| Summary: | 碩士 === 大同大學 === 通訊工程研究所 === 96 === ZigBee is a wireless network standard based on IEEE 802.15.4. It is a wireless sensor network technology that the main characteristics is low transmission range (usually less than 100 meters), low data rate (250 kbps), low price and low power consumption. ZigBee is expected to use in light meter sensor, temperature and humidity sensor, remote monitoring, home control, and industrial automation.
To reduce the costs and energy consumption always is one of the most important purposes about the research correlated with ZigBee. The main subject of this paper is to reduce energy consumption and latency by reformed the routing algorithm. For the tree routing of ZigBee, when a nodes went to transmitting a packet that will follow the tree topology to the destination even if the destination is located nearby. In order to solve this problem, Taehong Kim, Daeyoung Kim, Noseong Park*, Seong-eun Yoo and Tomás Sánchez López propose the shortcut tree routing algorithm to reduce the routing cost of ZigBee tree routing by using neighbor table. This algorithm can find the shortest path by compute the routing cost for all of router that stored in neighbor table, and transmit the packet to the neighbor router that can reduce the hop count of transmission. It is a useful routing algorithm, but not perfect. Since the characteristics of Zigbee are low price, low complexity, and limited resources, it has limited capability of compute. So it may be effect delay by compute the routing cost that if neighbor table store overmuch router’s information.
In order to get better efficiency, this paper analyzes the difference of efficiency that use shortcut tree routing between different conditions, and proposes an optimized neighbor table algorithm. The simulation results show that our improved shortcut tree routing algorithm can achieve more stable and better efficiency then the previous shortcut tree routing algorithm. For our various test environments, the proposed algorithm can improve the hop counts by 4 to 10 percent, and as a result can achieve benefit of reducing energy consumption and latency.
|
|---|
