能量采集认知无线网络中频谱效率和能量效率优化研究

The energy harvesting technology can harvest renewable energy from ambient sources that powers wireless communication devices. This novel idea marks the revolution of the operating mode for wireless communication networks, its motivation is to solve the energy scarcity problem and promote development and implementation of green communications. Applying energy harvesting technology to cognitive radio networks is a novel idea to solve the energy and spectrum scarcity problem, it helps to achieve the grand goal that 5G communication networks will have 1000 times higher capacity compared to 4G communication networks. However, the subsequent challenges should not be underestimated. These challenges include: 1) The ambient energy sources include RF (Radio Frequency) and non-RF sources. As energy supply for both of them is random, it is difficult to establish energy harvesting models; 2) Utilizing the harvested energy to power the cognitive radio networks needs to satisfy the constraint that the harvested energy is larger than the consumed energy, thus it is difficult to design the corresponding sensing-access and power allocation mechanisms; 3) These is a tradeoff between spectrum efficiency and energy efficiency, thus it is difficult to solve the optimization problems; and so on. Therefore, this project will investigate the energy harvesting and consumption models in wireless cognitive networks. On the condition that the constraint of harvesting energy and energy consumption is satisfied, we design the sensing-access and power allocation mechanisms to achieve collaborative optimization of spectrum efficiency and energy efficiency for cognitive radio networks, thus to provide technical contributions for the development of 5G networks and green communications.

能量采集技术利用周围环境的可再生能源为无线通信设备供电。这一新思想表征了无线通信网络工作模式的变革，其动因是解决能源紧张问题，促进绿色通信的发展与实现。将采能技术用于认知无线网络，是解决能源和频谱资源短缺问题的新思路，有利于实现5G通信网络的容量1000倍于4G通信网络这一宏伟目标。梦想是美好的，但随之而来的挑战也不容小觑。例如：1）周围环境能源包括射频和非射频，两者的供给均具有随机性，导致采集模型建立难；2）利用采集的能量为认知无线网络供电，需要满足采集能量大于消耗能量这一约束，导致相适应的感知频谱接入和功率配置机制设计难；3）频谱效率和能量效率的提升相互牵制，导致优化问题求解难；等等。为此，本项目将研究认知无线网络的能量采集与消耗模型，在满足能量因果约束的前提下，设计感知频谱接入和功率配置机制，实现认知无线网络频谱效率和能量效率的协同优化，为5G网络和绿色通信的发展提供技术贡献。

将能量采集技术用于认知无线网络，可有效解决能源和频谱资源短缺问题，并且符合绿色发展理念。本项目首先聚焦于能量采集认知无线网络（EH-CRN），研究了提升其频谱效率和能量效率的模型、协议、策略、方法；考虑到用户间同道干扰和敌方恶意干扰必将导致通信系统损失谱效和能效，本项目还研究了基于盲源信号分离消除干扰的理论和系列算法。上述研究取得了多项创新性成果。.1)优化EH-CRN谱效与能效的频谱感知、数据传输、能量采集的协议设计和接入策略.提出了融合射频能量采集和消耗的全双工认知无线网络模型，设计了次级用户同时进行频谱感知、数据传输、能量采集协议；提出了协作频谱感知接入、位置信息辅助频谱感知接入、优化次用户发射功率和融合准则的功率配置机制等策略和方法，实现次级网络谱效、能效、采集效率协同最大化。.2）提升谱效与能效的认知无线供能传感网 (CR-WPSN)信道资源分配策略 .区别于传统单汇聚节点认知传感网络研究，面向更为实际的大规模多簇网络，综合考虑网络干扰复杂关系、用户决策强耦合、能量约束以及能量队列动态性等因素，研究了提升网络能效-谱效二维资源利用率的信道分配策略；在用户干扰关系描述、多簇网络谱效能效优化、节点间无交互信息下的稳健优化等方面形成创新，给出了：基于二元图博弈的局部交互多簇CR-WPSN频谱分配方法、基于加权有向超图的云辅助多簇CR-WPSN频谱分配方法、无信息交互下的多簇CR-WPSN稳健频谱分配方法和信道频谱与功率联合稳健分配方法。.3）减少谱效与能效损失的干扰消除方法.用户间共信道干扰和敌方恶意干扰必将导致通信网络谱效和能效受损。因此本项目拓展研究思路，重点研究了不需要额外频率资源和功率资源的盲源分离（BSS）干扰消除方法，根据源信号特性和信道混合特性，提出了多种解决方案；理论分析和仿真分析的结果表明，基于BSS的干扰消除方法，能有效阻止干扰导致的谱效和能效损失。