July 30, 2009 16:55 SPI-B837 Trimsize-9.75in x 6.5in b837-ch07 1st Reading Chapter 7 Cooperative Geographical Routing in Wireless Sensor Networks Min Chen ∗,‡ , Victor C.M. Leung ∗,§ and Shiwen Mao †,¶ ∗ Department of Electrical and Computer Engineering University of British Columbia, V6T 1Z4, Canada † Deptartment of Electrical and Computer Engineering Auburn University, Auburn, AL 36849, USA ‡ minchen@ece.ubc.ca § vleung@ece.ubc.ca ¶ smao@ieee.org Reliable delivery of sensory data to a sink node in large scale sensor networks is a challenging problem. This chapter tackles this problem by assuming dense deployment of sensors, which allows us to exploit diversity in choosing intermedi- ate nodes for reliability and energy-eﬃciency. The proposed reliable and energy- eﬃcient routing (REER) protocol is based on the geographic routing approach. The central idea of REER is the notion of reference nodes (RNs), which means the nodes closest to the ideal locations between the source and to the sink. The multiple Cooperative Nodes (CNs) around RNs will contend to relay data pack- ets; thus, there is no overhead of route discovery and REER is resilient to node failures and transmission errors. By adjusting the distances between RNs, we can control the trade-oﬀ between reliability and energy-eﬃciency, which is validated by both analysis and simulation. 7.1. Introduction Sensor networks are usually subject to high failure rate: connectivity between nodes can be lost due to environmental noise and obstacles; nodes may die due to battery depletion, environmental changes or malicious destruction. In such environments, reliable and energy-eﬃcient data delivery is crucial because sensor nodes operate with limited battery power and error-prone wireless channels. However, the goal of reliability and energy-eﬃciency often conﬂict each other. We consider two extremes of routing protocols in terms of these two design objectives: unicast routing and ﬂooding. Unicast routing is energy-eﬃcient for reliable networks, but is not robust for dynamic networks. Flooding is very robust for dynamic and error-prone net- works, but incurs a high overhead for sensor networks. Some routing protocols try 1