Construction of Heterogeneous WSN Based on Multi-sink Nodes and Its Fundamental Applied Research in Mine Environment

Author:Hu Chang Jun

Supervisor:yuan shu jie

Database:Doctor

Degree Year:2018

Download:15

Pages:118

Size:8744K

Keyword:

In view of the shortcomings of single sink sensor network in mine environment,this dissertation studies the multi-sink sensor network,mainly involving the optimal deployment of multi-sink nodes and multi-sink network routing technology.The main research results are summarized as follows:(1)The location of sink nodes has a great influence on the performance of sensor networks.The basic particle swarm optimization algorithm is improved.An adaptive optimization algorithm based on particle swarm is proposed to optimize the deploy-ment location of multiple sinks.The inertia weight coefficient in the standard algorithm decreases linearly with the number of iterations.It can not reflect the degree of evolutionary optimization and approximation to the optimal value in real time.The real-time search performance is poor.The evolution and aggregation characteristics of particles are analized in the particle iteration process,and evolutionary factors and aggregate factors are introduced in the inertia weight coefficient,which increase the adaptability of the algorithm.At the same time,in order to avoid the overlap of particles in the iterative process,the perturbation strategy is added to ensure the diversity of particles,which is more conducive to the search of the optimal solution.The fitness function of the particle and the iterative process of the algorithm are designed,and the optimal position of the specified number of sink nodes under N sensor nodes is determined.(2)A non-uniform clustering routing protocol for underground sensor networks based on multi-sink nodes is proposed,which includes communication range negotiation between sink nodes and clustering protocol.In the mine environment,the sink nodes sometimes can not be deployed in a pre-optimized or designated way.This paper studies the negotiation process between sink nodes when they are placed anywhere in the roadway.The neighboring two sink nodes estimate the distance to the farthest point in the area according to their position and negotiate with each other to determine their coverage radius to ensure seamless coverage of the monitoring area.The competition radius formula of node competition cluster head and the underground non-uniform clustering algorithm are proposed,at last the communication overhead of cluster is analyzed.(3)An adaptive and energy-balanced distributed data collection method is proposed by applying hybrid communication mode to underground environment.The monitoring area is divided into several rings,and the outer nodes transmit data to the sink node by multi-hop or single-hop mode through the inner forwarding node,or send data to the node in the ring closer to the sink.The relationship between the width of the ring and the communication radius of the node is determined by theoretica analysis,and the detailed steps of forwarding information from the outer nodes to the sink nodes are specified.The expression of the forwarding node timer is defined and the relevant parameters of the timer and the density requirement of the sensor nodes are determined by theoretical analysis.(4)A distributed fault-tolerant routing algorithm based on energy-balanced is proposed to increase the robustness of routing,with the premise of reduced and balanced network energy consumption.The cost function of the forwarding node is defined to determine the most suitable forwarding node in the routing.If the cluster head node dies in the stable phase,the member-node in the original cluster join in neighboring cluster according to the nearest principle,thus avoiding the overhead of re-clustering.The set of backup forwarding nodes is defined,and the process of continued routing through the set of backup forwarding nodes when nodes fail in routing process is given in detail to improve the fault tolerance of routing.Figure[47] table[16] reference[124]