| Summary: | 博士 === 大同大學 === 資訊工程學系(所) === 101 === Sensor networks help us not only to detect some useful events but also to collect important data for study. For the safety reason, researchers deploy the sensor nodes in the critical or dangerous locations to detect important events. They just receive the collected data sent by sensor networks in the laboratory. The sensor networks consists of sensor nodes, the node sink, communication (Internet), and client computers. The mission of the sensor sink is to collect the data sent from sensor nodes through Internet. Once the data collected, it can then be analyzed for the intended purpose.
In the wireless sensor network, routing, synchronization, and power saving are the most important issues. There are many schemes to implement the synchronization and routing functions in the wireless sensor network, such as time-sync protocol for sensor network (TPSN), pulse-coupled scheme, cluster header method, and so on. TPSN uses the tree structure to implement synchronization. However, as the levels increased, the synchronization time will also increase in the TPSN. Pulse-coupled scheme uses the full mesh structure. All nodes in the sensor network are the header nodes, and they broadcast its information to other nodes for synchronization.
In this dissertation, we propose an Ant Colony Optimization (ACO) scheme for implementing routing, synchronization, and power saving functions in the wireless sensor network. The synchronization level always keeps 3 levels to sync each other. For power saving, leapfrog approach has been applied for each header in the same routing path. The proposed scheme is compared with the TPSN and the Pulse-coupled methods in terms of the synchronization and power consumption. The experimental results show that comparing with the TPSN method, synchronization time can be reduced by 93.6% and power consumption can be reduced by 7.3% with applying the proposed scheme. While comparing with the Pulse-coupled method, by applying the proposed scheme, synchronization time can be reduced by 77% and power consumption can be reduced by 5.5%.
|
|---|
