基于Moran过程的认知中继网络多策略信任演化模型  

Multi-strategy Trust Evolution Model for Cognitive Relay Network Based on Moran Process

作　　者：邹世辰[1] 王慧强[1] 冯光升[1] 吕宏武[1] 

机构地区：[1]哈尔滨工程大学计算机科学与技术学院,哈尔滨150001

出　　处：《小型微型计算机系统》2014年第10期2209-2214,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61370212)资助;教育部高等学校博士点基金优先发展领域项目(20122304130002)资助;中央高校基本科研业务费专项资金项目(HEUCFZ1213;HEUCF100601)资助;黑龙江省自然科学重点基金项目(ZD201102)资助

摘　　要：认知中继网络通过引入中继传输,可以有效改善网络通信的性能.为了提高认知中继网络系统的自适应性,改善网络通信的整体质量,抑制网络节点的自私行为,结合演化博弈理论提出一种基于Moran过程的认知中继网络信任演化模型.首先建立认知中继网络信任演化基本框架,设计基于区分贡献度的实体全局收益函数,为信任演化模型的建立提供基础,以增强信任演化模型的自适应性;然后提出一种Moran多策略信任演化模型,激励信任策略的演化,以增强对信任演化预测的准确性.实验结果表明,该模型能够准确地反映认知中继网络中节点信任行为的特点,推动网络节点行为向协同合作策略的演化,使节点实体的信任演化快速收敛到稳定状态,以此提高网络系统的可信度.Cognitive relay network can improve the performance of network communication by introducing cognitive relay transmission into network. Combined with evolutionary game theory, a trust evolution model for cognitive relay network based on Moran process is proposed in order to improve the self-adaptive performance and the overall quality of communication of network system, so that the network system can suppress the generation of selfish nodes effectively. Firstly, the cognitive relay network trust evolutionary frame- work is established,providing a framework basis for establishing of the trust evolution model. Then an entity global revenue function based on dividing contribution degree is designed, so as to enhance the self-adaptive performance of the trust evolution model. Finally, a multi-strategy Moran trust evolution model is proposed to inspire the evolution of trust policy, enhancing the accuracy of the forecasts for the evolution of trust. Experiments results show that the model can accurately reflect the trust behavioral characteristics of cognitive relay networks, and effectively promote the network node behavior strategy evolution towards the trend of cooperation. The trust evolu- tion of node entities can convergence and reach a stable state fast. The model can optimize the cognitive relay network trust benefits, thus improving the overall dependability of the network system.

关 键 词：认知中继网络 区分贡献度 动态演化 多策略 信任演化 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]