SCMA Based Radio Resource Management For Device-To-Device Systems

With the continuous development of communication services, the current network capacity is still far from satisfying the mass demands of existing equipment. For more efficient exploiting bandwidth resources,3GPP proposed D2D (Device-to-device) communication as an underlayer added solution to cellular networks.5G wireless technologies will also be completed on the new radio access technology and the evolution of existing wireless technologies. Under the above background, this thesis studies resource allocation and optimization problems based on sparse code multiple access (SCMA).Firstly, this thesis introduces the basic technology about SCMA, including codebook design and reuse, definition of competition transmission unit (CTU), mapping mechanism of users to CTUs and competitive transport mechanism. In codebook design, we refer to spreading method of LDS (low-density signature) and QAM (Quadrature Amplitude Modulation) mapping, and combine to produce a new set of complex vector with the sparsity of SCMA, namely the codewordes. We encode with multi-dimensional constellation to generate multi-demensional codebook instead of QAM modulation and LDS spreading in SCMA. Moreover, this thesis introduces the method for resources reuse by codebooks and proposes a new concept of resource block:CTU. In addition, compared to traditional time-frequency resource block, there are two more dimensions of CTU:codebook and pilot. We design mapping rules to map users to CTUs and utilize the competition backoff mechanisms to reduce the probability of collision to improve the transmission effciency of the entire system.Then, we study the theory of resource allovation and optimization with getting the channel information. We solve the problem by three steps:1) Design D2D pair access guidelines:as the maximum transmint power and minimum SINR requirement of users have been defined, through solving the corresponding inequality problem, we simplify the access criterion for D2D pairs by taking the distance between cellular user and D2D pair, and get the piecewise function; 2) Allocate power to the multiplexed channels:we get the optimal power point with the theory of convex optimization in three situations of different feasible region; 3) Find the final optimal channel allocation:we can determine the total throughput of current D2D pairs according to the optimal power allocation. By means of weighted bipartite graph algorithm in graph theory and KM based on Hungary algorithm, each D2D user can get its best channal.Finally, based on coalitional game theory, we study the resource allocation scheme in downlink SCMA D2D communication systems. With the assurance of the QoS of cellular users and D2D pairs, we build the objective function to make a maximum sum-rate based on SCMA system.With coalitional game theory, we assign each D2D pair to an SCMA layer and solve the optimal power problem within the various members of the coalition by Lagrangian dual decomposition method to achieve the maximum effectiveness of the coalition. By simulation, we compare the system sum-rate of general cellular network (LTE as an example) with the SCMA system. It is obvious that SCMA technology has significant effect on enhancing system capacity and spectral efficiency. Moreover, we also compare the coalitional game algorithm to the greedy selection algorithm and random selection algorithm, and find that the coalitional game algorithm contributes to the system a better sum-rate.