Research Of QoS Algorithm Based On Various Traffic Transmissions For Wireless Sensor Networks

Wireless sensor networks consists of a group of sensors constructed in Ad Hoc model. It is introduced to detect, to gather and to deal with information of detected objects in certain regions, then send the information to users. Wireless sensor networks has its own characteristics: energy of the network is limited; data-centric and end-to-end communication mode; high degree of data redundancy; multi-hop wireless transmission model with densely distributed nodes; multi-user, multi-task, different types of data traffic. Now the dominating routing algorithms of WSN that have been proposed aim at the best effort model for single service. But along with the continuous deepening and expanding application, different types of traffic require different types of QoS(Quality of Service). So it is necessary to study special communication protocol for WSN to provide different types of QoS for different types of traffic.The thesis first proposed a residual Energy-based Directed Diffusion algorithm(REDD) which chooses transmitting nodes according to the residual energy of neighbor nodes. The improved algorithm sets up minimum hops field according to the hops of different sensor nodes to the sink node. At the course of setting up minimum hops field, every sensor node uses two arrays NeighborLow and NeighborHigh to save different neighbor nods. NeighborLow is used to save neighbor nodes which hops is 1 less than the node's to the sink node, and NeighborHigh is used to save neighbor nodes which hops is 1 more than the node's to the sink node. At the same time ,every node saves its neighbor nodes' residual energy. Before transmitting data packets, the node estimates needed energy to transmit the data packets, and compares to the residual energy of neighbor nodes in NeighborLow. Then transmitting data packets to the nodes which have enough residual energy. At the meanwhile, transmitting its own residual energy information to the neighbor nodes in NeighborHigh in order to update the residual energy information. Then based REDD algorithm on the above, the thesis proposed a QoS architecture for providing different types of traffic with different types of QoS-EMHNS (Energy and Minimum Hops-Based NeighborNode Scheduling).In this architecture, d various traffics are sent by the path that satisfies its QoS requirements and enough residual energy. So it can achieve the minimum time delay for real-time traffic transmission, and taking into account the network energy load balance when the best-effort traffic transmits. The simulation results indicate that this architecture can achieve lower time delay for real-time traffic than Directed Diffusion (DD) algorithm, and extend the lifetime of the network.