Research On Trail And Data-driven Distributed Routing Protocols For Wireless Sensor Networks With Mobile Sinks

Wireless Sensor Networks(WSN)are constituent of a large number of sensor nodes,which are self-organized and can deliver perception data to sink node(s)via wireless multi-hop paths formed by sensor nodes,and then to users or central entities via the Internet.Existing research showed that wireless sensor networks with static sink nodes has the hotspot issue such as sensor nodes closer to the sink nodes can consume energy much quickly than other nodes in the network,and also cause traffic concentration issue,which largely affect the performance of WSNs.Mobile sink(s)can solve the above problems effectively and improve the network performance by appropriately distributing the traffic in the network.However,mobility of sink(s)will cause in change of the network topology and also data transmission path frequently,increase network overhead,and affect network performance.This thesis conducts extensive research in design of routing protocols for wireless sensor networks with mobile sink(s),and the design goals are to improve the packet delivery ratio,reduce the network overhead,and also prolonging the network lifetime.In our work in this thesis,we assume that no location is known at all nodes(including sensor nodes and sink nodes)and also no any global network state information is known.This thesis accordingly designs two reactive routing protocols,and carries out extensive some simulations for performance evaluation.The following major work has been carried out in this thesis:Firstly,design a new reactive routing protocol,TBD,which combines trail based forwarding and data-driven route learning/updating for improved performance as sink(s)move in the network.In TBD,when a sensor node has a data packet to forward to a sink node,trail-based forwarding has the highest priority,data-driven routing is the second,and random walk has the lowest priority.Specifically,if the packet is on a sensor node which has trail information,it will be transferred to the sensor node with freshest time stamp along the trail;if the packet is on a sensor node which has data-driven learnt routing information,it will be transferred as such routing entry indicates;otherwise,random walk will be triggered.In addition,TBD uses a two-hop local broadcast mechanism to find alternate route when encountering a broken point on a trail.Second,design an energy-efficient version of TBD,referred to as E-TBD,in order to prolong the network lifetime.In E-TBD,in the next hop selection,residual energy and time stamp freshness of next hop candidates are considered.Following this line,we design energy-efficient trail-based forwarding procedure and energy-efficient random walk routing procedure.In this way,nodes with low residual energy can be largely protected and network lifetime can be prolonged.Finally,carry out extensive simulation experiments,and the results show that TBD can achieve high packet delivery ratio performance with low protocol overhead;compared with TBD,E-TBD can keep high packet delivery ratio performance with low protocol overhead,while prolonging the network lifetime remarkably.