Abstract: In wireless sensor networks, sensors are severely constrained by the amount of battery power. In most existing protocols, nodes are assumed to have the global information of networks, which will cost too much energy expenditure. In this article, several localized routing algorithms are proposed for the many-to-one routing. When routing, nodes make routing decisions solely on the information of their neighbors and destination. Besides, the concept of constrained backward transmission is proposed by the first time. We combine the idea of backward transmission with existing localized algorithms. Simulation experiments show that the backward minimizing reluctance algorithm performs effectively in terms of extending the lifetime of networks, which prolongs at least 12% and 16% rounds than others. At last, algorithms are adapted to the networks where nodes can not fuse data, the average-reluctance routing algorithm works efficiently.

Key words: wireless sensor networks, localized routing, algorithm