Study On Data Aggregation Technology And Routing Protocols Of Wireless Sensor Networks

Wireless sensor networks with low-cost cabling, remote control, high detection accuracy,fault tolerance, easy diagnostics and maintenance advantages, has broad application prospectsin the fields of military, environmental science, medical health, space exploration, smarthome and intelligent transportation. Meanwhile, wireless sensor networks is also aresource-constrained system with low computing power, storage capacity and transmissioncapacity, so in order to fully develop the advantages of the wireless sensor and cleverly useits limitations, designing appropriate routing protocols and conducting data aggregationproblem become very important.The dissertation first introduces the research background, significance, research progress,and routing protocols and key technologies of wireless sensor network. Then it discusses thebasic structure, characteristics and classification of wireless sensor networks and comparesthe currently more typical routing protocols. Based on the analysis of the existing networklayer routing protocol and key technologies, the main research directions of this article are thenetwork layer routing protocol and key technologies--data aggregation technology in wirelesssensor networks. The main contributions and innovations are as follows:1. A clustering hybrid routing protocol based on time-component for wireless sensornetworks.Sensor node has a strict energy constraints and battery usually can not be replaced.How to rationally make use of the limited energy in the long time term is becoming theprimary objective for designing routing protocols for wireless sensor networks. Sensor nodeclustering can help to reduce the energy consumption of the nodes and improve networkscalability, such as the LEACH protocol. Based clustering routing protocol, it proposes ahybrid model routing protocol based on time-component clustering (HRPTC). Energyconsumption minimum formula can help to find a time split point, before the time split point,multi-hop routing model is used for reducing the transmission distance away from the thesink node sensor to cluster head, which saves nodes energy and energy consumption. Afterthe time-division point, single-hop routing protocol is used to mitigate "workloads" of sensornodes that near the sink node for balancing energy consumption in wireless sensor networks.Simulation results show that the hybrid routing model network node number of survivingincreased about40%and60%than a single multi-hop routing and single-hop routing. Thehybrid routing model extends the start time of the first dead node, and obviously improves thenetwork life cycle. 2. An energy efficient heterogeneous clustered routing protocolBased on LEACH routing algorithm, it studies the heterogeneous networks includecluster heads election and routing tree establishment and proposes an energy efficientheterogeneous clustering algorithm-EHCA accordingly. The advantages of this algorithmfully take into account the probability of cluster head election, average remaining energy anddistance factors. The use of weighted makes the energy consumption more balanced. Routingtree establishment between cluster heads also greatly reduces the data transfer energy, therebyit extends the life cycle of the network. The simulation results show that EHCA extends thelife-cycle about95%-233%more than LEACH.3. A data aggregation technology based on the structure of the complete ternary tree anduse regression function model in wireless sensor networksThrough the establishment of the multiple variables complete ternary tree, a novelmethod of data aggregation:Complete ternary tree-based data aggregation using regressionmodel-CTETREG is proposed. Three sets of coefficients are calculated by regressionfunction. Each child node passes them together with the coordinate range to the parent node,the parent node of the information uses these packets and received data from non-tree nodesto do further data aggregation, and then get a new set of coefficients. It continues passing thedata packet to its parent node until the data packet is arrived at the roots-sink node. Such dataaggregation strategy reduces the number of the data transfer, energy savings, and can performfast query. The simulation results show that percentage error rate is5%, less than5.6%of thecomplete binary tree, and node output of the data length is a constant within an acceptablerange. The compression ratio of the data is substantially a fixed value and influence of thedepth of the ternary tree is very small, it greatly increases the scalability of the network and isconducive to the application of this approach in a wider range.Finally, it concludes the whole dissertation and proposes further research directions.