Research On Millimeter Wave D2D Communication Resource Allocation Schemes

Millimeter wave technology and D2 D communication are key technologies for future 5G communication systems.These two technologies are combined to meet the needs of users for high transmission rates and excellent user experience.However,with the dramatic increase in the number of user devices and the types of devices,there are a large number of devices within a limited range,which will result in complicated interference conditions.Therefore,how to scientifically and effectively characterize the interference situation and manage the communication link efficiently and reasonably to optimize the transmission rate of the system has become one of the key issues for the wide application of this communication method.Based on graph theory and optimization theory,combined with the characteristics of millimeter wave D2 D communication,this thesis studies the resource allocation problem in millimeter wave D2 D scene from reducing sidelobe interference and reducing the side effect of antenna heterogeneity.The specific research contents are as follows:(1)Firstly,the key issues to be solved in 5G communication are discussed,and the current status and main problems of current communication are summarized.It points out the characteristics of millimeter wave technology and D2 D communication,and the combination of the two technologies is important for solving the spectrum saturation problem and offloading the base station traffic.At the same time,the resource allocation scheme related to millimeter wave D2 D is summarized,and the strategies applicable to this scenario are few.Combining the characteristics of millimeter wave technology and D2 D communication,the research difficulties of further optimizing the performance of the scene system are pointed out,and the research direction is proposed.(2)Secondly,for the characteristics of high-density distribution of devices in millimeter wave D2 D networks,the sidelobe interference reduced resource allocation algorithm is designed.By analyzing the transmission characteristics of millimeter waves,the differential effects of the main and side lobes of the millimeter wave on the system interference are evaluated.On the basis of overcoming the main lobe interference,the resource allocation algorithm which can reduce the sidelobe interference is further proposed.Combined with the spatial sparsity of millimeter waves,the algorithm introduces a scheduling scheme using time and space partitioning.At the same time,the conditions for concurrent transmission of D2 D communication are redefined.On the one hand,the collision state is judged by the proprietary area and the received power,not just the distance and angle.On the other hand,the expression of the collision matrix is modified to introduce a new value to represent potential sidelobe interference.To quantify the degree of interference between pairs of communications,a parameter called decision threshold is introduced,which is defined by the ratio of the sum of potential interference power to the theoretical received power.(3)Finally,for the antenna heterogeneity of millimeter wave D2 D networks,a joint resource allocation algorithm for heterogeneous antenna arrays is designed.In order to study the influence of antenna heterogeneity on system performance,the existing antenna gain model is compared and the roll-off characteristics of the millimeter-wave antenna gain model are analyzed.The model that best matches the real scene is selected.Based on this model,the transmission angle is considered as a factor that can be adjusted and allocated,and it is considered that the interference to other links can be greatly reduced by sacrificing the transmission rate of a part of the target link.The interference graph is optimized,and three attribute values are designed to represent the initial values and solution sets of the network.Then the time slot and transmission angle joint resource allocation optimization problem is given,aiming at maximizing the total system throughput.By dividing the cluster and the virtual cluster,the local optimal solution is obtained by the greedy algorithm in the cluster.At last,the approximate optimal solution to the problem is obtained.