Performance of MIMO systems in measured indoor channels with transmitter noise

Castro, González-Coma, García-Naya, Castedo, Castedo Luis

Description

This study analyzes the impact of transmitter noise on the performance of multiple-input multiple-output (MIMO) systems with linear and nonlinear receivers and precoders. We show that the performance of MIMO linear and decision-feedback receivers is not significantly influenced by the presence of transmitter noise, which does not hold true in the case of MIMO systems with precoding. Nevertheless, we also show that this degradation can be greatly alleviated when the transmitter noise is considered in the MIMO precoder design. A MIMO testbed developed at the University of A Coruña has been employed for experimentally evaluating how much the transmitter noise impacts the system performance. Both the transmitter noise and the receiver noise covariance matrices have been estimated from a set of 260 indoor MIMO channel realizations. The impact of transmitter noise has been assessed in this realistic scenario.

Informations

• Publié par: biomed
• Publié le: 01 janvier 2012
• Nombre de lectures: 6
• Langue: English
• Poids de l'ouvrage: 2 Mo

Extrait

Castro et al. EURASIP Journal on Wireless Communications and Networking 2012, 2012:109
http://jwcn.eurasipjournals.com/content/2012/1/109
RESEARCH Open Access
Performance of MIMO systems in measured
indoor channels with transmitter noise
*Paula M Castro, José P González-Coma, José A García-Naya and Luis Castedo
Abstract
This study analyzes the impact of transmitter noise on the performance of multiple-input multiple-output (MIMO)
systems with linear and nonlinear receivers and precoders. We show that the performance of MIMO linear and
decision-feedback receivers is not significantly influenced by the presence of transmitter noise, which does not
hold true in the case of MIMO systems with precoding. Nevertheless, we also show that this degradation can be
greatly alleviated when the transmitter noise is considered in the MIMO precoder design. A MIMO testbed
developed at the University of A Coruña has been employed for experimentally evaluating how much the
transmitter noise impacts the system performance. Both the transmitter noise and the receiver noise covariance
matrices have been estimated from a set of 260 indoor MIMO channel realizations. The impact of transmitter noise
has been assessed in this realistic scenario.
1 Introduction In this study, the impact of residual non-deterministic,
Practical implementations of wireless transmitters suffer non-systematic transmit impairments on MIMO
signalfrom a large number of impairments such as quantiza- ing methods not considered in [1,2] is studied. First of
tion noise, sampling offset, phase noise, I/Q imbalance all, we focus on MIMO systems with either linear
recei... which can be classified into systematic and non-sys- vers [3-5] or linear precoders [4-8]. We will also focus
tematic effects. These impairments are normally ignored on MIMO systems with nonlinear decision feedback
when multiple-input multiple-output (MIMO) signaling (DF) receivers [9,10] and nonlinear
Tomlinson-Haramethods are designed. However, noise generated at the shima (TH) precoders [5,11,12]. Both DF receivers and
transmitter can significantly affect predicted perfor- TH precoders are widely used because of their good
mance in practical scenarios. For instance, the p trade-off between performance and complexity. The of a linear zero-forcing (ZF) MIMO detector modulo operator at the receiver of a MIMO system
affected by the transmitter impairments of a MIMO with TH precoding has also motivated the proposal of a
orthogonal frequency-division multiplexing (OFDM) more general MIMO precoding technique where the
hardware demonstrator is tested in [1], where it is data signal superimposed with a perturbation signal
shown how the performance achieved by such noisy sys- inputs a linear filter at the transmitter. This scheme is
tems suffers from a loss greater than 4dB for a bit error referred to as vector precoding (VP) in the literature
-2
rate (BER) of 10 . More specifically, the impact of resi- [5,13]. The optimum perturbation signal is found with a
dual transmitter radio-frequency (RF) impairments on closest point search in a lattice. Despite its larger
comboth MIMO channel capacity and receiver performance plexity, VP outperforms TH precoding.
Although the impact of transmitter noise has beenis analyzed in [2]. It has been demonstrated that
maxievaluated over spatially-white Rayleigh channels inmum-likelihood (ML) and Max-log a posteriori
probability (APP) MIMO detection suffer from a substantial [14,15] and some preliminary results obtained from
performancelossunderthepresenceofweaktransmit- testbed measurements in an indoor scenario have been
ter noise, whereas linear ZF receivers are much less presented in [15,16], only our study evaluates the
peraffected. formance of all the aforementioned schemes affected
by transmitter noise over measured indoor channels by
* Correspondence: [e-mail] means of the testbed developed by the University of A
Department of Electronics and Systems, University of A Coruña, Facultad de
Coruña [17]. Contrary to previous studies, not only the
Informatica, Campus de Elviña s/n, A Coruña 15071, Spain
© 2012 Castro et al; licensee Springer. This is an Open Access article distributed under the terms of the Creative Commons Attribution
License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited.Castro et al. EURASIP Journal on Wireless Communications and Networking 2012, 2012:109 Page 2 of 19
http://jwcn.eurasipjournals.com/content/2012/1/109
channel coefficients are measured using such a testbed, Nr (1)y[n]= Hx[n]+ η [n] ∈C ,r
but also a estimate of all the noise covariance matrices
Ntexisting in practical MIMO systems and of the signal- where x[n] ∈C represents the transmitted signals,
to-transmitter noise ratio (STxNR) are obtained and Nr is the noise vector introduced by the recei-η [n] ∈Cr
plugged into the robust filter expressions to guarantee vers (which will be referred to as the Rx-noise),
a complete reproducibility of a real indoor environ- N ×Nr t is the MIMO channel matrix, andH ∈C
ment and the evaluation in terms of BER performance Nr is the received signal vector. Note that wey[n] ∈C
of the proposed schemes under such realistic
assume a block-fading channel, where H remains con-conditions.
stant during the transmissionofadataframe.Rx-noiseIn this study, we show that noise generated at the
is complex-valued, Gaussian-distributed with zero meantransmitter significantly affects the performance of
Hand covariance matrix C = E η [n]η [n] , i.e.,ηMIMO systems with precoding. On the other hand, the rr r
performance of both linear and DF MIMO receivers is η [n] ∼N 0,C . Transmit energy is constrained tor C ηr
more robust against the presence of transmitter noise a value
(cf. findings in the FP6-IST project MASCOT on
mulE = tr(C ), (2)tiuser MIMO communication systems [18]). At a first tx x
glance, it seems natural that MIMO systems with
prewhere C is the covariance matrix corresponding toxcoding be more sensitive to transmitter noise since
the input symbols x[n]. Accordingly, we define the
sigchannel equalization is carried out by processing signals
nal-to-receiver-noise ratio (SRxNR) as
prior transmission. However, it will be also shown that
the performance of precoded MIMO systems can be E Ntx r
SRxNR = , (3)substantially improved if the transmitter noise is consid- tr Cηr
ered inside the precoder design.
This study is organized as follows. Section 2 intro- where we have assumed that the channel is
normalduces the MIMO signal model which takes into account ized to have a mean Frobenius norm equal to N N .t r
the aforementioned transmit RF impairments. The When the residual impairments at the transmitter are
design of MIMO linear receivers and precoders consid- taken into account, a more accurate model for the
ering the presence of transmitter noise is addressed in transmitted signal is
Section 3, whereas Section 4 focuses on MIMO
nonNt (4)x [n]= x[n]+ η [n] ∈C ,linear receivers (decision feedback) and precoders (Tom- t t
linson-Harashima and vector precoding). Section 5
Ntwhere will be denoted as the Tx-noise.η [n] ∈CbrieflyintroducestheMIMO testbed used to obtain t
The subscript will be used hereafter to denote signalsmeasurements of the noise and of the channel para- t
affected by Tx-noise. This noise encompasses differentmeters necessary to reproduce a realistic, indoor
scepractical effects, for example phase-noise. As explainednario. Next, the performance of the MIMO systems
in [2], Tx-noise is adequately modeled as an additiveanalyzedintheprevioussectionsisevaluatedforthis
Gaussian noise since it results from the sum of a largeindoor scenario. Some concluding remarks are presented
number of residual transmit impairments. Tx-noise isin Section 6. Finally, Appendix 1 details all derivations
assumed to be zero-mean with covariance matrix Ccorresponding to the expressions of all linear and non- ηt
linear filters derived including transmitter noise in the and the STxNR is defined as
optimizations.
EtxVectors and matrices are denoted by lower-case bold, STxNR = , (5)
tr Cηand capital-bold letters, respectively. We use E[?], tr(?), t
T H(?) ,(?) ,and ∥ ? ∥ for expectation, trace of a matrix,2
where E is fixed and given by Equation (2). As antxtransposition, conjugate transposition, and Euclidean
example, practical implementations of the IEEE 802.11
norm, respectively. The ith element of a vector v is
(WiFi) standard achieve STxNR values ranging from 22
denoted by v.i
dB to 32 dB (see [2] and references therein). The
Txnoise is also assumed to be statistically independent2 Signal model with transmitter noise
from the Rx-noise. Note that we use the STxNR instead
Let us consider a narrowband MIMO communication
of the transmitter error vector magnitude (EVM), which
system with N transmit and N receive antennas. Whent r
is another measure of the signal modulation quality (see
considering only the receiver noise, this system can be
[19]) that additionally considers systematic errors.
represented by the following discrete-time modelCastro et al. EURASIP Journal on Wireless Communications and Networking 2012, 2012:109