D2D通信模式下集中-分布式混合资源优化分配方案

Hybrid centralized-distributed resource allocation optimization scheme for device-to-device communication

设备终端直连(Device-to-Device,D2D)通信可以复用小区中蜂窝用户的频谱资源,提升了频谱效率和系统容量,同时也带来了更为复杂的干扰。文中提出了一种混合的集中-分布式的复用蜂窝用户上行信道的资源分配方案,将复杂的资源分配分成信道分配和功率控制两个独立的子问题求解。信道分配由基站统一控制分配,提出了基于聚类思想的信道分配算法,用路径损耗衡量距离,把相互距离小的D2D用户聚为一类。同时提出了基于能量效率最大化的Stackelberg博弈进行功率分配,基站端通过对资源定价以保证自身的服务质量,D2D端各自追求最大化自身能量效率,可以分布式求解,并证明了纳什均衡解的存在性和唯一性。整个分配过程中,基站和D2D用户都只需要知道很少量的信道状态信息就可以完成整个资源分配方案,系统的信令开销很小。

By reusing spectrum resources of cellular users in a cell, Device-to-Device （D2D） communication improves spectral efficiency and system capacity, but brings the complicated co-channel interference. A resource allocation scheme with hybrid centralized-distributed resource allocation is proposed to reuse the uplink channel for cellular users. The resource allocation problem is divided into two independent subproblems： channel assignment and power control. Channel allocation strategy is carried out by the base station. An algorithm based on clustering is developed. The path loss is used to character the distance of clustering,and the D2D users with small mutual distance are grouped together. Based on the assigned channel information, a power allocation method with Staekelberg game is proposed to maximize energy effi- ciency. In the method, the Quality of Service of base station is ensured by pricing. Each D2D end pursues to maximize its own energy efficiency and it is solved in a distributed manner. The existence and the uniqueness of Nash equilibrium are proved. Throughout the whole procedure, the resource allocation scheme can be complemented by only a little amount of channel state information, and thus the additional signaling overhead of the system can be ignored.