Cognitive-based Network Reconfiguration Reasoning And Decision-making Technology In Mobile Ad Hoc Networks

Cognitive mobile Ad Hoc network is a kind of cognitive network which is composed of multiple mobile nodes with equal status, it is a combination of mobile Ad Hoc network technology and cognitive network technology. Cognitive network’s intelligence mainly reflects on its reconfiguration capability, network reconfiguration technology helps the network adapt to the change of network state caused by the changes in internal and external environment, by dynamically optimizing the configuration of the network structure and network protocols, thus improve the network communication performance. The continuous development of wireless communication network is putting more and more requirement to cognitive network, as the key technology of cognitive network, network reconfiguration technology has gradually gotten the public attention with its importance and indispensability, it has become a research focus of cognitive network.The current research of network reconfiguration technology has not formed a system, and fully covered all aspects of the reconfiguration implementation, it basically distributes in the research point of the a few details. The research of cognitive mobile Ad Hoc network reconfiguration technology is still in the initial stage, the research content mainly concentrated in the optimization design of each layer protocol based on cognitive information. In this paper, we study the application and operation of network reconfiguration technology in cognitive mobile Ad Hoc network, on the basis of related literature research, we mainly study the network parameters reconfiguration and routing protocol reconfiguration of mobile Ad Hoc network.First of all, because of the diversity of network parameters, we classify the parameters of mobile Ad Hoc network, and introduce in detail the method of choosing the network parameters in different network scenarios.Second, due to the nature of network reconfiguration is a multi-objective optimization problem, we introduce the method of building optimization objective function, including the construction of the utility function of performance parameters and the selection of multi-index synthetic model.Then, we put forward a network parameter reconfiguration algorithm which is based on Bayesian Networks, Genetic Algorithm and Case-based Reasoning, including the construction of the strategy and knowledge base in some representative network scenarios based on Bayesian Network and Genetic Algorithm and the reasoning and decision-making method in new network scenarios based on Case-based Reasoning. Network parameter reconfiguration helps the network adapt to the dynamic environment by the dynamic adjustment of protocol parameters, the simulation results prove that the use of parameter reconstruction algorithm can improve network performance in dynamic environment, including the end-to-end delay and network throughput and packet deliver rate.Finally, we propose a routing protocol reconfiguration algorithm which is based on Adaptive Neural-network-based Fuzzy Inference System, the network can choose different routing protocols in different network scenarios through the construction of the performance behavior model of routing protocols, thus guarantee the performance of communication networks in dynamic environment. The simulation results show that the network routing protocol reconfiguration algorithm can select routing protocol which provides the optimal communication performance based on the network scenario, so as to verify the effectiveness of the algorithm.