On The Study Of Interference Management And Secure Wireless Information And Power Transfer For Multi-antenna Systems

Opportunistic interference alignment (OIA) is one of the advanced inter-ference management techniques, and the key idea of OIA is to exploit the po-tential multiuser diversity in the system. With low-complexity user schedul-ing and transceiver design, OIA is able to asymptotically achieve the optimal degrees-of-freedom (DoF). OIA can lead to a better understanding of the in-terference management technique in academia, and it has important theoretical and practical values. Moreover, with the development of wireless energy har-vesting, radio-frequency signals can be used as a vehicle for simultaneously transmitting information and energy, and thus the scheme named simultaneous wireless information and power transfer (SWIPT) comes into being. However,due to the broadcast nature of the wireless medium in the network with wire-less power transfer, the messages intended to the information receivers may be at risk of being eavesdropped by the energy receivers. Physical-layer security(PLS) has emerged as a complementary solution to the conventional higher-layer cryptographic methods, and it does not rely on the computational capabil-ity of the eavesdroppers. PLS is able to guarantee the perfect secrecy commu-nication from the information-theoretic perspective. By integrating PLS with SWIPT, secure transmission in the network with wireless power transfer can be achieved, and it is of great significance for in-depth study. The disserta-tion mainly focuses on the research of OIA based interference management and secure wireless information and power transfer for multi-antenna system.By properly controlling and utilizing the interference, the dissertation aims to achieve effective and secure information transmission and efficient power trans-fer in the system.The main contributions and innovations of the dissertation are summarized as follows:1) The dissertation proposes the intra-cell performance aware OIA schemes.Considering the multi-cell interference network, the dissertation proposes two OIA schemes. The proposed schemes use the predefined reference signal space of each cell in the network, and design the intra-cell performance aware schedul-ing metrics. The scheduling metrics can characterize both the inter-cell inter-ference and the desired signal strength of each user. Moreover, the disserta-tion provides theoretical analysis for the proposed OIA schemes, which demon-strates that compared with the conventional OIA schemes, the proposed OIA schemes can not only achieve the same optimal DoF and user scaling law but also achieve a better sum rate performance. Furthermore, the dissertation also proposes a limited feedback strategy for the proposed schemes to further reduce the feedback overhead.2) The dissertation proposes the OIA based wireless information and power transmission schemes. Considering the multi-cluster interference network, based on the idea of OIA, the dissertation proposes two node scheduling schemes to facilitate efficient information transmission and power transfer in the network.The scheduling metrics of the proposed two schemes can characterize both the information transmission ability and power transfer ability of each node.The information transmission ability of each node can be measured by the ra-tio of the desired signal strength and the leakage of interference. The power transfer ability of each node can be measured by the energy provided by the node. Moreover, the dissertation provides theoretical analysis of the proposed schemes, which verifies that compared with the conventional schemes, the pro-posed schemes can not only guarantee the information transmission but also enhance the energy harvesting performance and prolong the working time of the system. Furthermore, in order to further reduce the feedback overhead, the dissertation also proposes a limited feedback based node scheduling scheme.3) The dissertation proposes a secure wireless information and power trans-mission scheme in interference channels. First, considering the two-user inter-ference channel with SWIPT and assuming imperfect eavesdropper's channel state information (ECSI) at the transmitters, the dissertation proposes a robust resource allocation scheme. The proposed scheme aims to minimize the total transmit power subject to the secrecy rate constraint of the eavesdropped infor-mation receiver, the signal-to-interference-plus-noise ratio (SINR) constraint of the unclassified information receiver and the energy harvesting requirement at the energy receiver. Second, considering the K-user SWIPT interference chan-nel, the dissertation proposes a robust joint beamforming and artificial noise design. The goal of the proposed scheme is to maximize the secrecy rate of the wiretapped information receiver subject to the individual power constraint of each transmitter, the SINR constraints of the unclassified information receivers and the energy harvesting requirement of the energy receiver. For the above two cases, the original problems are non-convex, which cannot be solved di-rectly. With the aid of rank relaxation and one-dimensional line search, the original problems can be transformed into a sequence of convex problems, and the optimal solutions can be obtained.4) The dissertation proposes a secure wireless information and power trans-mission scheme in heterogeneous networks. Considering the two-tier downlink heterogeneous network, the dissertation proposes an artificial noise aided se-cure transmission scheme for the macrocell base station and the femtocell base stations. The proposed scheme jointly optimizes the beamforming vectors and artificial noise of the macrocell base station and the femtocell base stations to maximize the secrecy rate of the eavesdropped macrocell user under the SINR constraints of the unclassified macrocell users and femtocell users, the energy harvesting constraints of the energy receivers and the transmit power constraints of the macrocell base station and the femtocell base stations. First, the scenario of perfect ECSI at the macrocell base station and the femtocell base stations is addressed. The dissertation transforms the original problem into a two-stage problem. In particular, with the help of rank relaxation, the inner stage is trans-formed into a semidefinite programming problem. The outer stage turns to be a common one-dimensional line search problem. Therefore, the original secrecy rate maximization problem can be efficiently solved by dealing with a sequence of convex problems. Moreover, we also prove that the relaxed problem always yields a rank-one solution. Second, we consider the scenario of imperfect ECSI at the macrocell base station and the femtocell base stations. The dissertation transfers the original non-convex problems into convex problems by utilizing semidefinite relaxation and successive convex approximation, and proposes an iterative algorithm to obtain the optimal solutions.