Physical-layer Multicasting By Stochastic Beamforming And Space-time Coding

In this thesis,we design an effective transceiver that could approach channel capacity in a multi-input single-output(MISO)downlink multi-user multicast channel environment.Particularly,we have studied the multicast schemes under three different scenario settings,i.e.,channel state information(CSI)is known at the transmitter,channel state information is unknown at the transmitter and partial channel state information is known at the transmitter.We therefore propose to adopt transmit schemes that combine the stochastic beamforming and space-time codes to realize the approximation of multicast channel capacity.To be specific,under the scenario where perfect CSI is perceived at the transmitter,a classical multicast transmit approach is transmit beamforming,whereby the associated beamforming optimization problem is handled by a rank-one approximation method called semidefinite relaxation(SDR).The SDR-based beamforming has been shown to be promising when the number of users served is small,but will experience an obvious performance degradation when the number of users served is large.When combined with the Alamouti spacetime block code,the associated Alamouti-precoded transmit beamforming scheme can increase the number of beams and expand the degree of freedom of beam rank.However,it still cannot meet the increasing multi-user requirements,since it can only accommodate rank-two beams.Therefore,classical beamforming schemes are only suitable for a parameter environment with a small number of users.When the number of users served is large,two new multicast transmit strategies—the stochastic beamforming scheme and the Alamouti-precoded stochastic beamforming scheme,are proposed to improve multicast rates.It has been proven that,when these two new proposed schemes are adopted,their rate gaps to the MISO multicast channel capacity,in the worst case,can be reduced to 0.8314 bit/s/Hz and 0.39 bit/s/Hz,respectively.Based on the above research results,this paper firstly goes deeper to the research about stochastic beamforming and space-time block codes.In the closed-loop system,i.e.,when CSI is known at the transmitter,we employ the quasi-orthogonal space-time block code(QOSTBC)on scholastic beamforming so that the transmitter can employ more spatial degrees of freedom.Although the inter-symbol interference at the receiver side will somehow counteract the increased antenna gain,the QOSTBC still benefits the stochastic beamforming in some scenarios.The simulation results show that when the rank of optimal transmission covariance is large,the combination of the stochastic beamforming and the QOSTBC exhibits even better bit error rate(BER)performance than the original single-stream stochastic beamforming scheme.Secondly,this paper considers the open-loop scenario,i.e.,channel state information is unknown at the transmitter.Through simulation,we prove that the stochastic beamforming scheme can still provide better multicast rate compared with the classic space-time coding scheme and the rank-one transmit beamforming scheme.Among all stochastic beamforming schemes,the Alamouti-precoded stochastic beamforming scheme performs the best,followed by the QOSTBC-precoded stochastic beamforming scheme.Finally,this paper considers the multicast scheme when partial channel state information is known at the transmitter side.Specifically,we adopt the antenna subset selection strategy to serve the purpose.The simulation results show that the combination of antenna subset and stochastic beamforming performs better than the original antenna subset selection scheme.Moreover,the lamouti-precoded stochastic beamforming scheme exhibits the best performance among all the stochastic beamforming schemes.