Resrarch On Radio Resource Management Technology Of Device-to-Device Heterogeneous Network

In mobile information age,people have more demand for data services.This trend urges the future communication system to enhance system capacity.The traditional network depended on base station structure encounters resistance of capacity enhancement.With the features of business localization is more and more obvious,device-to-device(D2D)communication has attracted widely attention.Thus,this communication pattern has an advantage on resource reuse.The thesis focuses on the radio resource management and interference management technology of D2D communication reusing cellular users' resources.To maximize the spectrum efficiency,we evaluate the performance and computation complexities of different schemes and algorithms.The study includes the following three aspects.Firstly,the coexisting scenario of D2D communication and Multiple Users Multiple Input Multiple Output(MU-MIMO)is studied.To begin with,we present the design of interference alignment(IA)and minimum mean square error(MMSE)detection technology,and prove the feasibility of IA in order to acquire the user parameters as well as encoding and decoding matrices.Afterward,we proposed a two-step method by transforming resource allocation and user clustering optimization function into binary integer programming.Finally,a location-based allocation scheme considering the strengths of both IA and MMSE detection is proposed.The simulation results show that IA which is better than MMSE detection in terms of interference elimination can acquire higher spectrum efficiency.Besides,the two-step method with polynomial complexity is better than exhaustive search with exponential complexity,and both performances are similar.The location-based scheme is a good tradeoff of performance and complexity.It can obtain the solution equal to that of the two-step method.Moreover,its complexity is linear,which is suitable for practical use.Secondly,we study the scenario where each cellular user equipment(CUE)can share its resource with multiple D2D links with CUE QOS guarantee,by considering multiuser matching,resource allocation and power control.To begin with,we formulate the function of a mixed combinatorial integer programming.Afterward,an improved branch-and-bound method is proposed to figure out the discrete part,its course is faster to obtain a better initial feasible solution than before.Finally,we present a cluster-match algorithm which can eliminate the redundancy of the search space.The results show that D2D communication reusing cellular resource based on user-pairing can greatly improve spectrum efficiency.The proposed algorithms outperform than the auction method,the greedy algorithm and random mode,and their complexities are lower than exhaustive search's.In addition,the asymmetric allocation scheme is superior to the symmetric one in a small gap,but its superiority should be based on suitable methods and more computation cost.Thirdly,we study power control algorithms in the multiple user reuse scenario.We consider all users' QOS requirement compared to previous models.With the model's feature,we proposed different algorithms for the different D2D reuse number.When the amount is small,we can get the optimal result from the intersection point of the feasible region.When the amount is large,we propose an iterative method which is based on the interaction among the power range,optimization result and users' SINR.The method makes the solution toward the optimal direction with constraints satisfaction.The results show that the performance of proposed algorithms is equivalent to that of the interior point method,moreover,it can lead the success probability of power control to a higher level by fast convergence.In addition,users' QOS has little impact on the performance,but,it and D2D reuse number have large effects on the success probability.We summarize in the end,and illustrate the future works.