`ANALOGUE MICROWAVE RADIO NETWORK
`
`F. J. Witt , W. T. Barnett , J. D. Hubbard
`
`*
`
`A
`
`i
`
`ABSTRACT
`With the emergence of new high capacity digital services, new modern digital networks are
`necessary to satisfy customer applications. However, the task is to provide digital connectivity
`that gives the best performance at the least cost. One way is accomplish this is to retain as
`much analogue equipment as possible. The TD-90 system, described here, utilizes existing long
`distance analogue microwave radio and economically converts it to high performance, high
`capacity, digital radio by application of innovative technology and careful route planning.
`
`INTRODUCTION
`The TD-90 microwave radio system is a high performance, 9OMb/s, 64 QAM, digital
`transmission system created by modifying and augmenting already in-place TD-2 and TD-3D
`analogue radio systems. Both economical implementation and excellent performance have been
`achieved by utilizing a system architecture that is based upon reuse of existing radio equipment
`while incorporating both the latest 64QAM technology and new performance features.
`In any digital network the use of microwave technology complements the use of lightwave
`technology. Microwave technology provides quality transport on backbone routes with modest
`growth since microwave capacity can be implemented as needed. The cost of microwave (being
`primarily in the electronics) can be spread out over several years. Implementing only the
`capacity that is needed allows a constant high fill over the entire life of the facility.
`The reuse of analogue pieces is an economical means for providing high capacity digital
`transmission capability when in-place analogue radio already exists. AT&T Communications has
`already deployed 3100 microwave radio stations, Figure 1. These stations carry microwave
`traffic utilizing 4, 6, and 11 GHz common carrier frequency allocations, and are geographically
`
`FIGURE 1 THE ATaT COMMUNICATIONS
`4GHz ANALOGUtE NETWORK
`
`* AT&T Bell Laboratories, North Andover, MA
`' AT&T Bell Laboratories, Holmdel, NJ
`" AT&T Communications, Bedminster, NJ
`
`53
`
`PMC Exhibit 2045
`Apple v. PMC
`IPR2016-00755
`Page 1
`
`
`
`dispersed throughout the country. Thus application of the TD-90 system provides an effective
`vehicle for rapid expansion of AT&Ts digital connectivity throughout the USA by converting
`selected portions of their 50,000 route mile in-place analogue radio network.
`The new TD-90 performance features include automatic transmitter power control (ATPC), FM
`interference cancelation, and digital-monitoring-based maintenance at non-regenerative repeater
`locations. ATPC allows the system to be introduced into a congested interference environment.
`It controls TD-90s interference into other systems while helping the reduce the effect of
`interference on TD-90 during fading. The use of FM interference cancelers increase the TD-90
`signal to FM interference ratio substantially.
`Digital-monitoring-based maintenance at non-
`regenerative repeater locations allows maintenance decisions to be made remotely that are based
`on actual digital performance. This saves time and improves overall performance.
`This paper provides a technology overview, a description of the new performance features,
`route and performance engineering considerations, and reports on the excellent performance of
`the initial 500 mile system which has been in service since April 1986.
`TD-90 SYSTEM DEVELOPMENT
`The design philosophy of the TD-90 Radio System is to reuse as much of the radio station
`hardware now used for analogue transmission as possible. Thus the station antennas, tower,
`waveguide, and channel separation networks are all reused. The radio T/R bays may be field
`converted for digital use by the addition of a single shelf of equipment and other minor
`modifications. The converted TAR bay contains new digital equipment as well as the preexisting
`analogue equipment, Figure 2. The digital terminal and regeneration equipment is a part of
`AT&T's new family of 64 QAM digital radio equipmentEl'
`The system is designed in such a way that the eleven working radio channels may be used for
`either analogue or digital transmission. The protection channel is shared and protects all
`channels.
`A system block diagram of an eight-hop switching section is shown in Figure 3. Regeneration
`of the digital signal is necessary after four hops. Maintenance is centralized using telemetry
`processors at the ends of a switching section, which communicate with a maintenance center
`over voice frequency lines. In order to isolate troubles, remotely switchable digital receivers
`are installed at repeater sites which do not contain unnecessary regenerators. This arrangement
`avoids costly unnecessary deployment of regenerators.
`TD-90 TECHNOLOGY
`In order to convert the analogue radio channels of the AT&T 4 GHz network for digital
`service, several key technological developments were required. Some of these had been used
`earlier, including adaptive IF slope and baseband transversal equalization, space diversity and
`In addition, the TD-90 System employs
`predistortion.
`These will not be described here.
`ATPC, forward error correction and an FM carrier interference canceler.
`To meet linearity requirements, all radio bay circuits are operated at reduced signal levels
`during unfaded conditions. During either a down fade or an up fade the transmitter output
`power level is increased or decreased, respectively by ATPC. ATPC results in a system with
`optimal transmitted power during periods of normal propagation; degradations are avoided due
`to either excessive thermal noise during down fades or receiver non-linearity during upfades[21
`Forward error correction is a feature included in the digital terminals of the TD-90 System?13
`Through its use, background errors are virtually eliminated. For radio link error rates (BER)
`th.p3signal after error correction is essentially error free (significantly better than
`of 10
`BER = 10-
`). Improvements are not as dramatic for higher BER, however, significant systet
`fade performance mpr%vement is still obtained. For example, for uncorrected BER = 10-
`the corrected BER= 10
`Since the TD-90 Digital Radio System is being introduced into a nationwide sea of 4GHz FM-
`analogue signals, there are instances when FM cochannel interference is limiting, especially at
`This problem is substantialy reduced through the use of FM carrier
`junction stations.
`interference cancelers. The canceler is designed to provide fade margin improvement in the
`digital link when the FM receiver of the interfering signal is colocated with the digital receiver.
`A sample of the FM signal is used to cancel the interference through dynmic amplitude and
`phase optimization. In a typical 30 dB fade situation, if the signal to interference ratio is 31 dB
`
`54
`
`PMC Exhibit 2045
`Apple v. PMC
`IPR2016-00755
`Page 2
`
`
`
`without the canceler, it is improved to 50 dB with the canceler. The canceler is effective over a
`very narrow band, and cancels only the FM carrier. However, since the FM sidebands are
`typically 19 dB below the carrier (for 1800 circuit loading), substantial benefit accrues by
`canceling only the carrier.
`PERFORMANCE ENGINEERING
`The typical TD-90 Route Engineering
`4.
`Figure
`in
`schedule
`shown
`is
`64
`a
`engineering
`of
`Performance
`requires
`route
`QAM microwave
`detailed information of the operating
`environment. Since TD-90 is a retrofit
`technology, it also requires accurate
`reusable
`the
`about
`information
`the
`to
`addition
`in
`pieceparts
`environnment. For this reason, the first
`Route Engineering activity shown is
`This procedure
`Route Qualification.
`consists of two types of tests. The
`characterize
`the
`first
`to
`are
`tests
`analogue radio bays and thus ensure
`their suitability for TD-90 use. If the
`bays are not suitable, proper action
`can be taken early in the process to
`facilitate their upgrade to the necessary
`in Route
`level.
`The second tests
`Qualification are transmission tests to
`quantify both the transfer function of
`the antenna system/transmission path
`environment.
`the
`interference
`and
`Cancelers can be ordered, antenna
`blinders can be installed, and space
`if
`added
`be
`any
`diversity
`can
`deficiencies are discovered early in the
`That is the key to quality
`process.
`information
`The
`performance.
`obtained in Route Qualification is then
`used to engineer the route.
`The major detriment to digital radio is
`fading[41
`the
`Knowing
`multipath
`allows
`environment
`interference
`detailed outage calculations to be made
`and the proper amount of space and
`frequency diversity can be added using
`the concept of the composite fade
`margin351 [6]
`[7] After the route is
`engineered, the system can be ordered,
`the route can be upgraded as needed,
`and
`installed,
`be
`the
`system
`can
`begin.
`A
`finally,
`service
`can
`transmission outage, occurs when jhe
`bit-error-rate (BER) exceeds 10-
`The availability objective for long haul
`traditionally been an
`has
`facilities
`annual all cause, two way transmission
`outage of less than .02% of the time
`on a 4000 mi. route. The distance
`prorated, one-way objective for outage
`.8 seconds/year/mile.
`Routes are
`is
`meet this
`engineered
`typically
`to
`objective.
`
`-
`
`FIGURE 2 TYPICAL TD ANALOGUE RADIO
`CONVERTED TO TD90
`
`55
`
`PMC Exhibit 2045
`Apple v. PMC
`IPR2016-00755
`Page 3
`
`
`
`TERMINAL
`
`TERMINAL
`
`I
`
`I
`\ ~~I I
`
`PROCESSOR \
`El
`m
`MAINTENANCE
`CENTER
`
`I
`IlAl
`l
`l
`Il/
`I
`I
`I
`VOICE FREQUENCY LINE
`R-RADI0 BAY
`DR-DIGITAL RECEIVER
`FIGURE 3 TD90 ROUTE BLOCK DIAGRAM
`
`I REGEI
`
`I
`
`I
`
`X'TELEMETRY
`/ PROCESSOR
`aI]
`C
`MAINTENANCE
`CENTER
`
`ROUTE
`ENGINEERED
`
`IROUTE UPGRADE
`AS REQUIRED
`
`SERVICE
`
`-12
`
`+ +4-8
`SYSTEM
`ORDERED
`ROUTE QUALIFIED
`-RFI QUANTIFIED
`-EQUIPMENT CHARACTERIZED
`
`0 MONTHS
`
`-4
`SYSTEM**
`INSTALLED
`PRESERVICE TESTS
`COMPLETED
`
`FIGURE 4 ROUTE ENGINEERING SCHEDULE
`The use of cancelers and ATPC fights the effect of interference. The benefits of cancelers has
`In digital radio,
`been discussed above, but the benefits of ATPC need further elaboration.
`interference is an outage mechanism only during fading, when the signal to interference ratio is
`degraded. However the ATPC increases the system power during this time, keeping the signal
`to interference ratio fixed. Proper hop engineering requires maximum power to be transmitted
`when the system encounters fades sufficient to cause outage. Thus the ATPC system is a
`system that is transmitting maximum power when it counts. Because of this, when coordinating
`the interference into a system equipped with ATPC, the ATPC maximum power should be
`used. Thus a benefit of more than 10 dB is achieved when coordinating interference into a
`system equipped with ATPC. When coordinating interference from a system equipped with
`This is because when the hop is properly
`ATPC, the nominal power should be used.
`engineered, the amount of time that the power differs from nominal power conforms to the
`existing long and short term interference objectives. Thus the use of ATPC produces a win/win
`situation for both fixed power and ATPC systems, they both reap the benefits of a better
`interference environment.
`
`56
`
`PMC Exhibit 2045
`Apple v. PMC
`IPR2016-00755
`Page 4
`
`
`
`INITIAL APPLICATION
`The first installation of TD90 was turned up for service in April 1986. The total route was
`from White Plains, New York to Cleveland, Ohio, Figure 5. This amounts to a 750 route
`miles, and is a 10 switch section implementation, with some sections having 4 hops before
`regeneration.
`Beginning in November 1985, Spindle Hill to Albany was put into service as a first office
`application. This four hop switch section was then tested until the middle of January. A DS-3
`circuit (45 MB/s) circuit was fed in Spindle Hill with a pseudo-random digital signal and was
`monitored in Albany. An identical arrangement was in,lementeg in the reverse direction.
`This amounts to a 93 mile circuit. The total number of 10- and 10- BER seconds before error
`correction from midnight to 8:00 AM from November 7, 1986 to January 17, 1986 is shown in
`Figure 6. This is for one of four of the DS-3s on the route, the others exhibited similar or
`better performance.
`In terms of total errored seconds, there were only 13 errored seconds which amounts to better
`than 99.993% error free seconds. Of the 13 errored seconds, twelve were severely errored
`(BER>10- ). In total 87% of the days monitored were error free. This easily meets all digital
`service objectives for error free seconds.
`
`PINNACLE
`
`I!
`
`CAN
`
`O5FARNHAM
`IPLEY
`
`NY
`
`CHERRY
`VALLEY
`
`PA
`
`_~
`CLEVELAND
`
`OH
`
`i
`
`FIGURE 5 TD90 FIRST INSTALLATION
`
`TD90 FOA PERFORMANCE
`11-7-85 - 1- 17-86
`
`TOTAL TEST DAYS
`ERROR-FREE DAYS
`DAYS WITH 10-6 SECONDS
`DAYS WITH 10-3 SECONDS
`l0o6 SECONDS
`lo 3 SECONDS
`% ERROR-FREE SECONDS
`
`62
`54 (87%)
`8
`8
`13
`12
`99.9998
`
`FIGURE 6 SUMMARY OF PERFORMANCE ON FIRST
`TD90 SECTION AVAILABLE FOR TESTING
`
`57
`
`PMC Exhibit 2045
`Apple v. PMC
`IPR2016-00755
`Page 5
`
`
`
`CONCLUSIONS
`Analogue radio can be converted to digital microwave radio. This has been accomplished with
`TD-2 and TD-3D. The resulting system incorporates new performance features that provide
`state-of-the-art interference control and maintenance features. The performance of the system
`on the first 1000 mile route has been monitored and is meets the error free second objective.
`This technology is currently being used to provide customers with high quality digital services.
`
`References
`1. C. P. Bates, W. G. Robinson, III, and M. A. Skinner, "DR 6-30-135 System Design and
`Application," Proceedings of GLOBECOM-84, pp. 539-546, December 1984.
`2. A. J. Giger, "Transmitter Power Control Circuit," U. S. Patent No. 4,495.648, Issued Jan. 22,
`1985
`3. G. D, Martin, "Optimal Self Orthogonal Convolutional Codes with an Application to Digital
`Radio," Proceedings of the International Conference on Communications 1985, pp. 1249-1253, June
`1985.
`4. W. T. Barnett, "Multipath Propagation at 4, 6, and 11 GHz," BSTJ, pp.321-361, Vol. 51, February
`1972.
`5. A. Vigants, "Space Diversity Engineering," BSTJ, Vol. 54, No. 1, pp. 103-142, Jan 1975.
`6. H. Kostal, "A General Model For Frequency-Diversity Protection In Microwave Radio,"
`Proceedings of the International Conference On Communications, pp. 982-986, June 1985.
`7. W. D. Rummler, "A Comparison of Calculated and Observed Performance of Digital Radio in the
`Presence of Interference," IEEE Trans. Commun., pp. 1693-1700, Vol. COM-30, July 1982.
`
`58
`
`PMC Exhibit 2045
`Apple v. PMC
`IPR2016-00755
`Page 6
`
`