摘要
针对机会网络中缓存资源利用效率低的问题,提出一种消息替换风险动态感知的缓存管理策略,该策略综合考虑多维网络状态参数,其中包括消息的已生存时间、跳数、副本数及区域概率,以分布式的方式估计给定消息的投递状态,并采用K-均值聚类法动态地评估给定消息的替换风险,进而为不同投递状态的消息合理地预留资源。仿真结果表明,提出的缓存策略与其他缓存策略法相比,能够有效提高消息投递率,并改善网络负载率。
Aiming at the low efficiency of cache management in opportunistic network, A novel cache management strategy based on dynamical message replicating risk evaluation is proposed. The transmitting status of a given message is predicted by considering four parameters, including the time to live already, hop count, copies of message and regional delivery prob- ability. Consequently, the risk of message replicating can be determined by K-average clustering, and the limited resources can be reserved for the proper message. Compared with other cache management strategies, the simulation results show that the proposed strategy can efficiently increase the message delivery ratio while the overhead is improved also.
出处
《重庆邮电大学学报(自然科学版)》
CSCD
北大核心
2014年第2期233-237,共5页
Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)
基金
国家自然科学基金(61371097)
重庆邮电大学青年自然科学基金(A2012-93)~~
关键词
机会网络
缓存管理
消息替换风险
opportunistic network
cache management
message replicating risk