ADVANCED SIGNAL PROCESSORS FOR INTERFERENCE CANCELLATION IN BASEBAND RECEIVERS

Abstract

An interference canceller comprises a composite interference vector (CIV) generator configured to produce a CIV by combining soft and/or hard estimates of interference, an interference-cancelling operator configured for generating a soft projection operator, and a soft-projection canceller configured for performing a soft projection of the received baseband signal to output an interference-cancelled signal. Weights used in the soft-projection operator are selected to maximize a post-processing SINR.

Claims

1. A method for cancelling interference from a received baseband signal in a CDMA system, comprising: generating at least one composite interference vector (CIV) by combining estimates from interfering subchannels, generating a soft-cancellation operator, and performing a soft cancellation of the received baseband signal to output an interference-cancelled signal.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Embodiments according to the present invention are understood with reference to the flow diagram of FIG. 1 and the schematic block diagrams of FIGS. 2A and 2B.

[0014] FIG. 1 is a flow diagram of an interference-cancelling method for a particular multipath component.

[0015] FIG. 2A is a schematic block diagram of a circuit configured for cancelling interference and combining interference-cancelled multipath components.

[0016] FIG. 2B is a schematic block diagram of a circuit configured for cancelling interference from at least one finger of a Rake receiver that produces a CIV from signals received by all fingers of the Rake receiver.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] The present invention will now be described more fully hereinafter with reference to the accompanying drawings, in which preferred embodiments of the invention are shown. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0018] A received baseband signal at a user handset having K base stations (or subchannels), U users, L propagation paths, and a sequence of transmitted symbols {b.sub.k[m]} can be expressed by

[00001]$y\left[n\right]=\underset{k=1}{\overset{K}{.Math.}}.Math.\underset{m=-}{\overset{}{.Math.}}.Math.\underset{l=1}{\overset{L}{.Math.}}.Math.c_{k,l}.Math.s_k\left[n-\mathrm{Nm}.Math.-d_{k,l},.Math.b_k\left[m\right]\right]+v\left[n\right]$

[0019] where {s.sub.k[n, b.sub.k[m]]} is a discrete-time symbol-bearing waveform from base station k that has N samples per symbol period, the vector sequence {b.sub.k[m]} is a sequence of U user information symbols b.sub.k[m]=[b.sub.k,l[m], . . . , b.sub.k,U[m]] from base station k, the values C.sub.k,l and d.sub.k,l are the complex channel fading coefficients and the time delays characterizing the propagation channel linking the k.sup.th base station to the receiver, and v[n] is additive noise having power .sup.2. When a multi-code (e.g., CDMA, DSSS, WCDMA, DO) transmission is employed, a transmitted waveform can be represented as

[00002]$s_k\left[n,b_k\left[m\right]\right]=\underset{u=1}{\overset{U}{.Math.}}.Math.b_{k,u}\left[m\right].Math.w_{k,u}\left[n\right],m.Math.Nn<\left(m+1\right).Math.N$

where U is the number of users, b.sub.k,u[m] is a user data symbol (which is drawn from a finite constellation and is constant over symbol intervals of sample length N), and w.sub.k,u[n] is a user spreading code (including PN, covering, and filtering), which is typically time varying at the sample rate. The sampling rate corresponding to n is taken to be the normalized rate 1 and assumed to be greater than the chip rate. The received signal y[n] may be organized into a sequence of vectors at rate 1/N

[00003]$y\left[m\right]=\underset{k=1}{\overset{K}{.Math.}}.Math.\underset{m^{}}{.Math.}.Math.\underset{l}{\overset{L}{.Math.}}.Math.c_{k,l}.Math.W_{k,l}\left[m-m^{}\right].Math.b_k\left[m^{}\right]]+v\left[m\right],$

where b.sub.k contains symbols b.sub.k,u and the columns of the matrix W.sub.k,l comprise vectors of the form

w.sub.k,l,u=[w.sub.k,l,u[mNd.sub.l], . . . ,w.sub.k,l,u[(m+1)N1d.sub.l]].sup.T

Thus, the sampling rate corresponding to m remains 1/N.

[0020] The optimal receiver for a given user information sequence depends on the cellular network's operating mode (e.g., soft handoff, blocking). For example, if a particular handset is not in handoff and there is no inter-base-station interference (i.e., K=1), the optimal detection strategy for a single symbol of interest corresponding to a designated user is

[00004]$b_u\left[m\right]=\arg .Math..Math.\underset{b}{\max }.Math.\underset{\left\{b_u^{}\left[m^{}\right]\right\}.Math.b_u\left[m\right]=b}{\max }.Math.\mathrm{Re}.Math.\underset{l}{.Math.}.Math.\stackrel{\_}{c}.Math.s_l^{*}\left[m;\left\{b\left[m^{}\right]\right\}\right].Math.\left(y\left[m\right]-\frac{1}{2}.Math.s\left[m;\left\{b\left[m^{}\right]\right\},l\right]\right)$

where overbar denotes a complex conjugate and superscript * denotes a Hermitian transpose. The term s.sub.l[m;{b[m]}] is a received signal vector, delayed by d.sub.l corresponding to the vector-valued information sequence {b[m]}, and the vector

[00005]$s\left[m;\left\{b\left[m^{}\right]\right\},l\right]=\underset{l^{}l}{.Math.}.Math.c_l,s_l,\left[m;\left\{b\left[m^{}\right]\right\}\right]$

represents an interference signal formed from all of the paths not equal to path l. This exemplary embodiment impels approximations that cancel interference terms s.sub.l[m;{b[m]}] from received signals, in advance of Rake reception (i.e., the sum over l of c.sub.ls.sub.l[m]. The vector s.sub.l[m; {b[m]}] may be expressed as

s.sub.l[m;{b[m]}]=[s[mNd.sub.l,{b[m]}], . . . ,s[(m+1)N1d.sub.l,{b[m]}]]

[0021] When the complex baseband signal y[m] is resolved at a particular (l.sup.th) finger in a handset's Rake receiver, it can be simplified to a vector representation

y=cx.sub.ub.sub.u+x.sub.MAI+x.sub.INT+v

where y represents received data after it passes through a receiver pulse-shaping filter (e.g., a root raised-cosine pulse-shaping filter). The data y is time aligned to a particular path delay. The term c is a complex attenuation corresponding to the path.

[0022] When the modulation is linear, the term x.sub.u in path l, which represents a code waveform that typically includes an orthogonal basis code and an overlaid spreading sequence (e.g., a PN code) assigned to a user of interest, may be written as

x.sub.1,l,u[m]=c.sub.1,lw.sub.1,l,ub.sub.1,u[m]

The term W.sub.1,u is the spread and scrambled code for user u in cell k=1, and b.sub.1,u is an information symbol corresponding to the user of interest. The term x.sub.MAI is multiple access interference, and it may be expressed by

[00006]$x_{1,l,\mathrm{MAI}\left[m\right]=}.Math.c_l.Math.\underset{u^{}u}{.Math.}.Math.w_{1,l,u}.Math.b_{1,u}\left[m\right].$

The term x.sub.INT may include inter-finger (and possibly inter-base-station) interference terms that are similar in form to x.sub.MAI. The term v is a vector of complex additive noise terms. Each of the vectors x.sub.u, x.sub.MAI, and x.sub.INT is a signal resolved onto a Rake finger matched to the l.sup.th multipath delay of base station k at symbol period m.

[0023] A conventional Rake receiver resolves the measurement x.sub.u onto a user's code vector to form the statistic x.sub.u*y.sub.l. Such statistics are typically derived from multiple Rake fingers and coherently combined across the paths via a maximum ratio combiner (i.e. they are weighted by the conjugate of the channel gains and summed). Alternatively, more general combining may be used.

[0024] FIG. 1 illustrates a signal processing method in accordance with an exemplary embodiment of the invention that is configured to reduce ISI in a received signal from a particular Rake finger. A CIV s is generated 101 by combining soft or hard estimates of interference corresponding to the other delays and/or base stations not tracked by the particular finger. For example, the soft estimates may correspond to interfering user subchannels from each base station tracked by a cellular handset. Soft or hard estimates may be derived from a conventional Rake receiver, an equalizer, or any detector matched to the communication protocol and channel conditions of a received signal. Embodiments of the invention may be configurable to operate within receivers employing receive diversity, equalization, transmit diversity combining, and/or space-time decoding.

[0025] Embodiments of the invention may include one or more CIVs. Therefore, in parts of the disclosure that describe a CIV, it is anticipated that a plurality of CIVs may be used. For example, specific embodiments may employ a matrix whose columns are CIVs. The CIV s is constructed from known and/or estimated active subchannels and then used to compute a soft projection matrix 102,

F()=Iss*.

The matrix F() is configured to operate on a received data vector y 103 to produce an interference-cancelled signal =F()y, which is coupled to a Rake processor or combiner (not shown). The term I is an identity matrix, and the weight may be determined symbol-by-symbol in order to maximize a post-processing SINR,

[00007]$\left(\right)=\frac{|x_u^{*}.Math.F\left(\right).Math.x_u.Math.{|}^2}{E|x_u^{*}.Math.F\left(\right).Math.x_{\mathrm{MAI}}.Math.{|}^2.Math.+E|x_u^{*}.Math.F\left(\right).Math.x_{I.Math.N.Math.T}.Math.{|}^2.Math.+{}^2.Math.x_u^{*}.Math.F\left(\right).Math.F^{*}\left(\right).Math.x_u}$

In this expression, each vector of the form x.sub.u is x.sub.u[m], corresponding to symbol period m. Therefore, the post-processing SINR () is measured symbol period-by-symbol period. The user powers are absorbed into the component vectors x.sub.u, x.sub.MAI, and x.sub.INT. These powers are known or estimated.

[0026] At each symbol period, the SINR at a given finger can be expressed as

[00008]$\left(\right)=\frac{a+b.Math.+c.Math.{}^2}{d+e.Math.+f.Math.{}^2}$

The coefficients are

[00009]$a=|x_u^{*}.Math.x_u.Math.{|}^2.Math..Math.b=-2.Math.x_u^{*}.Math.x_u|x_u^{*}.Math.s.Math.{|}^2.Math..Math.c=|x_u^{*}.Math.s.Math.{|}^4.Math..Math.d=\underset{u^{}u}{.Math.}.Math.|x_u^{*}.Math.x_{u^{}}.Math.{|}^2.Math.+|x_u^{*}.Math.s.Math.{|}^2.Math.+{}^2.Math.x_u^{*}.Math.x_u$$e=-2.Math..Math.\mathrm{Re}\left(\underset{u^{}u}{.Math.}.Math.\left(x_u^{*}.Math.x_{u^{}}\right).Math.\left(x_u^{*}.Math.s\right).Math.\left(s^{*}.Math.x_{u^{}}\right)+|x_u^{*}.Math.s.Math.{|}^2.Math.s^{*}.Math.s+{}^2|x_u^{*}.Math.s.Math.{|}^2\right)$$f=\underset{u^{}u}{.Math.}.Math.|x_u^{*}.Math.s.Math.{|}^2|s^{*}.Math.x_{u^{}}.Math.{|}^2.Math.+|x_u^{*}.Math.s.Math.{|}^2|s^{*}.Math.s.Math.{|}^2.Math.+{}^2|x_u^{*}.Math.s.Math.{|}^2.Math.\left(s^{*}.Math.s\right)$

wherein each of the inner products may be computed from the user codes w.sub.k[m] and complex amplitudes b.sub.l,u[m] identified for user u at baud interval m. If orthogonal spreading codes are used, the expression x.sub.u*x.sub.u with uu is zero. Furthermore, the relevant inner product x.sub.u*s can be efficiently obtained for a CDMA/WCDMA system by passing the synthesized CIV s for the finger of interest through a fast Walsh transform (FWT). Computing the soft projection matrix 102 may include a step of maximizing the SINR () by setting its derivative (with respect to ) to zero (not shown), resulting in the following polynomial equation

(cebf)/.sup.2+2(cdaf)+(bdae)=O.

One of the roots of the polynomial equation corresponding to the maximum SINR is selected (not shown) and then used to scale ss* in the matrix F(). Once computed, F()y may be scaled to conform to downstream processing in a baseband receiver.

[0027] It should be appreciated that variations to the previously described process for determining the weight may be made without departing from the spirit and scope of the claimed invention. For example, when a cellular handset is in a soft-handoff mode, there is an additional quadratic term in the numerator of () corresponding to the received signal power from the second base station, and there is one less term in the denominator. This changes the function (), but it does not change the procedure for determining the value of () that maximizes (). Furthermore, algorithms for maximizing () may be incorporated into other receiver processing techniques, such as (but not limited to) Rake path tracking, active user determination, amplitude estimation, receive diversity, and equalizing. () may be approximately maximized with variations or stochastic gradients.

[0028] FIG. 2A is a schematic block diagram of a circuit in accordance with an alternative embodiment of the invention that includes a CIV generator 201, an interference-cancelling operator 202, and a soft-projection canceller 203. Inputs to the CIV generator 201 and the soft-projection canceller 203 are coupled to outputs of a Rake receiver 200. An output of the soft-projection canceller 203 is coupled to the input of a combiner 210.

[0029] The soft-projection canceller 203 is configured to cancel interference from at least one path (or finger) of the Rake receiver 200. Soft and/or hard estimates from at least one other path or finger are processed by the CIV generator 201 to produce a CIV s. For example, FIG. 2B shows signals from rake fingers 210.1-210.N being used to construct a CIV in order to cancel interference from one of the rake fingers (e.g., 210.1). The interference-cancelling operator 202 uses the CIV s and user code x.sub.u to compute a soft-projection matrix. The soft-projection matrix computes the weight value that maximizes the SINR of the interference-cancelled signal =F()y. The interference-cancelled signal output from the soft-projection canceller 203 may be coupled into the combiner 210 and combined with interference-cancelled signals from other paths or Rake fingers.

[0030] The functions of the various elements shown in the drawings, including functional blocks, may be provided through the use of dedicated hardware, as well as hardware capable of executing software in association with appropriate software. When provided by a processor, the functions may be performed by a single dedicated processor, by a shared processor, or by a plurality of individual processors, some of which may be shared. Moreover, explicit use of the term processor should not be construed to refer exclusively to hardware capable of executing software, and may implicitly include, without limitation, digital signal processor DSP hardware, read-only memory (ROM) for storing software, random access memory (RAM), and non-volatile storage. Other hardware, conventional and/or custom, may also be included. Similarly, the function of any component or device described herein may be carried out through the operation of program logic, through dedicated logic, through the interaction of program control and dedicated logic, or even manually, the particular technique being selectable by the implementer as more specifically understood from the context.

[0031] The method and system embodiments described herein merely illustrate particular embodiments of the invention. It should be appreciated that those skilled in the art will be able to devise various arrangements, which, although not explicitly described or shown herein, embody the principles of the invention and are included within its spirit and scope. Furthermore, all examples and conditional language recited herein are intended to be only for pedagogical purposes to aid the reader in understanding the principles of the invention. This disclosure and its associated references are to be construed as applying without limitation to such specifically recited examples and conditions. Moreover, all statements herein reciting principles, aspects, and embodiments of the invention, as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Additionally, it is intended that such equivalents include currently known equivalents as well as equivalents developed in the future, i.e., any elements developed that perform the same function, regardless of structure.