A wireless sensor network is a collection of sensor nodes that have the ability to sense phenomena in a given environment and collect data, perform computation on the gathered data, and transmit (or forward) it to their destination. Unfortunately, these sensor nodes have limited power, computational, and storage capabilities. These factors have an influence on the design of wireless sensor networks and make it more challenging. In order to overcome these limitations, various power management techniques and energy-efficient protocols have been designed. Among such techniques and protocols, geographic routing is one of the most efficient ways to solve some of the design issues. Geographic routing in wireless sensor networks uses location information of the sensor nodes to define a path from source to destination without having to build a network topology. In this paper, we present a survey of the existing geographic routing techniques both in two-dimensional (2D) and three-dimensional (3D) spaces. Furthermore, we will study the advantages of each routing technique and provide a discussion based on their practical possibility of deployment.
TopIntroduction
Recent advancements in wireless sensor networks (WSNs) have shown them as one of the most promising techniques for the future. Sensor nodes have different capabilities, such as phenomenon sensing and data generation, data processing, and data transmission or forwarding. All of these capabilities have made them attractive for practical deployment in addressing various applications, including military, health, environment, home, and other commercial areas. A wireless sensor network (WSN) consists of (hundreds or thousands of) sensor nodes deployed in either a deterministic approach or randomly in a dense fashion. In fact, deploying a large number of sensor nodes is not a burden to the infrastructure as they are low-cost devices and allow sensing a large geographical region with greater accuracy. From now on, we use the terms “sensor”, “node”, and “sensor node” interchangeably.
Sensor nodes are typically dispersed all around a deployment field which they are meant to cover using their sensing capabilities, such as seismic, thermal, acoustic, visual, and radar. Each sensor node consists of a sensing unit, processor, transmission unit along with some additional capabilities, such as a positioning estimating system (like global positioning system) and a mobilizer. Figure 1 (Al-Karaki & Kamal, 2004), (Akyildiz, Su, Sankarasubramaniam, & Cayirci, 2002) shows that each sensor node can collect data either by its own capability or by accepting some other sensor node’s data. Also, it may aggregate the collected data and forward it to the base station. The base station acts as the destination node (or sink) to the individual sensor nodes. The end user can communicate with the base station through the internet. In order for each individual sensor to communicate with the base station, these sensor nodes must form a connected structure and every sensor node must have at least one path that leads to the base station. Since the sensor nodes have limited power resources and could be deployed in environments that are unattended, they must be able to re-organize themselves even when some of them become inactive due to failure, die out of battery, are unable to communicate with their neighbors due to some unforeseen obstruction, or are in a sleep mode. Therefore, wireless sensor networks must ensure connectivity among the sensor nodes at all times, and the network operational lifetime must be maximized.
Figure 1. Components of a sensor node (Al-Karaki & Kamal, 2004)