| In this thesis,interference alignment(IA)and rate splitting(RS),two novel interference management technologies,are studied to design the robust,flexible and spectrum efficient cooperative transmission strategies for three new scenarios-ultra-dense cell,dynamic time division duplex system and energy harvesting network.RS-based and IA-based asymmetric uplink(UL)-downlink(DL)cooperative transmission strategies are proposed.The RS-based asymmetric UL-DL cooperative transmission strategy is proposed to the new asymmetric interference properties.To maximize the time-averaged sum rate,the transmit direction and the covariances of the private and common messages are jointly designed.The results show that the proposed strategy brings gain in time-averaged sum rate compared to the conventional schemes using symmetric UL-DL transmission without RS for different users’ locations and network sizes.And the results indicate that the proposed strategy improves the performance of the users at the cell boundary and is superior for dense SC networks.The IA-based scheme for asymmetric UL-DL multi-cell multi-user multi-antenna system uses the interference properties to implement partial IA and guarantee feasibility of IA.Through minimizing weighted sum minimum mean square error(MSE),a regularized IA transceiver is designed to compensate the performance degradation in low and moderate signal-noise-ratio(SNR)region.The numerical simulations demonstrate that the proposed schemes outperform the previous works for the asymmetric UL-DL multi-cell system.The effects of imperfect backhaul links on the performance of IA network are studied.First,a novel hybrid coordinated multi-point(CoMP)transmission strategy is proposed to mitigate the impacts of the backhaul delay for data sharing and switch the transmission modes adaptively based on the current arrivals of the shared user data.Then,analytical expressions for the average sum rate in an IA network are presented,in which the combining effects of quantization from limited capacity and connection uncertainty introduced by finite delay are considered.With the analytical results,the sum rate performance is evaluated under limited backhaul for the conventional CoMP transmission modes and the proposed strategy.The simulation results demonstrate that the proposed strategy functions better on overcoming the impacts of capacity limited backhaul than the conventional ones.Finally,two robust IA based gradient methods on Grassmann manifold are proposed by utilizing the temporal correlation of channels:minimal interference leakage(min-IL)and maximal signal-to-interference-plus-noise ratio(max-SINR)IA based Grassaman gradient methods.Numerical results clearly demonstrate the spectral efficiency gain of the proposed transmission strategy.A signal and interference alignment based RS(SIA-RS)is proposed.The closed-form expressions of average sum rate,outage probability and symbol error rate are presented.Through the asymptotic analysis,the differences and relationships between IA and RS are clarified.The simulation results verify the analytical expressions and the effects of the number of the quantization bits and power allocation factor on the performance.RS is more robust than IA,but alignment can improve SIC performance.RS-based robust transceiver is designed for multi-antenna interference channel(IC)with simultaneous wireless information and power transfer(SWIPT)and harvested energy cooperation(EC)via smart meters under the normbounded errors(NBEs),respectively.RS-based robust SWIPT transceiver is designed to joint optimize the beamforming vectors,power splitting ratios and target common rates via minimizing the total transmit power subject to the quality of service and energy harvesting requirements.The non-convex problem with infinite constraints is first transformed into finite-constraint convex one,and then jointly solved by a gradient-based algorithm.The results demonstrate that the RS based robust schemes outperform the conventional robust designs.RS-based robust EC-CoMP transceiver is designed to joint optimize the beamforming vectors and transferred energy via maximizing sumrate subject to decodability for successive interference cancellation(SIC)and maximum transmit power with energy cooperation.The non-convex problem with infinite constraints is first transformed into finite-constraint convex one via the equivalent relationship between sum-rate and weighted sum mean square error and then solved with alternative optimization.The results demonstrate that RS and energy cooperation improve the robustness of transmission. |
