通过将认知无线电(cognitive radio,CR)技术应用到车载自组织网络(vehicular ad hoc networks,VANETs)(也称车联网)中,认知无线车载自组织网络(CR-VANETs)可以缓解频谱资源稀缺问题,有效提高车对车通信的频谱资源利用率.由于车辆的高速...通过将认知无线电(cognitive radio,CR)技术应用到车载自组织网络(vehicular ad hoc networks,VANETs)(也称车联网)中,认知无线车载自组织网络(CR-VANETs)可以缓解频谱资源稀缺问题,有效提高车对车通信的频谱资源利用率.由于车辆的高速移动性以及认知无线电频谱资源的动态特性,使得传统的认知无线电网络或车载自组织网络中的路由协议无法直接应用到CR-VANETs中.目前,针对CR-VANETs的路由研究相对较少,如何最大效率地利用有限的频谱资源,同时降低跳数过多带来的频谱资源浪费,仍然是一个有待解决的问题.为此,提出了一种CR-VANETs中联合路由调度方案,结合了有限频谱资源调度研究与最小化路由跳数的优化目标.首先,建立了CR-VANETs中的网络模型和基于车对车通信的频谱感知模型,预测车辆间有效接触时间和频谱可用概率.其次,通过这些参数定义出通信链路消耗,并由此得出权衡链路质量的权重因子.通过分析优化目标,将其转化为有限频谱资源约束下的最小化路由跳数问题,并证明该问题为NP难问题.然后,针对这个联合路由调度问题提出一种混合启发式算法,结合了粒子群优化算法的快速收敛性和遗传算法的种群多样性,对有限频谱资源进行调度,同时优化路由跳数.最后仿真实验结果表明,与现有的CR-VANETs路由研究比较,有着更优的路由跳数并使其保持在一个相对稳定的值.展开更多
The performance of the graph-based scheduling for device-to-device communications overlaying cellular networks is studied. The graph-based scheduling consists of two stages, the frequency assignment stage and the time...The performance of the graph-based scheduling for device-to-device communications overlaying cellular networks is studied. The graph-based scheduling consists of two stages, the frequency assignment stage and the time slot scheduling stage. For such scheduling, a theoretical method to analyze the average spectrum efficiency of the D2D subsystem is proposed. The method consists of three steps. First, the frequency assignment stage is analyzed and the approximate formula of the average number of the D2D links which are assigned the same frequency is derived. Secondly, the time slot scheduling stage is analyzed and the approximate formula of the average probability of a D2D link being scheduled in a time slot is derived. Thirdly, the average spectrum efficiency of the D2D subsystem is analyzed and the corresponding approximate formula is derived. Analysis results show that the average spectrum efficiency of the D2D subsystem is approximately inversely linearly proportional to the second- order origin moment of the normalized broadcast radius of D2D links. Simulation results show that the proposed method can correctly predict the average spectrum efficiency of the D2D subsystem.展开更多
文摘通过将认知无线电(cognitive radio,CR)技术应用到车载自组织网络(vehicular ad hoc networks,VANETs)(也称车联网)中,认知无线车载自组织网络(CR-VANETs)可以缓解频谱资源稀缺问题,有效提高车对车通信的频谱资源利用率.由于车辆的高速移动性以及认知无线电频谱资源的动态特性,使得传统的认知无线电网络或车载自组织网络中的路由协议无法直接应用到CR-VANETs中.目前,针对CR-VANETs的路由研究相对较少,如何最大效率地利用有限的频谱资源,同时降低跳数过多带来的频谱资源浪费,仍然是一个有待解决的问题.为此,提出了一种CR-VANETs中联合路由调度方案,结合了有限频谱资源调度研究与最小化路由跳数的优化目标.首先,建立了CR-VANETs中的网络模型和基于车对车通信的频谱感知模型,预测车辆间有效接触时间和频谱可用概率.其次,通过这些参数定义出通信链路消耗,并由此得出权衡链路质量的权重因子.通过分析优化目标,将其转化为有限频谱资源约束下的最小化路由跳数问题,并证明该问题为NP难问题.然后,针对这个联合路由调度问题提出一种混合启发式算法,结合了粒子群优化算法的快速收敛性和遗传算法的种群多样性,对有限频谱资源进行调度,同时优化路由跳数.最后仿真实验结果表明,与现有的CR-VANETs路由研究比较,有着更优的路由跳数并使其保持在一个相对稳定的值.
基金The National Natural Science Foundation of China(No.61571111)the National High Technology Research and Development Program of China(863 Program)(No.2014AA01A703,2015AA01A706)the Fundamental Research Funds for the Central Universities of China(No.2242016K40098)
文摘The performance of the graph-based scheduling for device-to-device communications overlaying cellular networks is studied. The graph-based scheduling consists of two stages, the frequency assignment stage and the time slot scheduling stage. For such scheduling, a theoretical method to analyze the average spectrum efficiency of the D2D subsystem is proposed. The method consists of three steps. First, the frequency assignment stage is analyzed and the approximate formula of the average number of the D2D links which are assigned the same frequency is derived. Secondly, the time slot scheduling stage is analyzed and the approximate formula of the average probability of a D2D link being scheduled in a time slot is derived. Thirdly, the average spectrum efficiency of the D2D subsystem is analyzed and the corresponding approximate formula is derived. Analysis results show that the average spectrum efficiency of the D2D subsystem is approximately inversely linearly proportional to the second- order origin moment of the normalized broadcast radius of D2D links. Simulation results show that the proposed method can correctly predict the average spectrum efficiency of the D2D subsystem.
