Reserch On New Mechanism And Key Issues For Energy Efficiency In Heterogeneous Wireless Sensor Networks

Heterogeneous Wireless Sensor Network(HWSN),due to its ability of self organizing,real time sensing,cooperative processing and wireless transmission,has continuously developed in many fields.However,since the resource,especially the energy resource,of HWSN is limit;energy efficiency is a great challenge for the network design and also the key issue of HWSN's research.The energy efficiency of MAC and routing layer play an important role in the whole network since they are located in the lower layer of the protocol stack.Therefore,how to design MAC and routing protocols with high energy efficiency is of great importance.As for the HWSN with image transmission,how to lower the energy consumption during image transmission is also a key issue since the image transmission consuming a large amount of energy.In this paper,we study the energy efficiency network which consists of ordinary nodes and special nodes who transmits image and powered by solar energy,in the following aspect: media access control(MAC),clustering routing and image transmission.The specific content and contribution of the thesis includes:In the MAC layer,we propose an adaptive MAC protocol named DTI-MAC based on characteristic of MAC in HWSN.DTI-MAC is based on the dynamic duty cycle design,in which the network can adjust sleeping time based on the network's traffic load in order to save energy.In DTI-MAC,duty cycle is the dynamic function of network's traffic load,which can be predicted by a data estimation algorithm based on nodes clustering and Genetic Programming.The simulation results show that DTI-MAC can reduce the transmission delay and increase the network's throughput while saving energy.In the network layer,to alleviate the high energy consumption of cluster head,we propose a node clustering routing protocol named DEBCR which is based on nodes' energy,deployment density and the bandwidth of links.In DEBCR,the clustering of nodes is based on the weight consideration of the residual energy,the energy harvesting capability of solar powered nodes and deployment density of nodes.The node with the largest weight is selected as the cluster head.The routing schemed among cluster head is based on the Dijkstra routing scheme which can find the shortest route from the source node to the sink node.The DEBCR adopts a novel complex scale path metric criterion based on transmission power,residual energy,harvesting energy and links' bandwidth.The novel criterion set different route weight based on different Type of Service(ToS)to ensure both energy efficiency and transmission delay.Simulation results show that the DEBCR can adaptively select a proper route among multiple routes to balance the networks' load,optimize the energy consumption and prolong the life time of the network.In the image transmission,we propose a Multi Hop Image Transmission Scheme(MHITS)based on 3D object retrieval and recognition algorithm.MHITS reduces and balances the network's energy consumption by reducing the number of transmission image,and cooperative compressing the image.The nodes who responsible for data transmission conduct the preprocessing of the 3D object retrieval based on the reconstruction error and the weight of query images.After the preprocessing,nodes only transmit a part of the image which reduces the energy consumption.According to the hop count and distance,the image node determines whether or not the image should be compressed before transmission.If image compression is the better way,the compressing is cooperatively conducted by the ordinary nodes,and the compressed image is transmitted to the sink.Simulation results show that MHITS can solve the problem of energy hole in the HWSN,and it is well suited for long range image transmission for the large scale wireless sensor network.