Relay Selection And Resource Allocation In Cognitive And Cooperative Wireless Networks

As the rapid development of social informatization process, the demand on mobile data services like broadband, mobile and the Internet of Things (IoT), etc. is growing rapidly, which brings sharp increases in demand of bandwidth of wireless communica-tion networks. From the first generation (1G) of cellular communication systems in the1980s to the coming fourth generation (4G) mobile communication systems, great progresses have been made in the wireless communication technologies. However, com-paring to the rapid growth of business and market applications, the information science and communication technologies are lagging behind and facing with many difficulties and challenges, e.g., security, scalability, spectrum shortage, Quality of Service (QoS), etc.Radio spectrum resource:With the rapid development of wireless communi-cation networks, the conflict between the limited spectrum resources and the rapidly increasing in demand for spectrum is becoming increasingly acute. On one hand, almost all of the spectrum resource has been occupied under current spectrum management policy. On the other hand, there exists great waste on the valuable spectrum resources because of the underutilization of assigned spectrum band in both time and spatial dimensions. In1999, Cognitive Radio (CR) was first proposed as the most promising technology to solve the conflict. The main idea of CR is to break the fixed spectrum management mechanisms and to allow flexible and dynamic spectrum access of the unauthorized users under the constraints that the authorized users are not affected, which improves the spectrum efficiency in both time and spatial dimensions.Link reliability and network connectivity:The time-varying wireless chan-nel, the dispersion and distance attenuation of radio signal propagation, and the in-terferences in the heterogeneous network all challenge the wireless link reliability and network connectivity in different ways. A virtual multi-antenna scheme, known as co-operative communication, has been proposed to deal with the challenge by allowing cooperation among multiple users each equipped with only single-transceiver antenna to explore the spatial diversity.It can be predicted that the future wireless communication system will be a het-erogeneous network consisting of different network, different application requirements and the mutual penetration of different coverage. The afore mentioned conflicts have become the main factors that restrict on the development of the wireless communica-tions, and the difficulties and challenges in wireless communications can not be solved by any single technology. As a result, this dissertation relies on the combination of the CR and cooperative relay technologies to improve the spectrum efficiency, the wireless link reliability and network connectivity. CR enhances the system efficiency from the perspective of degree of freedom (DoF), while cooperative communication enhances the system reliability from the perspective of diversity gains. Researches on relay selection and resource allocation in cognitive and cooperative wireless networks aim at provid-ing theoretical guidance for the design of future wireless communication networks and a theoretical basis for the network performance analysis. Such researches also pro-vide theoretical and algorithmic support for the design of radio resource management protocol like relay selection and subcarrier allocation. The communication scenarios considered in this dissertation vary from single-user cooperative relaying wireless net-work to multi-user cognitive and cooperative wireless network, the corresponding for-mulated problems change from relay selection algorithm design to the algorithm design of joint relay selection and subcarrier allocation. To solve these formulated nonlinear non-convex0-1integer programming problems, based on combinatorial optimization, artificial intelligence algorithms, and convex optimization theory and methods, on one hand we design the globally optimal solution algorithm that can be used for perfor-mance analysis and to provide a performance benchmark, and on the other hand, we design practical solutions with lower complexity. Research in this dissertation was sup-ported by the National Natural Science Key Foundation of China "Research on Cogni-tive and Cooperative Communications"(No.60832008) and National Natural Science Foundation of China "Theory Research and Implementation of Cooperative Multime-dia Communication in Heterogeneous Wireless Sensor Network "(No.61061122). Main contributions of the dissertation are as follows. 1. The relay selection and resource allocation strategies in cooperative relay wireless networks are studied. For a single source-multiple relay nodes-single destination network, with relay nodes working in two-hop amplify-and-forward (AF) mode, we design the multiple relay selection algorithm to maximize the received SINR at des-tination. Based on branch and bound architecture and linear relaxation technique for nonconvex items, we design global optimal multiple relay selection algorithm. Moreover, in order to decrease the computational complexity, we design an adap-tive genetic algorithm which can be verified to reach the optimal performance with lower complexity. For multiple source-multiple relay nodes-multiple destination, with relay nodes working in two-hop decode-and-forward (DF), by designing the chromosome representation, survivor selection, crossover and mutation procedures, we design genetic algorithm based relay selection and subcarrier allocation algo-rithm to maximize the sum capacity of all users. This contribution corresponds to Chapter2and paper4,5and6as listed at the end of the dissertation.2. The relay selection algorithms in single-user cognitive and cooperative wireless net-works are studied. The cognitive network which is composed of single secondary user pair and multiple two-hop AF cognitive relays work in the transmission scope of primary users. Cognitive network works in underlay mode to share the spec-trum with primary users. We design a relay selection algorithm to maximize the secondary suers'capacity subject to the interference temperature constraint of the primary user. When maximal ratio combining (MRC) is adopted by secondary destination, the problem is formulated as a0-1knapsack problem, based on greedy scheme and best-single scheme, we propose a hybrid scheme. The simulation re-sults show that the proposed scheme can overcome the drawback of greedy scheme and get near optimal performance. When equal gain combining (EGC) is adopted, the problem is formulated as0-1nonlinear nonconvex integer programming prob-lem, we first design the optimal relay selection scheme based on branch and bound architecture. Some simplified schemes and an iterative scheme with quadratic computing complexity are then proposed to decrease the complexity with slight sacrifice of performance. This contribution corresponds to Chapter3and paper1,2and3listed at the end of the dissertation. 3. The relay selection and subcarrier allocation problem in multicarrier based mul-tiuser cognitive and cooperative wireless networks is studied. The network consists of multiple primary user pairs (each including one transmitter and one receiver), multiple secondary user pairs and multiple secondary two-hop DF relay nodes, all the secondary users share spectrum with primary users in the underlay mode. The problem is first formulated as a nonconvex problem with complicated constraints, after equivalence transformation, the two two-dimension matrices are converted to a three-dimension0-1coefficient matrix and the problem is equivalently reformu-lated as a0-1linear programming problem. Based on branch and bound procedure, the proposed relay selection and subcarrier allocation algorithm can reach optimal solution within very limited number of iteration (less than2.5times after average). Simulation results prove that this algorithm has far less computing complexity than exhaustive scheme. This contribution corresponds to Chapter4and paper7listed at the end of the dissertation.