`Bar-Ness
`
`USOO6137785A
`Patent Number:
`11
`(45) Date of Patent:
`
`6,137,785
`Oct. 24, 2000
`
`54 WIRELESS MOBILE STATION RECEIVER
`STRUCTURE WITH SMARTANTENNA
`
`75 Inventor: Yeheskel Bar-Ness, Marlboro, N.J.
`73 Assignee: New Jersey Institute of Technology,
`Newark, N.J.
`
`21 Appl. No.: 09/042,948
`22 Filed:
`Mar 17, 1998
`51
`Int. Cl. ........................... H04J 3/14
`52 U.S. Cl. ............................. 370/328; 375/346; 455/63
`58 Field of Search ..................................... 375/346, 347,
`375/348,349; 455/67.3, 67.1, 63,524,
`525, 501, 517, 575, 132; 370/252, 241,
`334, 328, 342, 343, 345, 310
`References Cited
`U.S. PATENT DOCUMENTS
`
`56)
`
`5,634,199 5/1997 Gerlach et al. ........................... 455/63
`5,659,584 8/1997 Uesugi et al. .......................... 375/347
`5,819,168 10/1998 Golden et al. .......................... 455/303
`5,905,721 5/1999 Liu et al. ............................ 370/342
`6,006,110 12/1999 Raleigh ................................... 455/561
`Primary Examiner Huy D. Vu
`Attorney, Agent, or Firm Woodbridge & Associates, P.C.;
`Richard C. Woodbridge; Stuart H. Nissim
`57
`ABSTRACT
`-
`This invention relates to a System and method utilizing a
`receiver architecture with a Set of at least two antennae
`followed by a Rake demodulator at a mobile station for
`interference cancellation and diversity combining. Such a
`structure can work well only when the channel vector of
`desired signal is correctly estimated. The present invention
`makes use of the identifying spreading codes (as in IS-95 for
`example) to provide an adaptive channel vector estimate, to
`thereby cancel cochannel interference and improve the SyS
`tem capacity.
`
`5,471,647 11/1995 Gerlach et al. ........................... 455/63
`
`11 Claims, 2 Drawing Sheets
`
`12
`
`-- - - - - - 16
`
`O
`
`
`
`ANTENNA 2V
`
`
`
`ANTENNA V
`
`
`
`PARALLEL
`DEMODULATORS
`
`ERICSSON v. UNILOC
`Ex. 1034 / Page 1 of 6
`
`
`
`6Sts
`
` SUOLY
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`01-‘914
`
`(ym)soo|/\TYNNGINY
`/\¢YNNSINY
`
`U.S. Patent
`U.S. Patent
`
`Oct. 24, 2000
`Oct. 24, 2000
`
`Sheet 1 of 2
`Sheet 1 of 2
`
`6,137,785
`6,137,785
`
`TATWed1
`
`INGOWIO
`
`
`
`
`
`
`
`
`G2a(nigLS
`
`
`
`- - is as as as as a
`
`
`
`
`JONSUI4IY
`
`TWNOIS
`
`6
`
`ERICSSONv. UNILOC
`Ex. 1034 / Page 2 of 6
`
`ERICSSON v. UNILOC
`Ex. 1034 / Page 2 of 6
`
`
`
`
`
`U.S. Patent
`
`Oct. 24, 2000
`
`Sheet 2 of 2
`
`6,137,785
`
`FIG. 3
`
`O
`
`9
`
`8
`
`7
`
`SINR (DB)
`AT OUTPUT G
`OF RAKE
`
`5 r - - - - - - - CTO- - - - - - - -
`
`o-o ONE ANTENNA
`4-rrrr - TWO ANTENNAS, USING MRC
`: - SMART ANTENNA
`;
`:
`I
`2
`4
`6
`8
`10
`12
`14
`16
`ITERATION NUMBER: SYMBOL INTERVAL (Ts)
`
`3
`
`O
`
`18
`
`20
`
`FIG. 4
`
`
`
`III 1.
`0-0 ONE ANTENNA
`- TWO ANTENNAS, USING MRCr1-r
`2- SMART ANTENNA
`/
`------------------
`:::::::::::::::::::::::::::::::::::::::::::: Z.
`1
`
`1O
`
`
`
`PROBABILLITY OF ric for 1 or 1.
`BITERROR
`103-:::::::::::::::::::::::
`
`O
`
`1O
`
`2O
`
`30
`
`40
`
`50
`
`GO
`
`NUMBER OF USERS PER CALL
`
`ERICSSON v. UNILOC
`Ex. 1034 / Page 3 of 6
`
`
`
`1
`WIRELESS MOBILE STATION RECEIVER
`STRUCTURE WITH SMARTANTENNA
`
`BACKGROUND OF THE INVENTION
`1. Field of the Invention
`This invention relates to a System and method of inter
`ference cancellation for use with respect to a mobile Station
`having a Smart antenna and a Rake demodulator.
`2. Description of Related Art
`The capacity of wireless Code Division Multiple Access
`(CDMA) systems in the forward link direction (i.e., from a
`base station to a mobile receiver) is limited by both intra-cell
`and inter-cell cochannel interferences. In particular, when
`the mobile unit is close to a cell boundary, the desired signal
`is disturbed by relatively strong interference from neighbor
`ing base Stations. Antenna arrays have been previously
`Suggested for base Stations of CDMA Systems to improve
`the capacity in the reverse link through space diversity and
`interference cancellation. LeSS attention is given to the
`forward link due to the reliance on Orthogonal spreading
`codes to handle the cochannel interference. The perfor
`mance in the forward link is limited, however, by multipath
`fading and inter-cell interference.
`In IS-95 CDMA, signals from the same base station and
`same path are separated by a set of orthogonal codes (Walsh
`codes), which eliminate the interference of other users
`Signals in the same signal path from the home cell (see, for
`example, J. D. Gibson, The Mobile Communications
`Handbook, Boca Raton, Fla., CRC Press, Inc., 1996 and T.
`S. Rappaport, Wireless Communications. Principles and
`Practice, Upper Saddle River, N.J., Prentice Hall PTR,
`1996). The other signal paths from a home base station,
`however, create Self-interference. In addition, Signals from
`different base Stations are identified by a Special short
`pseudo random code. Such base Stations share the same
`short code, but with different shifts, and hence due to
`nonzero autocorrelation there exists inter-cell interference.
`The Worst case occurs at the cell boundary point, where the
`desired signal is the weakest and the inter-cell interference
`is the Strongest.
`Similar to its use in the reverse link, an antenna array at
`the mobile Station can be used as a diversity combiner to
`maximize the Signal-to-interference plus noise ratio (SINR).
`Due to packaging and cost considerations, however, Such an
`array needs to be Small. A dual antenna mobile Station for
`wireleSS communications has been Suggested and its imple
`mentation was studied in a paper by M. Lefevre, M. A.
`Jensen, and M. D. Rice, (“Indoor measurements of handset
`dual-antenna diversity performance,” in IEEE 47" Vehicular
`Technology Conference Proceedings, (Phoenix, Ariz.), pp.
`1763–1767, May 1997).
`The preferred embodiment of the present invention relates
`to a receiver with a two-element array, referred to as a Smart
`antenna receiver. Adaptive arrays are employed to utilize the
`known direction of arrival and Signal waveform Structure of
`desired signal for interference cancellation in point-to-point
`communication (see, for example, S. P. Applebaum and D.
`J. Chapman, “Adaptive Arrays With Main Beam
`Constrains,” IEEE Trans. Antennas Propagat., vol. 24, pp.
`650–662, September 1976 and J. R. T. Compton, “An
`Adaptive Array in Speed-Spectrum Communications,”
`Proc. IEEE, vol. 66, pp. 289-298, March 1978), wherein by
`using the pointing vector, the desired signal is co-phased and
`removed, prior to the application of weighting for interfer
`ence cancellation. To reduce Sensitivity to pointing vector
`error in a point-to-point communication application, a Self
`
`15
`
`25
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`6,137,785
`
`2
`correcting loop was Suggested to minimize the error by Y.
`Bar-Ness and F. Haber (“Self-Correcting Interference Can
`celling Processor for Point-to-Point Communications,” in
`Proceedings of the 24" Midwest Symposium on Circuit and
`Systems, (Albuquerque, N.Mex.), pp. 663-665, June 1981).
`In multi-user wireleSS and other similar communication
`applications, Such a pointing vector is not well defined, and
`hence cannot be used easily or accurately used by a mobile
`receiver for interference cancellation.
`
`SUMMARY OF THE INVENTION
`Briefly described the invention comprises a novel receiver
`Structure for a mobile Station that Simultaneously estimates
`the desired signal channel vector and adaptively controls the
`weights, thereby increasing the SINR at array output. Adap
`tive control continually corrects the channel Vector making
`it a more meaningful parameter than the prior art's pointing
`vector in modeling the received signal at mobile Stations of
`cellular CDMA systems.
`In a typical IS-95 CDMA mobile station receiver, for
`example, multiple paths are weighted and combined by a
`Rake demodulator to combat Small-scale fading (ref. Gibson
`Supra and A. J. Viterbi, CDMA Principles of Spread Spec
`trum Communication, Reading, Mass.: Addison-Wesley
`Publishing Company, 1995). In the current invention, the use
`of a Small antenna Structure along with Rake demodulator
`provides higher SINR at the output for symbol detection,
`and improves capacity in the forward link.
`These and other features of the invention will be more
`fully understood by reference to the following drawings.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`FIG. 1 illustrates the Smart antenna structure for BPSK
`demodulation at a mobile Station according to the preferred
`embodiment of the present invention.
`FIG. 2 illustrates a Rake demodulator, according to the
`preferred embodiment of the present invention, wherein the
`demodulator has L parallel Rake fingers for the jth user.
`FIG. 3 is a graph of the SINR output versus iteration
`number for three different antenna Systems.
`FIG. 4 is a graph of the probability of bit error versus the
`number of users per cell for three different antenna Systems.
`
`DETAILED DESCRIPTION OF THE
`INVENTION
`During the course of this description, like numbers will be
`used to identify like elements according to different figures
`which illustrate the invention.
`This invention (10) relates to the receiver structure and
`the Signal model used in the analysis of the proposed
`receiver Scheme. The disclosed Smart antenna receiver Sys
`tem of the present invention results in a dramatic capacity
`improvement over the receivers of the above disclosed prior
`art.
`Several assumptions are made in the development of the
`Signal model. According to the preferred embodiment, the
`receiver consists of a Small antenna array with two elements
`at a mobile station of a wireless cellular CDMA system. The
`disclosure assumes that there are one home base Station
`(n=0) and N neighboring base stations, wherein each n (n=0,
`1,...,N) base station Serves J, active users. The signal from
`each base Station is composed of J, users information
`waveforms and one pilot waveform. There are L resolvable
`paths for each Signal. The Signal from different paths and
`
`ERICSSON v. UNILOC
`Ex. 1034 / Page 4 of 6
`
`
`
`3
`different base Stations are assumed to independently undergo
`Rayleigh fading, while the J-1 waveforms that arrive from
`the same path and the same base Station at a given mobile
`receiver propagate over the same fading characteristics.
`According to the above model, the complex envelope of
`received signal at two antenna elements of the mobile Station
`is given by a 2x1 vector r(t):
`
`W
`
`L.
`
`ro-XXII)
`y P. d, ( – ta)u, ( – ta) + VP, u, ( – t} + w(t)
`
`where c, (t) represents complex channel vectors, d(t) are
`transmitted information bits, P., and P are, respectively,
`the received powers of the users Signals and pilot Signals,
`u(t) and u(t) are the spreading codes, V(t) is AWGN, t,
`are delays, and n=o refers to desired home base Station. The
`following assumptions are adopted in the analysis:
`communication performance is examined for user 1 in cell
`0, for which the channel vector can be written as:
`c'=cacal 1, . . . , L), where the SuperScript
`denotes transpose.
`slow fading is assumed for Signals from all paths and all
`base Stations.
`FIG. 1 shows the dual antenna receiver at mobile station.
`After down conversation (12) and matched filtering (14)
`(matched to the transmitting pulse), the signals received at
`two antennas are demodulated by L. parallel demodulators
`(16) (Rake fingers). The output of L Rake fingers are
`combined for a symbol detection. In FIG. 2, all the param
`eters which are referred to are used in ith Rake finger, hence
`the path index is ignored. When the weight S=c/c2, the
`desired signal is co-phase combined at y, and blocked at Z.
`That is, the Z signal output of antenna hybrid (18) consists
`of only interference as no desired signal is present. In this
`case, the weight S is used to estimate interference, which is
`subtracted from y, and higher SINR can be obtained at the
`array outputy and demodulated output b,C). The weight S
`can be updated, for example, by Direct Matrix Inversion
`(DMI) (20).
`When Szc/c2, the channel vector estimate is
`erroneous, and S will not result in a null difference between
`the desired Signals received at points X and r. Consequently,
`the residual desired signal contributions at Z will be inter
`preted by the array as interference, and hence cancelled. This
`results in performance degradation of the canceller
`(reference J. R. T. Compton, “Pointing accuracy and
`dynamic range in steered beam array," IEEE Trans. Aero
`space and Electronic Systems, vol. 16, pp. 280-287, May
`1980). To overcome this effect, the preferred embodiment of
`the present invention uses the Spreading code of the desired
`Signal. AS shown in the preferred embodiment depicting in
`FIG. 2, the processor at the mobile Station, using the
`Spreading code parameters and the correlator 22, despreads
`the array output y(t) (marked y) to yield y. This resulting
`Signal, y, is then accumulated over one Symbol interval and
`that result, g is then respread by correlator 24 using the
`Same respreading code to get a reference Signal g. When the
`kth Symbol of desired signal is received, the control Signal
`h(k) is generated by accumulating the multiplication of g(k,
`t) and Z(k, t) over one symbol interval, and h(k) is obtained
`Similarly from the reference Signal g(k, t) and the received
`
`6,137,785
`
`4
`Signal at the first element. These two control Signals can be
`expressed as
`
`5
`
`15
`
`25
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`where N is the number of code chips per bit, and the
`SuperScript “*” denotes complex conjugate operation. In
`FIG. 2 items labeled 25 denote accumulators. The weights
`is then updated by complex weight controller (34) as:
`
`It can be shown that when a pure reference Signal is
`available at g, this algorithm gives an estimate of the channel
`vector error between two antenna elements and the weight
`S=c/c. Since y has a higher SINR than the array input,
`the matched filter (30) (matched to the channel attenuation
`and phase delay) estimated from y is more accurate than the
`one estimated from the array input r. The array output y is
`then despread using correlator 32 using the pilot and jth
`users Sequences to generate the Ith Rake finger's output
`b().
`Based on these assumptions and analysis, Simulation
`results were obtained in light of the following additional
`assumptions. The Signal employed the same short code as in
`IS-95. It was also assumed that 20% of total transmitted
`power from each base Station is used for pilot. Three paths
`for each Signal are present, the relative delay between paths
`from Same base stations is two chips. In each of the Rake
`fingers, there are one desired Signal path from home base
`Station, two interfering paths from home base Station (Self
`interference), and three interfering paths from each of neigh
`boring base Stations. For comparison, three receiver models
`were examined:
`1. One antenna followed by Rake demodulator
`2. Two antennas with maximum ratio combining (MRS)
`followed by Rake demodulator
`3. The Smart antenna of the preferred embodiment, fol
`lowed by the Rake demodulator
`In the data depicted in FIG. 3, 20 active users per cell is
`assumed, with the curves obtained from 1000 Monte Carlo
`runs. For receiverS 1 and 2, the curves are also the average
`output SINR over bits 1 to 20. For a receiver with Smart
`antenna, from bit 1 to 5, the initial beam Steering weight
`s=1, the output SINR is averaged from bit 1 to 5. Starting
`from bit 6, the Smart antenna uses the algorithm of the
`preferred embodiment to control the weight S, and the
`output SINR shown in FIG. 3 is averaged over bits 6 to 20.
`The curves show that after the weight (S) correction starts,
`the receiver with the Smart antenna of the preferred embodi
`ment achieved 1.5 dB and 3.5 dB higher output SINR
`compared to receiverS 1 and 2, respectively.
`To See the capacity improvement due to proposed receiver
`of the preferred embodiment, FIG. 4 gives the curves of
`probability of bit error, P., versus number of users per cell.
`For performance requirement P=10, the System capacity
`is 24, 37 and 50 users per cell for receivers 1, 2 and 3,
`respectively. With the Smart antenna of the preferred
`embodiment at the mobile Station, the System capacity
`increases 108% and 35% compared to receivers 1 and 2,
`respectively. The proposed receiver Structure of the pre
`ferred embodiment, therefore, can provide improved capac
`ity over conventional receiverS 1 and 2.
`
`ERICSSON v. UNILOC
`Ex. 1034 / Page 5 of 6
`
`
`
`S
`While the invention has been described with reference to
`the above preferred embodiment thereof, it will be appreci
`ated by those of ordinary skill in the art that various
`modifications can be made to the Structure and function of
`the individual parts of the System without departing from the
`Sprit and Scope of the invention as a whole. In particular, the
`receiver model can be easily extended to the case of QPSK
`demodulation (rather than the BPSK demodulation depicted
`in FIGS. 1 and 2) as well as to the cases of Time Domain
`Multiple Access (TDMA) and Frequency Domain Multiple
`Access (FDMA) and to usage of polarization instead of
`Spatial information in implementing the interference cancel
`lation.
`What is claimed is:
`1. An apparatus for reducing neighboring base Station
`interference in a wireleSS mobile Station which receives a
`Signal transmitted from a home base Station, Said apparatus
`comprising:
`(a) at least two antennas;
`(b) a means for correction that continually corrects a
`channel vector model of the received signal by utilizing
`an antenna hybrid device and a pilot Signal received
`from the home base Station in order to process the
`output signal of each antenna;
`(c) an interference cancellation means for using the cor
`rected channel Vector model to adaptively control can
`cellation weights,
`(d) a despreading means for despreading a signal obtained
`from the output of the antenna hybrid device;
`(e) an accumulator means for accumulating the output of
`the despreading means over one symbol interval; and,
`(f) a respreading means to respread the accumulator
`results to yield a reference Signal.
`2. The apparatus of claim 1 wherein Said correction means
`further comprises a control Signal generating means for
`generating a control Signal using Said reference Signal.
`
`25
`
`35
`
`6,137,785
`
`15
`
`6
`3. The apparatus of claim 2 wherein Said cancellation
`means comprises a complex weight controller using Said
`control Signal to update Said cancellation weights.
`4. The apparatus of claim3 wherein Said correction means
`comprises a means for utilizing a spread code transmitted in
`a IS-95 CDMA signaling system.
`5. The apparatus of claim 3 wherein the received signal is
`a BPSK signal.
`6. The apparatus of claim 3 wherein the received Signal is
`a QPSK signal.
`7. The apparatus of claim 3 wherein the received signal is
`a Time Domain Multiple Access (TDMA) signal.
`8. The apparatus of claim 3 wherein the received signal is
`a Frequency Domain Multiple Access (FDMA) signal.
`9. The apparatus of claim 3 wherein the interference
`cancellation means comprises a means for utilizing spatial
`information.
`10. The apparatus of claim 3 wherein the interference
`cancellation means comprises a means for utilizing polar
`ization information.
`11. An apparatus for reducing neighboring base Station
`interference in a wireleSS mobile Station receiving a signal
`transmitted from a home base Station, Said apparatus com
`prising:
`(a) at least two antennas;
`(b) means for using a pilot signal received from the home
`base Station and an antenna hybrid device to continu
`ally correct a channel Vector model of the received
`Signal; and,
`(c) an interference cancellation means for using the cor
`rected channel vector model to adaptively control can
`cellation weights.
`
`ERICSSON v. UNILOC
`Ex. 1034 / Page 6 of 6
`
`