(12) United States Patent
`Ling et al.
`
`USOO6172970B1
`(10) Patent No.:
`US 6,172,970 B1
`(45) Date of Patent:
`Jan. 9, 2001
`
`(54) LOW-COMPLEXITY ANTENNA DIVERSITY
`RECEIVER
`(75) Inventors: Curtis Chih-Shan Ling, Kowloon
`SESSR NE. yiel,
`Water Bay (HK)
`(73) Assignee: The Hong Kong University of Science
`and Technology (HK)
`
`(*) Notice:
`
`Under 35 U.S.C. 154(b), the term of this
`patent shall be extended for 0 days.
`
`21) Appl. No.: 08/851,543
`pp
`1-1.
`May 5, 1997
`(22) Filed:
`(51) Int. Cl. .................................. H04J 3/02; H04L 1/02
`(52) U.S. Cl. ........................ 370,347, 375/347, 455,277.2
`(58) Field of Search
`370/347,321
`375/345 346,347,348,349.366:455/133.
`134,135.136. 137.13s. 136.277.1.2772
`27s. 273.276.1.3.04.101.671.672.
`67.26.1226.
`•-1s a- a
`•
`
`(56)
`
`References Cited
`U.S. PATENT DOCUMENTS
`455/138
`3,934,204 * 1/1976 Hill
`... 455/135
`4,450,585 * 5/1984 Bell ...................
`4,972,434 * 11/1990 Le Polozec et al. ................ 375/347
`
`2- - -2
`
`OZalSKI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`7/1991 Akinson et al. ..................... 375/347
`5,031,193
`4/1993 Yamao .............................. 455/277.1
`5,203,024
`5,263,180 * 11/1993 Hiramaya et al. ................... 455/139
`E. : 3.E. CN al. ...
`S.
`5,796,777 * 7/1998 Terlep et al. ........................ 455/67.1
`* cited by examiner
`Primary Examiner-Chi H. Pham
`Assistant Examiner Steven Nguyen
`(74) Attorney, Agent, or Firm-Burns, Doane, Swecker &
`Mathis, L.L.P.
`(57)
`
`ABSTRACT
`
`A low-diversitv antenna diversitv receiver Suitable for
`y
`y
`TDMA PCS handset implementation employing two diver
`sity branches. The receiver is capable of Selecting a diversity
`Scheme which is anticipated to give optimum signal recep
`tion among a plurality of diversity Schemes installed on the
`receiver. This receiver, more conveniently termed multi
`diversity receiver comprises a Single conventional wireleSS
`digital receiver chain augmented with a few additional
`low-cost passive RF components and minor control circuits.
`A plurality of diversity algorithms, for example, Selection
`diversity (SD), equal-gain combining (EGC) or interference
`reduction combining (IRC) scheme, which are suitable for
`implementing on this multi-diversity receiver are also
`described. Simulation results showing performance of this
`multi-diversity receiver are also presented.
`
`16 Claims, 5 Drawing Sheets
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`THEIRC ALCORTHM WITH TWO ANTENNAS
`(Low-FSGNAAFIERA)
`
`--
`
`a
`
`a ------
`
`
`
`INTIAL SETUP,
`= +
`(CANCEL INTERFERER)
`
`INITIAL SETUP (B1, B2, B3)
`ESTIMATE A1, A2, 6
`(FIND COPHASED STARTING POINT)
`
`ESTMATE A, ADURING
`BURSTS (B4, B5)
`
`SEARCHFOR ODURING
`BURSTS (B2, B1)
`
`COHERENT DEMODULATION (BC)
`
`NO
`(CONTINUE
`SEARCH)
`
`YES
`(SEARCH
`IS LOST)
`
`USNG
`SELECTION
`DIVERSITY
`NO
`(REVERSE PHASE
`TO CANCEL INSEAD
`OF ADD)
`
`
`
`OCKED TO
`STRONG INTERFERER)
`
`(OPTREEED DAAC
`FRAMEUSING CURRENT
`SEARCH LOCATION
`
`s - - 1
`
`CONTINUE NEXT
`
`m- an a
`
`ERICSSON v. UNILOC
`Ex. 1009 / Page 1 of 13
`
`

`

`U.S. Patent
`
`Jan. 9, 2001
`
`Sheet 1 of 5
`
`US 6,172,970 B1
`
`BLOCK DACRAM OF THE TWO-BRANCH ANTENNA MULT-DIVERSITY RECEIVER,
`
`ANT. 1
`
`FRONT END
`RF CIRCUIT
`
`FULLY-DIGITAL
`TDMA BURST
`DEMODULATOR
`
`
`
`
`
`MULT-DIVERSITY
`BASEBAND PROCESSOR
`
`
`
`PACS DOWNLINK FRAMESTRUCTURE
`
`
`
`SYNCH: SYNCHRONOUS BITS
`CRC: ERROR DETECTION
`
`SYNCHSC
`
`PC DATA
`
`CRC PC
`
`SC: SYSTEM CONTROL AND SUPERVISORY BITS
`PC: POWER CONTROL
`FIG. 2
`
`FC: FAST CHANNEL DATA
`
`ERICSSON v. UNILOC
`Ex. 1009 / Page 2 of 13
`
`

`

`U.S. Patent
`
`Jan. 9, 2001
`
`Sheet 2 of 5
`
`US 6,172,970 B1
`
`THE IRC ALCORTHM WITH TWO ANTENNAS
`
`
`
`INITIAL SETUP,
`= 0 + T
`(CANCEL INTERFERER)
`
`INITIAL SETUP (B1, B2, B3)
`ESTIMATE A1, A2, 0
`(FIND COPHASED STARTING POINT)
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`ESTMATE A1, A DURING
`BURSTS (B4, B3)
`
`SEARCH FOR O DURING
`BURSTS (B2, B1)
`
`
`
`COHERENT DEMODULATION (BO)
`
`
`
`- OUPUIRECEIVEDDAIA
`
`conut Niyi
`FRAME USING CURRENT
`SEARCH LOCATION
`
`IS LOST) 3rdPOOR.
`(LOCKED TO goE
`STRONG INTERFERER)
`PASS
`
`YES
`(SEARCH
`
`NO
`(CONTINUE
`SEARCH)
`
`USING
`SELECTION
`DIVERSITY
`NO
`(REVERSE PHASE
`TO CANCEL SEs
`OF ADD
`
`
`
`ERICSSON v. UNILOC
`Ex. 1009 / Page 3 of 13
`
`

`

`U.S. Patent
`
`Jan. 9, 2001
`
`Sheet 3 of 5
`
`US 6,172,970 B1
`
`LINK PERFORMANCE UNDER FLAT. FADING
`x NO DIVERSITY (DIFF)
`o SELECTION DIVERSITY
`+ NO DIVERSITY (COH)
`: EQUAL GAIN COMBINING
`
`
`
`10
`
`12
`
`14
`
`22
`20
`18
`16
`AVERAGE Eb/No (dB)
`FIG. 4
`
`24
`
`26
`
`28
`
`50
`
`ERICSSON v. UNILOC
`Ex. 1009 / Page 4 of 13
`
`

`

`U.S. Patent
`
`Jan. 9, 2001
`
`Sheet 4 of 5
`
`US 6,172,970 B1
`
`LINK PERFORMANCE UNDER FLAT. FADING WITH CO-CHANNEL INTERFERENCE
`
`o SELECTION DIVERSITY
`x NO DIVERSITY
`+ EQUAL GAIN COMBINING - SMART COMBINING
`
`
`
`- fo = 6HZ
`
`----- f = 1Hz
`
`5
`
`10
`
`20
`
`15
`AVERAGE SIR (dB)
`FIG. 5
`
`25
`
`50
`
`ERICSSON v. UNILOC
`Ex. 1009 / Page 5 of 13
`
`

`

`U.S. Patent
`
`Jan. 9, 2001
`
`Sheet 5 of 5
`
`US 6,172,970 B1
`
`LINK PERFORMANCE UNDER FREQUENCY-SELECTIVE FADINCE
`
`x NO DIVERSITY
`+ EQUAL GAIN COMBINING
`
`o SELECTION DIVERSITY
`le SMART COMBINING
`
`
`
`- fo = 6HZ
`
`----- f = 1HZ
`
`10-2
`
`10-1
`NORMALIZED DELAY SPREAD BY SYMBOL PERIOD
`FIG. 6
`
`100
`
`ERICSSON v. UNILOC
`Ex. 1009 / Page 6 of 13
`
`

`

`1
`LOW-COMPLEXITY ANTENNA DIVERSITY
`RECEIVER
`
`FIELD OF THE INVENTION
`The present invention relates to an antenna diversity
`receiver for radio communication Systems, and more par
`ticularly to a low-complexity antenna diversity receiver
`having implemented on a Single receiver unit a plurality of
`inter-switchable diversity schemes. Furthermore, this inven
`tion relates particularly to an antenna diversity receiver
`especially Suitable for use as a portable handset for Personal
`Communication Systems (PCS), such as Time Domain
`Multiple Access (TDMA) Communication Systems.
`BACKGROUND OF THE INVENTION
`It is well known that antenna diversity can improve the
`reception quality of communications in a wireleSS environ
`ment and yield increased System capacity. Conventionally,
`Selection diversity is the Simplest diversity Scheme which
`operates on the principle of Selecting the antenna diversity
`branch which provides the Strongest received signal level or
`the best eye-opening. However, it is known that Selection
`diversity does not provide any useful gain in a line-of-Sight
`(LOS) environment since the two branches are correlated.
`In a recent paper by Cox and Wong, “Low-Complexity
`Diversity Combining Algorithm and Circuit Architectures
`for Co-channel Interference Cancellation and Frequency
`Selective Fading Mitigation', IEEE Trans. Comm. Vol. 44,
`no 9, pp. 1107–1116, September 1996, it is shown that two
`antenna optimum-combining diversity produces a signal-to
`interference ratio (SIR) improvement of at least 3-dB over
`conventional two-antenna Selection diversity in Personal
`Access Communication Systems (PACS). This is attractive
`Since combining diversity can be applied to cancel
`co-channel interference and boost the desired signal even in
`an LOS environment.
`Qualitatively Speaking, in an LOS environment, an
`optimum-combining receiver adjusts the joint Signal of a
`plurality of antennas, resulting in an adaptive joint antenna
`pattern or polarization which attenuates co-channel interfer
`ence while amplifying the desired Signal. In a multi-path
`environment, the antennas may be receiving Signals from
`Separate paths and this picture is not entirely applicable, but
`the concept is the Same.
`While optimum-combining diversity offers attractive per
`formance improvement over Selection diversity, it is noticed
`that existing antenna diversity researches concentrate on
`Selection diversity. Such a preference is probably due to that
`fact that many of the So-called adaptive antenna array
`Solutions rely on algorithms which require well character
`ised antenna patterns. In contrast, most mobile PCS handset
`antennas possess patterns which are not carefully controlled
`and are quite dependent on the position of the antenna with
`respect to the user's hand and head. Thus, if optimum
`combining diversity is to be devised and implemented on
`mobile PCS receiver handsets, the first task would be to seek
`optimum-combining diversity algorithms which do not
`require well characterised antennas as a prerequisite.
`Hitherto, System complexity together with the associated
`power consumption, cost and Size has been a significant
`barrier to the wide-spread commercial implementation of
`diversity schemes in PCS portable handsets since most
`proposed diversity handset Schemes require one receiver
`chain for each branch of diversity which means that receiver
`circuitry from RF to baseband has to be duplicated. This dual
`receiver chain design approach is contradictory to the indus
`
`15
`
`25
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`US 6,172,970 B1
`
`2
`trial trend of circuit Simplification and consumer appetite of
`miniaturisation and cost reduction. This limitation, unless
`circumvented, would continue to hinder implementation and
`further development of diversity schemes in mobile hand
`SetS.
`In the Cox & Wong publication above, there is shown a
`Symbolic diagram, i.e. FIG. 1, which discloses the concept
`of a simplified ideal diversity receiver design in which the
`RF signal from two antenna branches are combined after
`level adjustment but before further processing presumably
`by a Single channel device for baseband processing.
`However, this disclosure merely shows a future receiver
`topology hopefully to be implemented but the underlying
`algorithm proposed in that publication does not actually
`Support implementation of a diversity receiver using Single
`channel baseband Signal processing.
`Furthermore, while Selection diversity algorithm does not
`offer Significant Signal quality improvement in the circum
`stances mentioned above, it is nevertheless very fast and
`energy efficient. In circumstances where the Signal quality
`received by one of the antennas is Superbly high, Selection
`diversity would be beneficial and it would be highly desir
`able that the Simpler Selection diversity can be chosen and
`utilised. Thus, it would be highly desirable if a diversity
`receiver handset can accommodate a number of modes of
`diversity algorithms which can be chosen according to the
`reception conditions. This would of course require the
`preSupposition that the prime constraints of low-cost, low
`complexity and low-weight are observed.
`
`SUMMARY OF THE INVENTION
`It is therefore an object of the present invention to provide
`a low-complexity antenna diversity receiver design consis
`tent with the prime design constraints, i.e. low-cost, low
`complexity and low-weight, which is particularly Suitable
`for handset implementation in PCS. To be implemented as a
`practical and Sophisticated mobile PCS handset, it would be
`appreciated that the design has to meet the following prac
`tical design constraints:-
`Firstly, the design should utilise only a single receiver
`chain and baseband combining processor together with
`Standard baseband processing techniques.
`Secondly, the only additional RF frontend components
`required are low-cost passive components for combining RF
`Signals received through a plurality of antenna branches at
`the RF front-end.
`Thirdly, the system is sufficiently robust to handle poorly
`defined, user dependent antenna patterns.
`Fourthly, the System is capable of providing Several
`modes of diversity algorithm on Single receiver without
`physically changing the hardware or baseband processing,
`and can choose the most appropriate diversity mode given
`the mobile usage and Signal environment. For convenience,
`Such a receiver would be referred hereinafter to as “multi
`diversity receiver”.
`Finally, the techniques can be applied to an increased
`number of antennas, though at the cost of decreased mobility
`and lower tolerance to fading.
`According to the present invention, there is therefore
`provided A portable receiver for time division multiplexing
`access (TDMA) personal communication Systems in which
`a wanted Signal burst and a plurality of unwanted Signal
`bursts are transmitted in a time-multiplexed manner within
`the same signal frame comprising first and Second antenna
`diversity branches, Signal combining means and Signal pro
`
`ERICSSON v. UNILOC
`Ex. 1009 / Page 7 of 13
`
`

`

`3
`cessing means, wherein each Said antenna diversity branch
`comprises a low-noise amplifier and means for Signal ampli
`tude variation and one of Said diversity branches comprises
`phase shifting means, Said Signal combining means is
`adapted to combine the Signal outputs from Said first and
`Second diversity branches before Said Signal outputs have
`undergone any frequency conversion, and Said Signal pro
`cessing means is adapted to process the Signal output from
`Said Signal combining means.
`Preferably, the receiver further comprises controlling
`means, wherein Said controlling means is adapted to control
`Said means for Signal amplitude variation and Said means for
`adjusting phase shift, the amount of amplitude to be varied
`and the phase to be shifted being dependent on the Signal
`quality (SO) of unwanted signal bursts which were respec
`tively received by said first and second diversity branches.
`Preferably, Said Signal quality is a factor indicating the
`eye-opening of the received unwanted Signal bursts and is
`preferably determined by using a Square-law Symbol timing
`Preferably, wherein Said receiver comprises means to
`Select a diversity Scheme among a plurality of diversity
`Schemes comprising Selection diversity (SD), equal-gain
`combining (EGC) and interference-reduction combining
`(IRC) algorithms.
`According to another aspect of the present invention,
`there is described a portable receiver for time division
`multiplexing access (TDMA) personal communication Sys
`tems in which a wanted Signal burst and a plurality of
`unwanted Signal bursts are transmitted in a time-multiplexed
`manner within the same Signal frame comprising first and
`Second antenna diversity branches, Signal combining means,
`Signal processing means and controlling means, wherein
`each Said antenna diversity branch comprises a low-noise
`amplifier and means for Signal amplitude variation and one
`of Said diversity branches comprises phase shifting means,
`Said Signal combining means is adapted to receive and
`combine the Signal outputs from Said first and Second
`diversity branches, Said Signal processing means is adapted
`to process the Signal output from Said Signal combining
`means, and Said controlling means is adapted to control Said
`means for Signal amplitude variation and Said means for
`adjusting phase shift, the amount of amplitude to be varied
`and the phase to be shifted being dependent on the Signal
`quality (SO) of unwanted signal bursts which were respec
`tively received by said first and second diversity branches.
`In yet another aspect of the present invention, there is also
`described an algorithm for operating an antenna diversity
`receiver comprisingi) determining the Signal quality of Said
`first and Second diversity branches using unwanted Signal
`bursts by firstly enabling said first and substantially dis
`abling said Second branch, ii) measuring the first signal
`quality of the burst received by first branch, secondly by
`enabling Said Second branch and Substantially disabling Said
`first branch, measuring the Second Signal quality of the burst
`received by Second branch; iii) comparing the signal quali
`ties thus measured against a pre-determined threshold value,
`Selecting Selection diversity if either of Said Signal qualities
`exceeds Said threshold value and Selecting combining diver
`sity if the Signal qualities of the Signals or their combination
`are below the threshold value but above a second predefined
`threshold value which corresponds to Signal which are too
`poor for demodulation, and iv) searching for better signal if
`the Signal qualities of the Signals and their combination are
`below the threshold value which corresponds to signal
`which are too poor for demodulation.
`BRIEF DESCRIPTION OF THE DRAWINGS
`The present invention will now be explained and illus
`trated in better detail by way of examples only and with
`reference to the accompanying figures, in which:-
`
`15
`
`25
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`US 6,172,970 B1
`
`4
`FIG. 1 is a block diagram Showing a two-antenna branch
`multi-diversity receiver,
`FIG. 2 shows a typical TDMA signal frame structure
`conforming to PACS standard for PCS communication,
`FIG. 3 shows a flowchart of a multi-diversity receiver
`algorithm with two antennas,
`FIG. 4 shows the simulation results for link performance
`under flat fading,
`FIG. 5 shows the simulation results for link performance
`under flat fading with co-channel interference, and
`FIG. 6 shows the simulation results for link performance
`under frequency Selective fading.
`DETAILED DESCRIPTION OF THE
`PREFERRED EMBODIMENT
`The preferred embodiment of the present invention is a
`two-antenna multi-diversity receiver System having imple
`mented a plurality of inter-Switchable diversity modes,
`including Selection diversity (“SD), equal-gain combining
`(“EGC) and interference-reduction combining (“IRC").
`The expression “multi-diversity' used in the present context
`is merely intended to indicate a receiver pedigree which is
`characterised by its ability to Select a diversity algorism or
`Scheme among a plurality of pre-installed diversity algo
`rithms or schemes. The hardware of this system is only
`marginally more complex than existing non-diversity
`receivers and its complexity is quite comparable to that of a
`receiver implementing only Selection diversity, while
`achieving performance comparable to that offered by more
`complex Systems under quasi-static multi-path channel and
`interference conditions.
`Hardware
`The multi-diversity receiver shown in FIG. 1 comprises
`two antennas, Ant.1 & Ant.2, each followed by a low-noise
`RF amplifier (LNA). Conventional antennas meeting the
`Spatial or polarization diversity requirements (un-correlated
`in a multipath environment) and conventional low-noise RF
`amplifiers meeting pre-determined performance criteria or
`technical specifications would be suitable for used. The
`amplified signals, after appropriate amplitude adjustment
`and phase shifting, for example by a pair of controllable
`variable Signal attenuators and a phase-shifter, is combined
`into a Single Signal Stream by an RF-combiner. The Signal
`stream thus combined is then processed by a front-end RF
`circuit which down-converts the RF-signal so that it can be
`processed by a demodulator and further operated on by a
`baseband processor which would in turn control the ampli
`tude attenuators and the phase shifter.
`The most noticeable extra hardware components which
`are to be added to a conventional Selection diversity receiver
`System in order to convert the same into a multi-diversity
`receiver are two RF attenuators, a RF phase-shifter, and an
`RF signal combiner. Voltage-controlled variable
`RF-attenuators having attenuation range between 0-20 dB
`and a Voltage-controlled RF phase-shifter having a shifting
`range of 0-360 degrees are selected for the present embodi
`ment for illustration purposes and convenience only. Other
`types of attenuator or shifter with the appropriate ranges can
`of course be used.
`Referring to FIG. 1, a variable RF attenuator is placed
`after the LNA in each diversity branch for relative amplitude
`Scaling, introducing a maximum possible amplitude varia
`tion of 40 dB across the two branches. The phase shifter is
`only required in one of the two diversity branches to
`introduce relative phase shifting between the Signal Streams
`in the two branches. The resulting Signals, after attenuation,
`
`ERICSSON v. UNILOC
`Ex. 1009 / Page 8 of 13
`
`

`

`25
`
`S
`phase shift or a combination thereof, are then Summed at the
`signal combiner, fed into the RF front-end circuit for down
`conversion and then to be processed by the baseband pro
`ceSSor which in turn controls the variable attenuators and the
`phase shifter via Some conventional interfaces.
`Fully-Digital TDMA Burst Demodulator
`For ease of understanding and because of the importance
`of TDMA systems in current PCS, the preferred embodiment
`of multi-diversity receiver architecture is explained with
`reference to a receiver which is compatible to and designed
`to operate under TDMA environment for PCS
`communications, such as PACS, WACS, GSM, PHS, DECT
`and other like Systems. However, it should also be appreci
`ated that the present receiver methodologies, concepts and
`topologies would equally be applicable for performance
`improvements of any PCS Systems, Such as direct-Sequence
`or frequency-hopped Systems, as well as TDMA Systems.
`For Sake of completeneSS and clarity, Some basic param
`eters of the selected TDMA environment together with a
`preferred demodulation technique are now described. The
`example transmission environment Selected for the present
`illustrative purposes is the US low-tier PACS standard using
`JL/4 DQPSK modulation, with 384 kbps channel bit rate, 120
`bits per time slot, 8 time slots (bursts) per frame, 312.5 us
`burst duration and 2.5 ms frame duration, i.e. a frame rate of
`400 Hz. The PACS signal downlink frame structure of this
`System is shown in FIG. 2. In this example implementation,
`JL/4 DQPSK modulation and square-root raised cosine
`(C=0.5) pulse shaping is used. The receiver also uses two
`Square root raised cosine (C=0.5) filters for in-phase (I) and
`quadrature (Q) baseband Signal to match the transmitter for
`optimal performance in additive white Gaussian Noise
`(AWGN) environments. It also follows that other digital
`phase modulation Systems can be treated Similarly.
`The preferred signal demodulation technique which is
`adopted in the instant System for explanation purposes is the
`fully digital coherent demodulation technique proposed by
`Chuang and Sollenberger in “Burst Coherent Demodulation
`with Combined Symbol timing, Frequency Offset
`Estimation, and Diversity Selection”, IEEE Trans. On
`40
`Communication, Vol. 39, no. 7 July 1991, and in “Low
`overhead Symbol Timing and Carrier Recovery for TDMA
`portable Radio Systems”, IEEE Trans. On Communication,
`vol. 38, no. 10, pp. 1886–92, October 1990.
`This coherent demodulation technique is unique and
`preferred because it jointly estimates both Symbol timing
`and carrier frequency offset by operating on an individual
`TDMA burst without requiring a training Sequence. These
`estimates produce a signal quality factor (SQ) measurement
`which is a good indicator of the degree of Signal impairment
`caused by noise, delay Spread or interference which closes
`the eye-opening of the detected Signals. Unlike using the
`maximum average eye-opening as Symbol timing as Sug
`gested in the above Chuang and Sollenberger paper of 1991,
`Square-law Symbol timing Scheme as proposed by J. G.
`55
`Proakis in his book “Digital Communications”,3' ed., New
`York, McGraw-Hill, 1995, is used to estimate timing
`because of its Superior performance. The values of I and Q
`thus obtained at the Sampling output are then used to
`calculate the SQ and carrier phase (cp). A novel low
`complexity diversity combining processor is added into the
`receiver to control the combining circuits. At the same time,
`the Signal Strength is measured through received signal
`strength indicator (RSSI) circuits.
`A low IF bandpass signal at 768 kHz (4 times the symbol
`rate) is sampled with an A/D converter at 3.072 MHz (4
`times the IF), resulting in an oversample of 16 Samples per
`
`45
`
`50
`
`60
`
`65
`
`US 6,172,970 B1
`
`15
`
`35
`
`6
`Symbol. This is required to achieve Symbol timing recovery,
`Signal quality measure, frequency offset estimation and
`carrier phase recovery without overhead, as Suggested in the
`Chuang and Sollenberger paper above. With the same down
`converter architecture, coherent and differential detection
`can be achieved for J/4 DOPSK.
`While frequency offset estimation is not addressed in this
`implementation, it can for example be removed either
`through a RF frequency Synthesiser or baseband frequency
`estimation. Since signal bursts for PACS and the like stan
`dards are very Short relative to channel variation, a quasi
`Static channel approximation can be assumed. Such an
`assumption means that the channel is Static during the burst
`period is realistically applied to a flat fading Study. It should
`be appreciated that the assumption of a quasi-static channel
`is used through out this description and in the Simulation
`work that follows.
`Antenna Diversity Modes
`Most PCS downlinks, including PACS, utilise continuous
`time division multiplex (TDM) transmission which is par
`ticularly known for the increase in transmission rate by time
`multiplexing data from a number of Sources. A characteristic
`feature of TDM transmission is that, within a time frame,
`there are a number of extra time-slots in addition to the
`time-slot allotted to the burst which contains the wanted
`communication burst.
`Referring to FIG. 2, each Signal frame duration is 2.5 ms
`and comprises eight slots each of 312.5 LiS duration. Each
`Such 312.5 us time slot is designated to transmit communi
`cation a data burst of 120 bits.
`Thus, up to eight communication bursts, usually all origi
`nating from different Sources, can be transmitted within a
`Single frame. ASSuming for convenience that burst B0 in the
`third time slot of the instant frame is the desired communi
`cation burst which is preceded by a plurality of un-wanted
`signal bursts, namely for example B3 & B4 from the
`previous frame and B2 & B1 from the present frame
`respectively, which precede the desired burst B0. It will
`become apparent below that these Seemingly irrelevant data
`bursts, i.e. B1-B4, can be utilised to determine the channel
`and receiver parameters and Set the diversity combining
`parameters A, A, 0, i.e., the attenuation and the phase
`shifting factors in FIG. 1.
`Upon determination of the parameters from the Seemingly
`irrelevant bursts, the appropriate diversity modes, i.e. SD,
`EGC and IRC, which is anticipated to give the best reception
`according to Some pre-determined criteria is to be Selected
`and implemented for instantaneous reception of the desired
`burst. The manner how this is done is explained below.
`In the description to follow, a general description of the
`various diversity algorithms which are applicable in a diver
`sity receiver are discussed, the symbols P, SQ, and stand
`respectively for the received signal power, Signal quality and
`carrier phase of the ith diversity branch.
`Diversity Mode I: Selection Diversity (SD mode)
`In this diversity mode, the receiver Simply Selects the
`diversity branch which has the best Signal quality for
`demodulation. Selection of the diversity branch is usually
`based on a signal quality factor (SQ) which indicates the
`quality of the received signal with reference to the Signal
`Strength or eye-opening of the received signal. AS eye
`opening is widely accepted to be the more accurate indica
`tion of Signal impairment, it will be used in the present
`embodiment.
`In this mode, the SQ of the first and second diversity
`branches is determined independently and Sequentially by
`any two preceding unwanted bursts, for example B2 & B1
`
`ERICSSON v. UNILOC
`Ex. 1009 / Page 9 of 13
`
`

`

`US 6,172,970 B1
`
`8
`Algorithm for Determining Phase Shift for Co-Phasing
`The following provides an example of the Steps which
`could be used as a reference to operate on the unwanted
`bursts B3-B1 for the co-phasing procedure.
`B3: Firstly, Ant.1 is selected and Ant.2 substantially
`disconnected by setting A=1 (0 dB), A=0.1 (20 dB
`attenuation) and 0=0. SQ, and d are calculated from the
`unwanted burst B3.
`B2: Secondly, Ant.2 is selected and Ant.1 substantially
`disconnected by setting A=0.1 (-20 dB), A=1 (0 dB) and
`0=0. SQ and d are calculated from the unwanted burst
`B2.
`B1: Thirdly, during the 312.5 us duration of burst B1, the
`four possible phase difference values, i.e. d-180, d-90, did
`and d+90, are tested and the one which yields the smallest
`combined power after phase inversion (+180°) is then the
`exact phase difference which will be used to provide phase
`shifting in the Second branch before the Signals are com
`bined.
`B0: Finally, A and A are both set to 0 dB and the phase
`shift is set equal to 0, the true phase difference. The desired
`communication burst B0 is then received and demodulated.
`Diversity Mode 3: Interference-Reduction Combining
`In a high capacity PCS, it is known that, for a given
`bandwidth, co-channel interference (CCI) limits system
`capacity. Usually, CCI is dominated by one co-channel
`interferer because of Shadowing phenomenon, which is
`known to have a log-normally distributed local mean of
`received signal power. In order to cancel the primary Source
`of CCI, it is preferable that the attenuating factors A and A
`are adjusted So that the interferences I and I from each of
`the branches are Substantially equal in amplitude. The
`adverse effect of the interference can then be substantially
`eliminated or cancelled out by out of phase addition.
`It is known, for example from the Cox and Wong paper,
`that Signal-to-Interference Ratio (SIR) and SQ are related.
`In particular, when SIR is between 7-13 dB, simulation has
`shown that there exists an approximate linear logarithmic
`relationship between SIR and SQ which is given by:
`
`1O
`
`15
`
`25
`
`35
`
`40
`
`45
`
`50
`
`7
`above. The branch having the better SQ is selected for
`demodulating the Subsequently arriving desired communi
`cation burst B0. The following provides an example of steps
`which could be used for independently obtaining the SQ and
`selecting the better antenna diversity branch in the SD mode.
`Determining the SQ of first branch:
`Firstly, the SQ of the first antenna branch, Ant.1, is
`determined by evaluating the preceding burst B2 which is
`received through Ant. 1. This is done by setting A=1 (0 dB
`attenuation), A=0.1 (20 dB attenuation) and 0=0 (no phase
`shit). The SQ of the first branch, SQ, as obtained from the
`received burst B2, is then calculated.
`Determining the SQ of second branch:
`Secondly, the SQ of the Second antenna branch, Ant.2, is
`determined by evaluating another preceding burst B2 which
`is received through Ant.2. This is done by Setting A=0.1
`(-20 dB), A=1 (0 dB) and 0=0. The SQ of the second
`branch, SQ, as obtained from the received burst B1, is then
`calculated.
`Communication burst reception:
`After SQ and SQ has been determined, the antenna
`branch having the larger or better SQ is Selected to receive
`and demodulate the desired burst, B0.
`Diversity Mode 2: Equal-Gain Combining Mode (EGC
`mode)
`Predictions and simulation results show that EGC pro
`vides performance advantage in a flat fading environment in
`which noise is the dominant and at a constant level. In this
`mode, the Signals received by the two antenna are Subject to
`equal amplification and the overall noise figure is reduced by
`having the phase of the Signals from the two branches
`equalised before Summing. EGC is useful for minimising
`noise figure in the present embodiment Since attenuators
`rather than variable gain amplifiers are used. In this mode,
`the phase of the Signals received via the two branches are
`first determined by utilising a number of unwanted bursts
`and their phases are then equalised by relative phase shift
`ing. Since A and A can be set 0 dB, the key remaining
`proceSS is then to have the Signals in the two branches
`co-phased before combining.
`Co-Phasing
`Local crystal oscillators are known to have very high
`short-term stability. It can therefore be safely assumed that
`its frequency and phase remain constant during Several
`bursts and can be used as a phase reference.
`Firstly, Ant.2 is disconnected and Ant. 1 is Selected to
`receive a preceding unwanted burst B2, and the phase d is
`recovered. Secondly, Ant.1 is disconnected and Ant. 2 is
`Selected to receive another preceding unwanted burst B2 and
`phase dba is recovered. Here, the phases d and dB have an
`ambiguity equal to an integer multiple of 90 degrees intro
`duced during the phase recovery process. This ambiguity
`causes no problems in coherent detection Since it can be
`removed by deferential decoding. However, the absolute
`phase difference between the two branches are required if
`they are to be properly co-phased.
`Now, let d=d-d (ph