Implementation of Blind Sequence Detection for Time Varying Channel in the Analysis of MIMO Capacity for Spatial Channel Model with Partial CSI Knowledge

Abstract

Listen

Translate article

Multiple-input multiple-output (MIMO) antenna systems provide high capacity gain over a singleantenna system. Much attention has been given to MIMO systems in recent past. In this paper, new blind estimation technique is developed to jointly estimate the partial Channel state information (CSI) of MIMO channel system. It is known that channel gain is maximised when channel state information (CSI) is available at the transmitter. It is found that in the case of partial or imperfect CSI the capacity depends on the statistical properties of the error in the CSI. It is shown that, in addition to improving the spectral efficiency, the proposed technique offer better channel estimation accuracy. It is shown that using the proposed blind estimation method for partial CSI knowledge of transmitter based on capon estimation technique with the Maximum-likelihood(ML) or the minimum variance(MV) MIMO results in a better performance as compared with the case when the standard Least Squre(LS) channel estimation. The results are comparable with perfect CSI knowledge spatial channel system. Introduction MIMO Antenna system is an attractive technique for high data rate transmission over frequency-selective fading channels due to its capability to combat the intersymbol interference (ISI), low complexity, and spectral efficiency [1][2]. Using MIMO technology at both the transmitter and receiver results in further increase in the capacity, provided that the environment is rich scattering [3]. SPACE-TIME coding is a powerful means to exploit spatial diversity and to combat fading in multiple-input multipleoutput (MIMO) wireless communication systems. Among various space–time codes developed to the date, orthogonal space–time block codes (OSTBCs) [4], [5], [6] are of particular interest because they achieve full diversity while offering very simple maximum-likelihood (ML) decoding in the single-user case. In most of the estimation techniques, it is known that the CSI of the transmitters is available at the receiver. This implies that pilot symbol should be used to estimate the transmitter CSI. Wheras it is known that use of pilot symbol may reduce the spectral efficiency and therefore blind and semiblind techniques can be used for spatial channel model with partial CSI knowledge. Most of the techniques are not directly implemented to multiuser MIMO systems since they are based on the point-to-point MIMO assumption and treat multiple access interference as a white noise and because of it, its performance may degrade in multiuser scenario. In this paper, a new approach to simultaneously estimate the partial CSI of MIMO system that uses Orthogonal Space Time Block Codes (OSTBC) to communicate with single transceivers has been proposed. New approach has been tried on the concepts of Capon technique that has been kept using in array processing and spectral analysis to the problem of spatial channel estimation in multiuser MIMO communication. It is shown that using such a similarity between array & multiuser MIMO channel estimation problem, the subspace that contains the spatial channel matrices can be estimated in a fully blind way. Results for improved spectral efficiency using the proposed blind channel estimation technique along with the ML or MV linear multiuser receivers are shown which also shows lower channel estimation errors as compared with the LS estimation method. Wireless Channel Model Assume sufficient antenna element spacing so that the subchannels between different transmit-receive antenna pairs are spatially uncorrelated, and have impulse responses of equal maximum resolvable paths L. Each subchannel is assumed to be slowly timevarying frequency-selective fading so that it can be considered as a constant during one OFDM symbol. The time-variant complex baseband CIR between the pth transmit and the qth receive antennas can be described as −