Learning-based Power Efficient Routing Algorithms Balancing Network Lifetime and Packet Delay for Wireless Sensor Networks  

作　　者：WU Wandeng WU Wandeng XU Deju XU Deju ZHU Yihua ZHU Yihua SHEN Zhenwei[1] SHEN Zhenwei[1] TANG Yiping[1] TANG Yiping[1] 

机构地区：[1]College of Information Engineering, Zhejiang University of Technology, Hangzhou 310032, China [1]College of Information Engineering, Zhejiang University of Technology, Hangzhou 310032, China [2]Science Academy of Mathematics, Capital Normal University, Beijing 100037, China [2]Science Academy of Mathematics, Capital Normal University, Beijing 100037, China

出　　处：《Chinese Journal of Electronics》2009年第4期703-707,共5页电子学报（英文版）

基　　金：This work is supported by the National Natural Science Foundation of China （No.60873228 and No.60673177）, Zhejiang Provincial Natural Science Foundation of China （No.Z105185）, and Zhejiang Provincial Qianjiang River Talent Plan （No.2007RlOG2020022）.

摘　　要：Power-saving is a critical issue for Wireless sensor networks （WSNs）. A Learning-based power efficient routing （LPER） algorithm is proposed. In the LPER, a fitness function, which balances network lifetime, energy consumption, and packet delay, is constructed and used in an ant colony system to establish the optimal route. In addition, reinforcement learning is applied in predicting the energy consumption of neighboring nodes. The LPEl~ is able to optimize network lifetime of WSNs, while keep- ing energy consumption and packet delay in a relative low level. Numeric experiments show the LPER outperforms the Minimal spanning tree （MST） and the Least energy tree （LET） based routing algorithms in terms of network lifetime and packet delay.

关 键 词：Wireless sensor networks Power-aware routing Power-saving Energy reservation. 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置] TP393[自动化与计算机技术—控制科学与工程]