基于社会网络分析的无线Mesh网络资源安全使用机制研究

The issues of security and resource consumption become the most significant concerns when any new communication technique is applied in the real world. Recently, Social Network Analysis (SNA) emerges as a novel and effective approach, which studies the relationship between actors. SNA can quantify and analyze the relationship exactly and it is considered as the most simple and convincing approach, especially for the problem of behavior structure. In this project, we focus on the research of secure resource utilization issues in real Wireless Mesh Network (WMN) circumstance. We build the effective and efficient secure resource utilization schemes and algorithms based on SNA in WMN, to improve the security of resource utilization and reduce resource consumption. The main research contents include: Firstly, we present a low resource consumption attack detection mechanism to solve the malevolent resource consumption issue based on hot data mining, considering the robustness and adaptability. Secondly, we propose a node collaboration based incentive mechanism based on clique, which can enhance the cooperation between nodes and prevent excessive resource saving. Finally, we develop an optimization strategy based on link risk assessment to improve the resistance ability of network, which can give more comprehensive support on secure resource utilization in WMN.

任何通信技术一旦进入实用阶段，安全和资源消耗问题就成为最受关注的课题之一。近年来，社会网络分析技术作为一种研究行动者之间关系的分析方法，能够精确量化和分析各种关系，被视为研究行为结构问题最简单明朗、最具有说服力的研究手段。本项目聚焦实际应用场景下无线Mesh网络的资源安全使用问题，引入社会网络分析技术，从资源恶意消耗和过度保护两方面出发，建立高能效的无线Mesh网络资源安全使用算法及策略，在提高资源使用安全性的同时，降低网络的资源消耗。研究内容包括：探索基于攻击热点发掘的低能耗攻击检测机制，兼顾鲁棒性和适应性要求，解决资源恶意消耗问题；攻关基于凝聚子群的节点协作激励机制，增强节点的协作主动性，预防资源过度保护问题；研究基于链路风险评估的优化策略，更深入地保障资源合理使用，进一步提高网络抗攻击性能，形成对无线Mesh网络资源安全使用技术更全面的支持。

本项目聚焦实际应用场景下无线Mesh网络的资源安全使用问题，建立高效能的无线mesh网络资源安全使用算法及策略。提出了基于互信息量的监测节点选择算法，该算法综合考虑了节点拓扑位置和节点间的联系紧密程度，利用信息论理论定义节点重要性，从而确定监测节点；并以信息相似度作为边权值，建立影响力模型，将联系频繁的节点分到同一个监管区域以避免冗余的检测开销，保证检测的实时性和监测信息的有效性。在此基础上提出了基于流量分析的监测节点动态选择算法，该算法在选取监测节点时，除了考虑拓扑结构的影响，还考虑到节点的繁忙程度，进而建立了节点中心值全局模型；并基于贝叶斯模型，按照节点的流量转发情况确定了影响力边权值，建立基于流量的影响力模型。另外，引入节点风险评估机制，研究优化的选路策略以提高网络可用性。最后搭建由Mesh路由器组成的测试平台，对其中的关键技术进行评测和验证。