P2P系统中节点的动态特性对搜索性能的稳定提出了很大挑战。借鉴社会学和组织学相应原理,提出拓扑演化模型捕获节点的运动规律。节点模拟人类在社会网络中的行为,根据自身和他人的兴趣变化,随搜索过程的进行,利用自身局部信息,动态、自适应地调整网络拓扑,及时反映资源分布以及搜索内容的动态变化,维持资源合理组织。在此基础上提出的智能搜索机制ITSON(InTelligent Search based on tOpologyevolutioN),利用消息转发的智能性进行路由方向决策,从而迅速定位提供资源的节点。仿真结果说明,ITSON能自动优化网络,具有良好的搜索性能和自适应特性。
In low-duty-cycle wireless sensor networks,designers have to cope with unreliable links and limited communication capacity.In this work,we propose COST,a coding scheme that leverages spatial-temporal diversity to achieve higher energy efficiency and lower delay of packet transmissions.We particularly address long sleeping intervals in low-duty-cycle networks by exploiting multi-path diversity.Specifically,we propose to employ an erasure-coding scheme to improve reliability.With respect to energy efficiency and delivery timeliness,we formulate the problem in optimal allocation of coded blocks over multiple paths,which is then proved to be NP-hard.We further propose a near-optimal algorithm to solve the allocation problem.Through extensive simulations,we evaluate the impact of network parameters and demonstrate the effectiveness of our proposal.
