Q-Probabilistic Routing in Wireless Sensor Networks

• DOI: 10.1109/issnip.2007.4496810
• Publication Date: 2008-04-29T18:57:47Z
• Authors: Rocio Arroyo-Valles; Rocio Alaiz-Rodri...; Alicia Guerrero-C...; Jesus Cid-Sueiro
• Keywords: Geographic routing; Policy-based routing

Unpredictable topology changes, energy constraints and link unreliability make the information transmission a challenging problem in wireless sensor networks (WSN). Taking some ideas from machine learning methods, we propose novel geographic routing algorithm for WSN, named Q-probabilistic (Q-PR), that makes intelligent decisions delayed reward of previous actions local interaction among neighbor nodes, by using reinforcement Bayesian decision model. Moreover, considering message importance embedded itself can be adapted to traffic importance. Experimental results show Q-PR becomes policy that, as function importance, achieves trade-off expected number retransmissions (ETX), successful delivery rate network lifetime.