Research On Energy Efficiency Optimization Of APTEEN Routing Protocol In Wireless Sensor Networks

Wireless sensor networks are composed of nodes with limited energy,through improving the routing protocol to reduce the communication energy consumption of network,prolong the network lifetime has been the key problem of research.APTEEN routing protocol with cluster structure is widely used because it has the characteristics of both active and reactive routing.However,APTEEN routing protocol still has the problems of low network energy utilization,unbalanced network energy consumption,and massively redundant information in the data transmitted in the network.Based on the in-depth study of APTEEN routing protocol,bat algorithm and deep autoencoder,this paper improves APTEEN routing protocol in selection of cluster head,data transmission and data fusion.Aiming at the problem of APTEEN routing protocol that selection of cluster heads randomly and multi-hop communication mechanism between clusters leads to the network energy consumption is too fast,and the nodes close to the base station die early because of energy depletion,this paper weighs the factors such as energy and distance to propose an improved APTEEN routing protocol based on bat algorithm(BA-APTEEN).The bat algorithm cannot be directly used to optimize the discrete problem of selection of cluster heads since bat individuals search in the continuous domain.Therefore,by transforming the real coding of bat individual position vector into binary coding,replacing the original local disturbance in the continuous search space with an improved exchange mutation operator,and transforming the explosion operator of fireworks algorithm into binary version to incorporate into the global search stage,a hybrid binary firework bat algorithm(HBFBA)is proposed.BA-APTEEN routing protocol adds remaining energy of nodes,energy weighted distance between nodes and adjacent clusters,and distance within the cluster as selection factors to find the optimal arrangement of selected cluster heads.According to the distance between cluster head and base station,inter cluster communication adopts a hybrid communication mode with single hop and multi hop,multi hop communication considers the energy factor when selecting the relay node to balance the energy consumption of inter cluster communication.Simulation results show that contrasted with bat algorithm,HBFBA significantly improves the convergence speed and search accuracy,BA-APTEEN routing protocol is 1.65 times of APTEEN routing protocol network lifetime,and improves the network energy efficiency.Aiming at scenarios of sensor nodes sensing single type data and sensing multiple types of data,the Application of Data Fusion based on Stacked Autoencoder to APTEEN routing protocol(SAE-APTEEN)and the Application of Data Fusion based on Convolutional Autoencoder to APTEEN routing protocol(CAE-APTEEN)are advocated respectively to decrease the energy waste of redundant data transmission in WSN.SAE-APTEEN and CAE-APTEEN design the opportune moment when the base station sends the deep autoencoder parameters to adapt the characteristics of the ST and HT control node sending data.At the same time,the data fusion algorithm is deployed in the cluster head and the cluster member nodes hierarchically pursuant to the clustering structure to limit the data redundancy in the cluster and get the small packets containing the effective information.The simulation results show that SAE-APTEEN and CAE-APTEEN,obtained by improved data fusion,BASAE-APTEEN and BACAE-APTEEN,obtained by integrating selection of cluster head and data transmission improvement measures,prolong the network lifetime contrasted with APTEEN.