used on occasion for military
base training.
Some military radar is of the “frequency hopping” type, in which the radar’s operating
frequency is changed frequently and in a random pattern. Note that this means the BWA
operator can’t necessarily avoid such a DFS event merely by channel frequency planning
if they are operating a DFS-enabled unit within ~ 20 miles of the radar’s site.
Maritime Radar
Although common operating procedures suggest that vessels turn off their radar when in
port, this is not done consistently – by both military and commercial vessels.
Radar equipment may be used while in port for routine maintenance, servicing or repair
work or to stabilize the unit before leaving port. Such vessels may trigger a DFS event
while in port or within ~ 20 miles of the DFS-enabled devices.
It is common practice for short stay vessels to put radars onto “standby” while in port. No
radar emissions occur while on “standby.”
Most radar is configured for a particular frequency within the radar’s band, and with a
particular pulse width and particular pulse repetition rate. These settings can be re-
configured by the operator as needed but it is not known whether this is a real concern.
Outside of the US the radar situation is much the same. A big difference is that certain
countries, such as Canada, Australia, and the European Union countries, have deployed
C-band based weather radar nationally. All indications suggest that the 5.60 to 5.65 GHz
sub-band is also commonly used internationally for weather detection. However, other
frequencies might also be licensed for this use.
For example, Canada and Australia employ C-band weather radar and so choose to
prohibit the private use of the 5.60-5.65 GHz sub-band. Consequently BWA vendor
equipment must “notch out” this particular 50 MHz sub-band from any commercial use.
Canopy products automatically account for this when an operator chooses a Regional
Setting.
Figure 3 shows the locations of these Canadian weather radar sites; note that range of
these extend into USA territory where use of 5.60-5.65 GHz is not prohibited explicitly
(but may trigger the FCC’s DFS rules). Canopy operators located sufficiently close to
USA’s national borders must comply with any Canadian regulation regarding DFS and
channel notching.
A more complete description of the Canadian weather radar network can be found at
http://en.wikipedia.org/wiki/Canadian_weather_radar_network.
Several technologies are not a DFS issue. For example, police traffic radar uses
frequencies well above Canopy’s unlicensed bands. Microwave ovens operate in the
unlicensed 2.4 GHz band.
Motorola strongly recommends adequate spectrum analysis and channel planning prior
to deployment to ensure the best operating environment. This is discussed in greater
detail below.
Although the current Canopy spectrum analyzer is not intended to detect the energy in a
very short radar pulse, it may. Both Motorola and customer testing indicate that it, in fact,
does detect much of the radar energy present.
Figure 4 shows a Canopy 5.4 GHz FSK spectrum analyzer screen. Notice that the
spectrum captured ranges from 5.47 to 5.85 GHz meaning both the 5.4 GHz band and
5.7 GHz band. A Canopy 5.4 GHz product is limited to 5.47 to 5.725 GHz (channel
centers from 5.495 to 5.705 GHz).
While not explicitly designed to capture radar-type signals the spectrum analyzer seems
to indicate maximum powers.
For example, in Figure 4, the frequencies from 5.605 to 5.625 GHz all show the red ticks
at the extreme right side indicating strong reception. These frequencies are typical of
current C-band weather radar operation. Note these high power ticks are shown even if
the “bar” establishing them is not.
By employing the Canopy spectrum analyzer, an operator may conclude that any
additional 5.4 GHz band Canopy deployment should avoid these two 5.4 GHz sub-bands:
Recall that even this limitation leaves still much more available bandwidth than the entire
5.7 GHz band (125 MHz).
Of course, the safest approach to understand the potential for radar interference in the
5.2 GHz and 5.4 GHz bands is simply to deploy an AP in the intended operating area to
determine whether actual DFS detection is experienced.
To that end, Motorola recommends purchasing a DFS-band Access Point (AP) and
deploying it in the intended site(s), in multiple directions and on multiple channel
frequencies, over a period of time. A period of one week is strongly suggested.
This pre-service AP deployment can log DFS events by direction and channel frequency
and so, after some period of time (e.g. one week) can provide the Canopy operator with a
level of confidence concerning the tested channel frequencies and sites.
Of course even this testing is not definitive as the DFS environment is subject to change
over time, but this pre-service testing is a low cost and effective way to determine the
viability of reliable service in the 5.2 GHz or 5.4 GHz bands.
In this section Canopy’s implementation of DFS is described at a high level. This DFS
description is applicable to both “regular” 20 MHz and 25 MHz channel single carrier
(FSK) Canopy and the OFDM-based 10 MHz channel 5.4 GHz Canopy. The only DFS
distinction between the two Canopy product lines is the narrower channel of the OFDM-
based Canopy which makes channel planning more flexible.
Canopy’s SM-Lite currently supports both the FCC and ETSI DFS rules. The operator
needs to choose the proper region code to enable the proper DFS functionality.
Currently Canopy’s DFS operation is dependent upon geographic region as national DFS
rules require. In Canopy Release 8.2 AP the region is configured in Configuration →
General → Regional Setting webpage as shown in Figure 5; dropdown menu options are:
Other, United States, Canada, Europe, Brazil, Australia, Russia, and None. Selecting the
factory default None results in a message “please select a valid region code.” An
administrator is required by local law to select the correct Regional Setting.
This Regional Setting configuration will also automatically block channels as required by
that region or nation as discussed earlier (e.g. 5.6-5.65 GHz channel frequencies in
Canada and Australia).
In general a Canopy device has the ability to “listen” for the specific radar signal profile,
and, if detected, change transmission to an alternate channel as regulation requires. The
Canopy operator must pro-actively select these alternate channels.
In FCC situations, if radar is detected and the AP’s or BHM’s primary channel frequency
is “shut down”; the set of sector SMs no longer “hear” the AP’s beacon signal so they
cease regular operation and enter a “pre-registered scan mode”. Naturally the customers
will not have service at this instant. However the SMs will soon “hear” the new beacon
signal on some alternate channel frequency (as configured on the AP by the network
operator) and so re-register automatically and so restore customer service. Note the
sector downtime is just over one minute in this case.
2
We confine discussion to the point-to-multipoint (PTMP) base station, but the same DFS rules apply to
Canopy BH10/BH20 operation (in the 5.2 and 5.4 GHz bands). Note, however, that the Canopy BHM must be
configured using the same menus associated with an AP.
Note: Eleven non-overlapping channels are available on 20 MHz centers for standard
“1X” Canopy operation.
The Canopy recommended nine channel center frequencies are shown in Figure 6. The
colored channels impinge on the 5600-5650 MHz “weather radar” notch.
A channel plan may include all of these channels or a subset if DFS testing indicates
certain channel centers should be avoided due to radar interference.
In general all channel center frequencies infringing between 5.600 GHz and 5.650 GHz
cannot be deployed in Canada and Australia. Thus only six channel centers are
recommended in Canada and Australia: 5495, 5520, 5545, 5570, 5670, and 5695 MHz.
A six-sector Canopy 5.4 GHz base station is normally configured for a 3x2 channel reuse
plan (ABCABC) using three stipulated primary channels; and leaving the other six non-
overlapping channels for DFS alternatives.
A sample Canopy 5.4GHz ABCABC channel plan is show in Figure 7 As stated above all
the nine non-overlapping 20 MHz channels on 25 MHz channel centers are used in this
plan.
In the DFS channel plan illustrated, each of the six base station APs has two DFS
alternatives, and each alternatives are distinct for each of the three primary channel
frequencies. In this case, multiple AP’s could detect a radar signal and continue
operation in an alternate channel ensuring more continuous operation.
This channel plan means that when the primary channel detects radar, it shuts down the
current frequency and enters a one-minute “listen” on the first alternate “DFS1.” By
placing this alternate as “distant” as possible from the now silenced channel frequency, it
is less likely to detect radar on the alternate channel. Consequently the customers being
served in this sector are “down” for only about one minute before the alternate frequency
is up. Note: the sector SMs are default configured to “scan” all the 5.4 GHz channel
frequencies, and so will re-register quickly to the new DFS1 alternate channel frequency.
Note also that the channel center frequency separation remains at 25 MHz in the DFS1
and DFS2 columns in Figure 7. Thus “co-channel” interference is even less likely than
among the primary channels. In Canopy FSK devices a “schedule whitening”
configuration mitigates this interference (discussed in the Best Practices section).
Note that the configuration of these alternate channel frequencies must be done
independently for each AP/BHM. Consequently great care must be taken when
configuring all the AP’s of a 5.4 GHz base station cluster.
Each of these alternates has a dropdown menu of all possible 5.4GHz channel center
frequencies. After selecting the primary channel frequency center, the two alternate
frequencies can be configured.
Many other channel plan schemes can be devised for example to account for “weather
notching” when 5600-5650 MHz must be avoided. Figure 8 illustrates a “one alternate
DFS channel” plan suitable for Canada and Australia (or elsewhere by choice). Note the
second alternate DFS is left “not applicable.”
Consequently Motorola recommends the 5.4 GHz band channel centers shown in Figure
9 for OFDM deployment:
Motorola suggests that Canopy OFDM operators do the same pre-deployment DFS radar
detection as described above. After testing is completed for all four sectors, an OFDM
channel plan can be produced for a 2x2 channel reuse as shown in Figure 10.
Each channel is offset 60 MHz to avoid any co-channel interference. Note also this
particular OFDM 10 MHz channel plan avoids the weather radar notch so is applicable
worldwide.
In general, with a 2x2 channel reuse plan and 10 MHz channel width on 10 MHz channel
centers, the 5.4 GHz band (of 255 MHz width) has great flexibility. The particular channel
plan illustrated in Figure 10 is but one of many that could have been devised.
Figure 11 ABCD channel plan for 5.4 OFDM; avoids weather notch
Alternately a Canopy operator may opt for an ABCD channel plan; this avoids having any
potential “back to back” self-interference issues. Figure 11 shows such a channel plan.
Note how the channel centers are spaced such that, even if a DFS event forces channel
switching, the adjacent sectors are sufficiently far apart.
A good practice is to select channel center frequencies that are spread over the complete
band thereby satisfying the FCC need for “uniform channel spreading.”
All of these standards are automatically configured for Canopy products by choosing the
correct Regional Setting.
This section emphasizes the differences in Canopy DFS operation in the EU/UK and
operation in the US.
Primarily this difference consists of the Canopy SM and BHS playing an active – rather
than passive – role in DFS operation.
Additionally a primary and two alternate channel frequencies must be configured, again
as per the Canopy FCC DFS configuration.
The EU/UK SM is different in that, like the AP, the SM actively “listens” for a radar signal.
Note the ramification is that an AP sector could be operating on a particular channel
frequency (because its particular AP does not detect radar from its location) but one or
more of its sector SMs does detect radar on that same channel frequency.
In the EU/UK the SM scans through its configured list of scan frequencies; these SM
frequencies should correspond to those configured DFS channel frequencies configured
into its AP.
However, prior to “listening” for an AP’s beacon, the SM must enter a one-minute “listen”
mode to determine if a radar signal is present on that particular channel. If radar is not
detected after one minute the SM attempts to register (by transmitting).
If, instead, radar is detected then that particular channel must enter a 30-minute “lock
out” period. At that point the SM can immediately switch to another one of its scanned
frequency set, and listen one minute there. If radar is not detected then the SM will listen
for an AP beacon and, if found, register and enter a normal operational state with the AP
and SM now operating on the same alternate channel.
If radar is detected on its alternate channel, then the SM “locks out” this channel and
proceeds to scan again into its pre-selected frequency list.
Caveat: It is important to note that in the rather unlikely event that the SM (or BHS)
experiences a DFS event and the AP (or BHM) does not, the SM (or BHS) will have
switched to an alternate channel and the AP (or BHM) would still be operating on its
original primary channel. If that happens, the SM (or BHS) would not be able to register
to its AP (or BHM). In this case, the particular SM would have to wait to complete its 30-
minute lock-out time and then re-register to the AP’s primary channel (assuming that this
channel was radar-free). Important: Given that the AP (or BHM) and associated SM’s (or
BHS) are most commonly located within 2 miles of each other, it is somewhat unlikely
that this scenario would occur.*
In “FCC nations” the DFS detection and recovery times are as follows:
This time interval is due to the mandatory 1 minute “listening” time (on the DFS alternate
channel frequency) followed by the approximate scanning and re-registration time of the
impacted SMs and BHS.
Note if an ETSI AP or BHM detects a DFS event, it must listen for 1 minute before
moving channels and must additionally have the SM or BHS listen for 1 independent
minute.
These are likely recovery times for 5.4 GHz Canopy equipment in ETSI DFS regions:
• 2.5 minutes for the SM to resume after its AP has a DFS detection
• 3.5 minutes for the BH link to resume after the BHM has a DFS detection
(requires two channel availability checks)
*Recovery times for ETSI regions are as follows should an SM or BHS experience a DFS
event and the AP or BHM does not as described at the end of Section 3.2:
• 32.5 minutes for the BH link to resume after the BHS has a DFS detection
Canopy’s DFS detection is in effect a “sampling” mechanism and as such has an innate
“failure” probability. As an operator, you would like to avoid any “false positive”
detections in order to maximize system performance.
Schedule whitening (discussed earlier) can help to avoid self-interference and “false
positive” triggers caused by other Canopy equipment.
Motorola has also attempted to “design out” error possibilities and so reduce “false
positives” to close to “zero” and continues to reduce this risk overtime through software
upgrades. Therefore, operators are strongly encouraged to upgrade to the latest version
of software available. Motorola advises Canopy operators that it is impossible to
eliminate “false positive” detections altogether and that all BWA equipment vendors are
subject to this caveat.
Prizm 3.1 supports Release 8.2 features such as DFS. In this section an example of
Prizm’s ability to inform a Canopy operator of DFS detection events is illustrated. Details
are fully described in the Prizm Users Guide.
The Canopy devices Management Information Bases (MIBs) now include the additional
DFS-related information. This allows the Prizm-based management of DFS remotely via
Simple Network Management Protocol (SNMP).
In particular Canopy and Prizm have four DFS-related MIB entries, namely:
• dfsStatus
• dfsStatusAlt1
• dfsStatusAlt2
• dsfStatusPrimary.
• whispRadarDetected
• whispRadarEnd
In particular the use of Prizm (using SNMP) allows the operator to be alerted to DFS
events. This allows the operator to see a complete DFS log for all Canopy equipment
triggering DFS events so avoids the need to “visit” every Canopy device individually via
its web GUI interface. Prizm displays a complete log of all past DFS events, including the
shift to alternative channel frequencies, each completely time and date stamped.
Figure 12 shows a screenshot of a Prizm sequence of DFS Alerts. Note the date and
time stamp and the channel frequencies (in MHz). Also note that the most current Alert is
Importantly SNMP traps are said to be asynchronous, that is they are unilaterally sent
from the Canopy device to the Prizm server (without a Prizm server initiated “poll” of the
devices). Therefore, the operator can receive an e-mail alert that a DFS event has
occurred in “real time”.
4.0 Summary
With proper channel planning, site surveys and configuration, DFS events should be rare
to non-existent.
Based on Motorola’s experience and that of its customers, we can offer the “best
practices and considerations” outlined in Section 2.2.
The impact of DFS can be minimized by a “pre-deployment” check of the 5.4 GHz bands
by purchasing one FSK-based SM and utilizing its spectrum analyzer capability.
Additionally, purchasing an appropriate Canopy AP and probing for DFS events (as per
Section 2.4) can further help you determine the likelihood of a successful Canopy
deployment in the 5.4 GHz band.
The Motorola Canopy product line is compliant with all known DFS regulatory
requirements in all applicable nations. Canopy’s Prizm EMS supports configuring
Canopy devices remotely via SNMP for proper DFS operation and event management.
From a regulatory standpoint, DFS rules in the unlicensed spectrum are an operational
fact. Fortunately the 5.4 GHz band is wider than the 5.2 GHz and 5.7 GHz bands
combined which offers operators many options when navigating DFS compliance. This
provides unlicensed BWA operators the business opportunity to profit as never before in
newly available, uncongested spectrum.
Master Mode: Operating mode in which the U-NII device has the capability to transmit
without receiving an external control signal and can perform Network Initiation.
Client Mode: Operating mode in which the transmissions of the U-NII device are under
control of the Master Device. A U-NII device operating in Client Mode is not able to
initiate a network.
Channel Availability Check Time: The period of time during which a Channel
Availability Check is performed.
Channel Move Time: The time to cease all transmissions on the current Channel
upon detection of a Radar Waveform above the DFS Detection Threshold.
In-Service Monitoring: A DFS function that monitors the Operating Channel for
the presence of a Radar Waveform above the DFS Detection Threshold.
Network Initiation: The process by which the Master Device sends control
signals to Client Device(s) that allow them to begin transmissions.
Non-Occupancy Period: The time during which a Channel will not be utilized
after a Radar Waveform is detected on that Channel.
Applicable Canopy APs, SMs, BHMs, and BHSs are, in operation, both a Master Station
and a Slave Station by the definitions above.
a) The Master Device will use DFS in order to detect Radar Waveforms with
received signal strength above the DFS Detection Threshold in the 5250 - 5350
MHz and 5470 - 5725 MHz bands. DFS is not required in the 5150 − 5250 MHz
or 5725 − 5825 MHz bands.
c) The Master Device initiates a U-NII network by transmitting control signals that
will enable other U-NII devices to Associate with the Master Device.
d) During normal operation, the Master Device will monitor the Channel (In-
Service Monitoring) to ensure that there is no radar system operating on the
Channel, using DFS described under a).
f) Once the Master Device has detected a Radar Waveform it will not utilize the
Channel for the duration of the Non-Occupancy Period.
a) A Client Device will not transmit before having received appropriate control
signals from a Master Device.
b) A Client Device will stop all its transmissions whenever instructed by a Master
Device to which it is associated and will meet the Channel Move Time and
Channel Closing Transmission Time requirements. The Client Device will not
resume any transmissions until it has again received control signals from a
Master Device.
• For the Short Pulse Radar Test Signals this instant is the end of the Burst.
• For the Frequency Hopping radar Test Signal, this instant is the end of
the last radar Burst generated.
• For the Long Pulse Radar Test Signal this instant is the end of the 12
second period defining the Radar Waveform.
Note 3: During the U-NII Detection Bandwidth detection test, radar type 1 is used
and for each frequency step the minimum percentage of detection is 90 percent.
Measurements are performed with no data traffic.
The FCC requires DFS detection for four types of Short Pulse radar waveforms (types 1-
4), one type of Long Pulse radar waveform (type 5), and one type of Frequency Hopping
radar waveform (type 6). These are numbered respectively in Table 3 below.
Testing requires the Pulse Widths be stepped randomly in 1µs intervals, along with other
stipulations not necessary to this overview.
We remind the reader that any detection scheme is subject to two kinds of errors:
1. False positive: Detecting a signal when that signal is not present.
Note in the first case the device’s radar interference “falsely” continues when it should be
terminated. In the second case the device’s operation is “falsely” terminated when no
radar interference is occurring.
The FCC’s and EU’s rules concerning these error occurrences are met by Canopy.
1 1 1428
2 1-5 150-230
3 6-10 200-500
4 11-20 200-500
5 50-100 1000-2000
6 1 333
Note that this description of the FCC DFS rules only stipulate that a device detects the
prescribed radar signals and what must then be accomplished. The FCC does not
specify how or even what a vendor should do in order to implement the requirement.
Consequently Canopy’s current DFS deployment may not be identical to other vendors.
In addition Motorola may subsequently change its DFS deployment as experience
dictates.
A connectorized 5.4 and 5.7-GHz module provides an Antenna Gain parameter. When
you indicate the gain of your antenna in this field, the DFS algorithm (where applicable,
as in the UK) calculates the appropriate sensitivity to radar signals, and this reduces the
occurrence of false positives (wherever the antenna gain is less than the maximum).
Figure 13 provides a summary of the FCC DFS requirements. In the FCC context a
“master” is a Canopy Access Point (AP) and Backhaul Master (BHM), and a “slave” is a
Canopy Subscriber Module (SM) and Backhaul Slave (BHS).
In the European Union (EU) the European Technical Standards Institute (ETSI) standard
EN 301 893 V1.2.3, the EU’s harmonized radio standard for unlicensed devices operating
in the 5150 – 5350 MHz and 5470 – 5725 MHz frequency bands, was the first standard to
require DFS. Canopy will support the newer ETSI 1.3.1 DFS specification as required
after March 2009.
Figure 14 shows a summary of the ETSI DFS requirements. Again the “master” is a
Canopy AP (and BHM) and a “slave” is a Canopy SM (and BHS). This is Table 5 in the
ETSI EN 301893 document stipulating DFS. In the ETSI context a Canopy SM is a
“slave with radar detection” by necessity and design.
Additionally the UK Interface Requirement IR2007 requires DFS in the 5.7 GHz band as
well. Norway requires 5.7 GHz band DFS as defined by Annex 1 of ITU-R
Recommendation M.1652 / EN 301 893. Brazil requires 5.4 GHz band DFS. Other
relevant documents are issued by: WRC-03, ITU-R Recommendation M.1652, National
Telecommunications and Information Administration (NTIA).
“ETSI states”: Armenia, Austria, Azerbaijan, Belgium, Belize, Bolivia, Brazil, Bulgaria,
Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Georgia, Germany,
Hungary, Iceland, Ireland, Italy, Liechtenstein, Lithuania, Luxembourg, Monaco,
Netherlands, New Zealand, Norway, Poland, Portugal, Slovak Republic, Slovenia, South
Africa, Spain, Sweden, Switzerland, Trinidad & Tobago, Tunisia, Turkey, United
Kingdom, and Venezuela.
This listing is not guaranteed to be correct and is subject to unilateral national change.
http://www.ieee802.org/18/Meeting_documents/2007_Nov/WFA-DFS-
Best%20Practices.pdf (weather radar background)
http://www.faa.gov/about/office_org/headquarters_offices/ato/service_units/techops/spec
_management/engineering_office/rfb.cfm (lists all spectrum used in FAA-based aviation)
Alternatively, the 5.2 GHz Canopy six-sector base station may employ a 2x2 channel
plan (ABABAB) and specify each sector switch to the alternative channel frequency C
upon radar signal detection. Motorola testing suggests that this 2x2 channel plan does
work satisfactorily.
Note that if an adjacent A and B sectors both detect a DFS event then they both switch to
the same DFS alternate; this is unworkable. However, if such a channel plan is not
deployed a DFS detection means a 30 minute sector shutdown.
Source: TDWR_SPG_ICD_v43.pdf