Research And Implementation Of Software Defined Wireless Mesh Network Networking System Based On WiFi6

Wireless Mesh network is a multi-hop network,which has the advantages of high speed,easy networking,self-healing and self-organization.Compared with traditional wireless networks,this kind of network with low cost,high scalability,flexible networking and easy maintenance has been widely used in home,enterprise,smart power grid,industrial Internet of things and other scenarios.With the certification of the sixth generation wireless network technology(Wi Fi6),wireless speed and wireless resource utilization have greatly improved,and network stability under dense signals has also been enhanced.These performance improvements are very suitable for wireless Mesh networks,which can effectively improve routing speed and stability.Although Wi Fi6 based wireless Mesh networks have excellent performance,with the explosive growth of user volume and diversification of business types,if traditional wireless ad hoc network routing algorithms are directly applied to such network scenarios,there will be many problems such as network congestion,high packet loss rate,and excessive end-to-end latency.In response to the above issues,this thesis proposes a Wi Fi based wireless network algorithm based on Wi Fi6,and implements the software and hardware development and performance testing of the wireless Mesh network networking system through the Open Wrt system.The specific work is as follows:(1)Aiming at the problem that the wireless Mesh network cannot route on demand according to the service type,which makes the service Qo S difficult to guarantee,a wireless Mesh routing algorithm based on cross layer information awareness is proposed.The data link layer parameter signal strength is combined with the network layer parameter delay and packet loss rate as the routing metric to achieve more accurate link quality assessment.At the same time,the idea of software defined routing is used to send the routing policy uniformly by the controller,Integrating network dynamics with business dynamics into the routing process to achieve on-demand routing based on different business types.The test results show that compared with the wireless Mesh network using BATMAN-ADV protocol,the algorithm can reduce the Transmission delay by about 15%,and improve the service throughput by about 10%,effectively guaranteeing the Qo S of the service.(2)After analyzing the shortcomings of traditional wireless Mesh networks such as network congestion,high packet loss rate,and difficulty in ensuring reliable transmission of services,a wireless Mesh reliability routing algorithm based on deep reinforcement learning is proposed.The algorithm in the field of deep reinforcement learning is introduced into network intelligent routing,and the optimal path is adaptively selected based on network state to improve routing reliability.The controller continuously collects various parameters that affect the reliability of the Mesh network,such as node congestion level,network frequency band,latency,etc.The DQN algorithm processes these parameters to ultimately obtain the most reliable routing strategy in the current network,in order to improve the reliability of routing.The test results show that compared to traditional algorithms,this algorithm performs well in dealing with network congestion and business packet loss rate,with a decrease of about 20% in packet loss rate.(3)We have completed the design of a software defined wireless Mesh network networking system and developed software and hardware using the latest Qualcomm chip’s Wi Fi6 router.The proposed routing algorithm was tested based on the constructed wireless Mesh network.The test results show that the software defined wireless Mesh network based on Wi Fi6 performs well in terms of throughput,latency,and packet loss rate,improving the overall performance of the Mesh network;And the software defined routing method can effectively ensure the Qo S of different services,achieving the expected design goals.