Research On Event-Driven Wireless Sensor Network Power Multi-Path Routing Protocol

The advancement of sensor technologies, micro electro-mechanical systems, network technologies and wireless communication technologies promote the rapid development of Wireless Sensor Network (WSN). WSN is a self-organized large scale dynamic network system that is composed of a great number of widely deployed sensor nodes using wireless communication, which can be used broadly in the field of military, environment monitoring, space exploring and security monitoring. In WSN, sensor nodes have the limitations of power, thus, maximizing network lifetime using finite nodes power becomes one of the most important goals of WSN design. Wireless sensor networks (WSNs) consist of low-cost, low-power tiny sensor nodes that can communicate with each other to perform sensing and data processing cooperatively. Network coverage and energy consumption are two primary problems in wireless sensor networks. The performance of a sensor network depends to a large extent on the sensor field coverage and its lifetime is determined by its energy consumption.Nowadays, sensor network use extremely widespread.The military, civil, the education, the scientific research, the medical service, cover each aspect which the humanity lives, the development future are limitless.For instance the militarily application, under the field operation condition, investigates enemy's military strength, the weaponry situation, may apply the sensor network, reduced the personnel casualty, has enlarged the detection reliability. In the medicine, surveys in the human body with the small sensor, discovery pathological change, prompt treatment.Studies the level along with the humanity the enhancement, the sensor network application scope can more and more extend.The wireless sensor network development also experienced from arrived simply complex, from preliminarily to high-level stage.US quite are at present many in this domain research achievement, the level quite is high.From the army to the place, from the theory to the practice, all has the breakthrough.Our country takes this domain the research to be late.The domestic some universities, the scientific research courtyard institute, have some explorations and the attempt, also has obtained some encouraging result, like designs the data for the core sensor network model, the development sensor network data management system.The domestic some companies also march into this domain, strives for produces, studies, grinds unifies, develops some wireless sensor network product.In this thesis, according to the characteristic of event-driven WSN, we propose anEvent-Driven Power Multi-Path routing protocol (EDPMP), which includes three parts: multi-path construction, data communication and route maintenance. In multi-path construction part, a Heuristic Local Gabriel Graph (HLGG) algorithm is proposed to do topology construction. HLGG employs Gabriel Graph simplification rule to adjust sensor nodes' transmission power, and uses heuristic mechanism to control node state. When an event happens, nodes that detect the event awaken themselves using random retreat mechanism and form a cluster. The earliest waken node is designated as the cluster header,which start the multi-path construction process and construct multi paths from itself to sink.In order to avoid the blindness of message forwarding, a "positive/negative position"concept is given, in which node makes forwarding decision according to its relationship with neighbors. During data communication phase, cluster header collects data of other nodes in the cluster and carries through data aggregation, and then it decides to use single-hop or multi-hop routing policy according to routing table. If there is a single-hop route exists, the cluster header sends data to sink directly. Otherwise the power metric function is used iteratively to select a local optimized next hop, at last a path is selected among the multi paths to do multi-hop data forwarding. Route maintenance is mainly responsible for power variation maintenance, power metric function maintenance and routing table maintenance.Simulation under ns2 shows that EDPMP has a good performance in saving power consumption: mean-square error of nodes' power consumption is less than 0.1, which means that power consumption of nodes is balanced. And network lifetime is extended by about 30% than that of LEACH.