认知网络中基于网络辅助的速率控制方法

Network-assisted optimal rate control methods in cognitive networks

面向多速率业务的认知网络,为了克服其动态性并实现速率控制的自主性,在改进IEEE 1900.4系统架构的基础上提出了速率控制框架。探讨不同层面不同尺度的速率控制方案。重点研究终端侧短期实时速率控制的问题。首先,基于非合作博弈提出分布式自主速率选择方法;进而,基于合作博弈提出基于网络辅助的中心式速率分配方法。仿真结果表明,后者较前者获得近60%的效能和一定的公平性改善,同时也验证了定价函数设计能够有效地改进公平性。

Orienting the multi-rate cognitive networks, overcoming the typical characteristics of dynamics to implement the autonomy and rationality of rate control, first the rate control framework based on the presented improved IEEE 1900.4 architecture was proposed. Meanwhile, different scaled rate control schemes on different levels were investigated. Then, the real-time rate control problem on the terminal we concentrate on. Most importantly, both the distributed rate selection of TRM towards RNRM and the centralized rate allocation of RNRM to TRM were investigated. Simulation results show that the latter can achieve 60% utility and certain fairness improvements, in addition, the rationality and fairness guaranteed by the newly-built pricing function is verified.