Research And Improvement Of Backpressure Routing Algorithm In Wireless Sensor Networks

In recent years,with the rapid development of wireless sensor networks,sensor networks have been maturely applied in various industries,organizations,and people's daily lives.Common sensor nodes are mainly powered by batteries and are highly limited in energy.Simultaneously,as the development of sensor networks,the increase in the demand for interactive traffic between nodes has led to a lack of network bandwidth resources,which requires an appropriate routing algorithm to effectively use resources to support high throughput and high quality communications.Much previous work has demonstrated that backpressure based routing algorithms can achieve superior performance in wireless sensor networks,such as throughput optimization,but delay performance is poor.And backpressure does not take into account issues about energy-efficient when resources are constrained.For this reason,this thesis proposes two improved backpressure routing algorithms for wireless sensor networks.The main work and innovation of this thesis are as follows:Firstly,this thesis proposes an improved backpressure routing algorithm based on delay and energy(DE-BP)to study how to improve the energy efficiency of backpressure algorithm.Backpressure algorithms that do not consider energy efficiency are difficult to deploy to resource-constrained wireless sensor networks.Different from the previous backpressure algorithm,this algorithm designs a new link weight calculation formula.In addition to the queue length,the last access node record of the packet and the neighbor node's energy state also serve as the primary factor for backpressure routing and scheduling decisions.In the DE-BP algorithm,data packets are more likely to go to neighboring nodes which are not been visited and with more residual energy.Moreover,the algorithm still maintains optimal throughput.The simulation results show that the DE-BP algorithm has improved the residual energy ratio,network throughput and delay performance compared with some existing work.Secondly,this thesis proposes an improved backpressure routing algorithm based on delay and distance(DD-BP),and studies how to effectively reduce the average end-to-end packet delay based on backpressure routing in sensor networks.The last packet problem,the packet starvation problem and the initial packet problem in the backpressure algorithm result in a large end-to-end packet delay.The algorithm has designed a new link weight calculation method.When thealgorithm makes backpressure routing and scheduling decisions,it considers three factors including the stay time of the head data packet in the node,the mutual position relationship between the nodes and the length of the queue.The DD-BP algorithm makes the node with long packet backlog time hold the priority of packet transmission,and the node closer to the destination node be more likely to be selected as the next hop.The algorithm can also maintain optimal throughput.Simulation results show that the DD-BP algorithm has improved throughput and delay performance as well as the data packet delivery ratio compared with previous work.