Research On Clustering Routing Algorithm Based On Software-defined In WSNs

Wireless Sensor Networks(WSNs)is usually used in environments with low bandwidth and delay tolerance.In most applications,sensor nodes are limited by energy supply and communication bandwidth.In order to achieve the purpose of energy saving and maximize the life of the network,it is very necessary to find a way for the relay node to connect to the base station in the network layer.Traditional WSNs routing technology has the problems of bloated sensor node design,local optimization of routing,and dedicated network.How to solve the above problems has become the key to the research of WSNs routing technology.Software-Defined Networking(SDN)is a new network architecture,which can realize the separation of control and data forwarding planes,which can significantly improve the flexibility of WSNs and extend the network life cycle of WSNs.Therefore,we design a routing algorithm suitable for the characteristics of Software-Defined Wireless Sensor Network(SDWSN)has become particularly important.Based on SDN-WISE(Software Defined Networking Solution for Wireless Sensor Networks),this article optimizes the SDWSN routing algorithm as follows:(1)Due to the heavy tasks of traditional WSNs hierarchical routing cluster heads leading to excessive energy consumption,based on the analysis of traditional WSNs and SDWSN-based routing algorithms,a Meanshift Non-Uniform Clustering Routing Algorithm Based on SDWSN(NUR-SDWSN)is proposed.This algorithm uses the advantages of SDN numerical control separation to reduce the energy consumption of cluster heads.Firstly,the controller uses the Meanshift algorithm to complete non-uniform clustering to reduce hot spots,refresh the network regularly,and re-cluster when the node degree of any cluster head is less than the node degree threshold.Secondly,calculate the probability of the nodes in the cluster becoming a cluster head through the remaining energy,position and node degree of the nodes,sort to form a cluster head rotation table,optimize the packet header format,and rotate the cluster heads according to the energy threshold.Finally,set up Rechargeable Super Nodes(RSNs)to interact with control information,adopt the branch and bound method to achieve the optimal path between cluster heads,and then form a flow table,and send control information such as flow table and cluster head rotation table to RSNs.Through simulation comparison with LEACH,UCR and SDN-based ESR-W algorithm,in terms of cluster head distribution,the NUR-SDWSN algorithm prevents cluster heads from appearing at the edge of the network and in areas where the density of neighbor nodes is very low,and increases the coverage of cluster heads.Compared with the energy consumption per round of cluster heads and the number of surviving cluster heads per round,the energy consumption of each round of cluster heads is relatively low,and the number of survivors is relatively large,which can reduce the problem of excessive energy consumption caused by the heavy tasks of the cluster heads.(2)In order to balance the energy consumption of sensor nodes in the network,an Optimal clustering Routing Algorithm for Entropy weight Assignment based on SDWSN(ORAEA)is proposed.The controller uses the entropy weighting method to objectively assign the influencing factors of the cluster head selection such as the expected life of the sensor node,the density of neighbor nodes,the distance between the node and the base station,the node transmission energy consumption and the node load,and select the cluster head that adapts to the dynamic changes of the network.Form the cluster head rotation table to deliver the data forwarding layer.The controller uses the link transmission cost calculated by the distance between the sensor node and the base station and the energy consumption as the heuristic function of the A~*search algorithm,and obtains the multi-hop transmission route between cluster heads with the smallest transmission cost,and forms a flow table to deliver the data layer for forwarding matching.The ORAEA algorithm is relatively reasonable in terms of cluster head distribution.In terms of network coverage,due to the relatively reasonable selection of cluster heads,the coverage rate increases accordingly.In terms of total energy consumption,the balance of energy consumption of nodes is realized.