MANET是一种分布式的移动自组网络，它凭借灵活、便捷和迅速的特点，成为了5G通信网络的关键技术。为增强MANET的组网性能和传输效率，项目组引入节点行为认知机制，为网络节点建立行为认知模型，提出了基于行为认知的节点行为预测方法，使网络具备节点行为感知能力。通过对用户行为特点进行分析和预测，项目组提出基于节点行为相关性的动态自适应分簇组网算法，增强了MANET网络拓扑的稳定性。在分簇的网络结构上，项目设计了基于行为认知的无线网络跨层性能优化框架，利用网络效用优化方法对MANET网络能效展开跨层多目标联合优化。为验证用户行为模型，项目组提出了合适的QoS形式化描述方法，设计了行为认知的MANET网络性能优化测试平台，从系统层面对基于行为认知的MANET模型进行仿真和评估。通过本项目的研究，可以探究行为认知对MANET性能的影响，为网络规划和优化提供解决方案，促进MANET技术的应用和推广。

MANET (Mobile Ad hoc NETwork) is a distributed wireless mobile network, which by virtue of flexible, convenient and rapid characteristics, has become the key technology of the fifth generation of communication network. It has been the industry and the academic community of widespread concern and in-depth the study. In order to enhance the networking performance and transmission efficiency of MANET, we introduce a node behavior based cognition mechanism, in which a behavior aware model for the network node is established. We also propose a node behavior prediction method, which makes the network node more intelligent. Based on the analysis and prediction of user behavior characteristics, we propose a dynamical adaptive clustering algorithm which considers node behavior correlation to enhance the stability of MANET network topology. Based on the cluster structure, we design a behavior aware optimization framework in which a multi-objective optimization method is proposed to strengthen the network performance and energy efficiency. In order to verify the node behavior model, we design a simulation platform for the behavior aware MANET, which can be used to evaluate the optimization framework from the system level. Through the research of this project, we can explore the impact of behavior awareness on MANET performance. It will provide solutions for network design and optimization, and promote the application and popularization of MANET technology.