Research On The Resource Allocation In The Dynamic Spectrum Access Systems

Nowadays, people have more and more requirements on the data rate and transmission relia-bility, while the limited bandwidth and complicated environments restrict the development. Under such condition, dynamic spectrum access (DSA) and multiple-input multiple-output (MIMO), co-operative communication (CC) technologies have drawn more and more attention. The DSA is proposed to solve the low utilization of the spectrum, and the MIMO and the CC techniques can improve the capacity and reliability of the wireless communications. Furthermore, The cogni-tive radio (CR) is regarded as the critical implement of DSA. Therefore, by exploiting the MIMO and CC techniques in the CR systems, how to allocate wireless resource in order to achieve the optimum spectrum efficiency is worth studying. Meanwhile, as to the cellular spectrum sharing, Device-to-Device (D2D) communications show a significant advantage over CR. What’s more, the energy efficiency can’t emphasize more in the D2D communications because the D2D terminals are always battery-supported. Therefore, how to optimize wireless resource in order to achieve the optimum energy efficiency in D2D communications is also worth studying.Firstly, several optimization algorithms and optimization metrics are introduced. We start with discussing the convex optimization theory, including basic concepts, dual theorem, the KKT condi-tion, strong duality property and the time-sharing condition. Then the particle swarming optimiza-tion (PSO) algorithm is described, including the fundamental principles, process implementation, the parameters selection, and other modified algorithms. Next we introduce the bandwidth-power product (BPP) optimization metric, and analyze its differences compared to the traditional water-filling (TWF) optimization scheme. At last, the expected quantity of transmission data during battery lifetime (EQoB) optimization metric is provided and its application is offered.And then, taking advantage of the MIMO in the CR systems and using the BPP metric, the optimal wireless resource allocation schemes are developed. The single cognitive user communi-cation scenario is first studied, then the scenario is extended to the multiple cognitive user commu-nications and fully analyzed. We propose the joint power and channel interactive iterative (JPCII) algorithm to deal with the core resource allocation problem, and the innovative idea of JPCII is to decompose the original problem into two separate sub-problems, which can be independently solved. The numerical results show that the provided algorithms achieve the acceptable perfor-mance with great reduction of computational complexity compared to the PSO algorithm, and improve the spectrum efficiency compared to the TWF algorithm.Furthermore, exploiting the CC technique in the CR systems and using the BPP metric, the optimal wireless resource allocation schemes, including the relay assignment, relay protocol se-lection, the shared channel and the transmit power, are introduced. We first analyze the single cognitive user communication scenario and propose the greedy and joint resource optimization (GJRO) algorithm, which includes:for the given channel resource set, achieve the optimal chan-nel assignment and the optimum power allocation; next, obtain the optimal channel resource set using the greedy-cluster algorithm. Then the multiple cognitive user communications scenario is studied and the joint optimal relay matching and greedy resource allocation (JRMRA) is proposed. The JRMRA algorithm includes:regulate the optimal relay selection in terms of statistic expec-tation; next, perform the resource allocation to minimize the overall BPP using the greedy-based clustering method. The simulation results demonstrate that the provided algorithms achieve the acceptable performance with great reduction of computational complexity compared to the PSO algorithm, and improve the spectrum efficiency compared to the TWF algorithm.Finally, for the D2D communications underlying cellular networks, using the EQoB metric, the optimal wireless resource allocation schemes are presented to achieve the optimum energy efficiency of D2D links while guaranteeing the quality of service (Qos) of cellular links. We first study the single D2D user communication scenario and propose the optimal energy-efficiency (Opt-EE) algorithm, which transforms the original problem to the nonlinear programming (NLP) problem using the fractional programming method and solves the corresponding NLP problem with the joint iterative optimization method. Then the multiple D2D user communications scenario is analyzed and the greedy optimal energy-efficiency (Greedy-Opt-EE) algorithm is introduced, where each unoccupied cellular channel is orderly added to the D2D links until the total EQoB of D2D links can not increase any more. The numerical results verify the superiority of provided algorithms, which achieve better energy efficiency compared to the rate maximization method.