08-17-2017, 12:45 AM
Abstract:
The combination of antenna array beamforming with multiuser detection can effectively
improve the detection efficiency for wireless communications under multipath interference,
especially for the applications in a fast fading channel. The authors study the performance of an
adaptive beamformer incorporated with a B-MMSE detector, which works on a unique signal
frame characterised by training sequence preamble and data blocks segmented by zero-bits. Both
the beamformer weight updating and B-MMSE detection are carried out by either LMS or RLS
algorithms. The comparison of the two adaptive algorithms applied to both the beamformer and
the B-MMSE detector is made in terms of convergence behaviour and estimation mean-square
Download full report
http://ieexplore.ieexpl/freeabs_all.jsp?...er=1370506
error. The final performance in error probability is also given. Various multipath patterns are
considered to test the rapidity of the receiver s response to changing multipath interference. The
performance of the adaptive B-MMSE detector is also compared with that of the non-adaptive
version, i.e. through matrix inversion. The results obtained suggest that the adaptive beamformer
should use an RLS algorithm for its fast and robust convergence property; whereas the B-MMSE
filter can choose either LMS or RLS algorithms depending on the antenna array size, multipath
severity and complexity.
1 Introduction
To obtain an increasingly high transmission rate for next
generation wireless communications, two major problems
have to be addressed: one is multipath interference and the
other is multiple access interference (MAI). It has been
shown that the use of an antenna array together with
multiuser detection is an attractive means to overcome the
above two problems. As a mobile channel is a fast varying
medium, beamforming weight updating and multiuser
detection should preferably be carried out by adaptive
algorithms. This paper addresses the issues related to
adaptive implementation of a joint block-based minimum
mean square error (B-MMSE) detector and antenna array
system. For both antenna beamforming and B-MMSE
detection, two adaptive algorithms, least mean squares
(LMS) and recursive least squares (RLS), will be considered.
Because a receiver in mobile communications in
general does not know the direction-of-arrival (DOA) of
incident signals (particularly in multipath scenarios), the
receiver should be made to work under DOA-blind
conditions.
Before introducing our system model, let us first review
some related work in this area. Pham and Vu [1] studies an
adaptive MMSE and antenna array system, where the LMS
algorithm was used for both adaptive beamforming and
MMSE detection. It is well known that the LMS algorithm
is rather sensitive to the eigenvalue-spread of an input
correlation matrix (which is related to MAI power) [2].
Thus, it may not be a good choice to only use an LMS
algorithm, due to its power-dependent convergence property
that is associated with the near far effect. On the other
hand, the convergence speed of an RLS algorithm is
interference power independent, making it more preferable
for applications in time-varying channels. The paper,
unfortunately, did not address the issues associated with
the application of the RLS adaptive method. Besides, the
MMSE detection concerned in the paper operates on the
entire message length and thus the complexity of MMSE is
another serious concern. Miller [3] presented a scheme for a
single-user direct sequence/code division multiple access
(DS/CDMA) receiver using a chip waveform matched filter
followed by an adaptive equaliser to perform the despreading
operation. The performance of such an adaptive receiver
was compared with a conventional receiver. Chin and Chao
[4] proposed an LMS-adaptive minimum least square error
(MLSE) equalisation scheme for signal detection under a
multipath fading channel. The core of the scheme consists
of an LMS estimator and a linear channel predictor. The
authors succeeded in deriving a new upper bound on the
block error rate by considering imperfect channel estimates.
J. Razavilar et al. [5] offered a comprehensive tutorial on
smart antenna systems with software radio implementations.
Their paper also, in particular, addressed the issues of
adaptive beamforming. In another paper [6], published by
F. Rashid-Farrokhi et al., the issues of joint power control
and beamforming were studied.