Admission Control And Resource Allocation Optimization Technology In The Hybrid Cellular And D2D Network

A key technique of improve the utilization of spectrum and remit the shortage of spectrum resource is cognitive radio (CR). Based on CR technology, Device-to-Device (D2D) communication technology is proposed. It does not need the base station for data forwarding in D2D communication, and it can save half of the system spectrum resources compared with cellular communication without consideration of signaling overhead. D2D transmission has features of low delay, high link quality, low transmission power, and reuse of cellular spectrum resources. So it can effectively reduce cellular network load by using D2D transmission mode in local business. Hence, hybrid cellular and D2D network will be used in the future, in which cellular network is the primary system and D2D network is the secondary system that use the licensed spectrum of cellular network. The main work of this paper include:First, simulate and analyse the interference introduced by D2D in the cellular network, especially when D2D reuse cellular resources. D2D can reuse intra-cell or inter-cell cellular resources, when D2D reuse uplink resource, cellular user will interfere D2D receiver and D2D transmitter will interfere base station; when D2D reuse downlink resource, base station will interfere D2D receiver and D2D transmitter will interfere cellular user; there also exist cumulative interference of many D2D users. So we need admission control and resource allocation optimization mechanisms to avoid or ease the interference. The interference introduced by D2D is not obviously in orthogonal manner, and the interference control mechanism can use the manner in traditional cellular system.Second, propose centralized Set-based Admission Control (SAC) and Distributed Power Optimization (DPO) algorithms for D2D links in hybrid network. In proposed algorithms, each D2D link is in one of three sets (admitted, unadmitted, or temporary remove user sets) according to its quality of service (QoS) level. In the presence of QoS requirement for D2D links, interference threshold for the cellular link, etc. User in the three sets use three kinds of power update manner in SAC algorithm, and three novel strategies of sets adjustment are presented to handle the violations of constraints. Through SAC, the admitted user set contains users as many as possible at last, and other users are in the remove set. Then, for the admitted user set, DPO algorithm uses penalty function and geometric programming to transform the power optimization problem to an unconstrained convex problem, and uses gradient decent method to update powers to optimize D2D system capacity. Through the numerical results, proposed algorithms have more admitted D2D links and higher D2D system capacity compared to existing algorithms we researched.Third, propose two kinds of Distributed Admission and Power Control algorithms, DAPC-Capacity and DAPC-Number algorithms, for D2D links in hybrid network to determine if the new D2D link can be admitted and optimize powers of D2D links to maximize the D2D system capacity. Both the two algorithms combine the admission control and power optimization to a single process to simplify the system operation, and can decrease the network signaling overhead compared to centralized algorithms. With QoS requirement for D2D links and interference threshold for the cellular link, the DAPC-Capacity algorithm is proposed with preference for the D2D system capacity, and the DAPC-Number algorithm is proposed with preference for the number of admitted D2D links. In proposed algorithms, D2D links cooperate with each other and only need to exchange limited informations to update powers. At last, we get satisfying performance of the two distributed algorithms through simulations compared to existing centralized algorithms.