Thesis

Diversity gain enhancement for extended orthogonal space-time block coding in wireless communications

Creator
Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2013
Thesis identifier
  • T13634
Qualification Level
Qualification Name
Department, School or Faculty
Abstract
  • Transmit diversity is a powerful technique for enhancing the channel capacity and reliability of multiple-input and multiple-output (MIMO) wireless systems. This thesis considers extended orthogonal space-time block coding (EO-STBC) with beamsteering angles, which have previously been shown to potentially achieve full diversity and array gain with four transmit and one receive antenna. The optimum setting of beamsteering angles applied in the transmitter, which has to be calculated based on channel state information (CSI) at the receiver side, must be quantised and feed back to the transmitter via a reverse feedback link. When operating in a fading scenario, channel coefficients vary smoothly with time. This smooth evolution of channel coefficients motivates the investigation of differential feedback, which can reduce the number of feedback bits, while potentially maintaining near optimum performance. The hypothesis that the smooth evolution of channel coefficients translates into smooth evolution of feedback angles is justified by simulations. The maximum attainable gain under optimum unquantised beamsteering angles is derived, which allows to experimentally assess the effect that quantisation in the feedback channel has on the system performance. In characterising the degradation experienced through time-variation and limited quantised feedback, we demonstrate that the new differential feedback approach offers a practical bandwidth-efficient scheme. Simulation results with Doppler spread conditions confirm that the proposed scheme achieves significant bandwidth savings over previously proposed systems. With a single feedback bit per beamsteering angle the proposed differentially encoded EO-STBC approach can achieve near optimum performance and exceed the performance of non-differential feedback schemes that employ a higher word length. We further propose combining differential encoding with channel estimation that is practically useful because the EO-STBC receiver requires knowledge of the channel coefficients for both detecting the transmitted symbols as well as for computing the optimum angles to be fed back to the transmitter. Channel estimation accompanied by a decision-directed (DD) tracking scheme by means of a Kalman filter has been adopted. The Kalman filter exploits the smooth evolution of the channel coefficients as a motivation for tracking as well as for differential feedback. Further we propose applying an auto-regressive (AR) predictor with order greater than one in the Kalman model. This can be shown to offer advantages in terms of temporal smoothness when addressing channels whose coefficient trajectories evolve smoothly. Simulation results show that the overall EO-STBC system achieves longer tracking periods with suitable bit error (BER) values, and that the performance of the proposed system offers a distinct advantage for lower Doppler spreads with the inclusion of second order AR model instead of the standard first order AR model. The earlier work on EO-STBC systems is for frequency-flat channels. However, in frequency-selective channel a multi-carrier approach can help to split into independent subcarriers. Therefore, the EO-STBC scheme is then applied for a dedicated chirp-based multicarrier based on a fractional Fourier transformer (FrFT) system over doubly dispersive channels, where FrFT-domain is developed to further increase robustness against channel time-variations. Applied in nearstationary channel conditions, the performance of orthogonal frequency division multiplexing (OFDM) receivers that mitigate crosstalk between individual subcarriers are evaluated for open and closed loop schemes. A higher degree of non-stationarity in mobile scenarios will destroy the orthogonality of subcarriers and result in intercarrier interference (ICI) and intersymbol interference (ISI). In this case, minimum mean square error (MMSE) of a reduced system matrix is considered for open loop EO-STBC. The equaliser complexity can be decreased even furtherby using least squares minimum residual (LSMR) iterative algorithm, equalisation are underlined by simulations, demonstrating the overall practical use if the contributions wihtin this thesis towards EO-STBC diversity schemes over both time- and frequency-dispersive channels.
Resource Type
DOI
Date Created
  • 2013
Former identifier
  • 1004516

Relations

Items