Although recursive localization approaches are efficiently used in wireless sensor networks (WSNs), their application leads to increased energy consumption caused by the important communication overhead necessary to achieve the localization task. Indeed, localization information coverage increases iteratively as new nodes estimate their locations and become themselves new reference nodes. However, the uncontrollable number evolution of such nodes leads, especially in high density networks, to wasted energy, important communication overhead and even impacts the localization accuracy due the adverse effects of error propagation and accumulation. This chapter proposes an efficient recursive localization (ERL) approach that develops a new reliable reference selection strategy to ensure a better distribution of the reference nodes in the network. ERL improves localization accuracy without incurring any additional cost. It allows conserving the energy and consequently prolonging the WSN life time.
TopIntroduction
A Wireless Sensor Network (WSN) is a collection of tiny sensor nodes with non-rechargeable batteries (Sivakumar et al., 2018) and typically equipped with processing, sensing, power management and communication capabilities. These sensor nodes collaborate to form a Wireless Sensor Network (WSN). The essential objective of WSN is to observe, assemble and process the knowledge of sensor nodes within the network scope (Kaur et al., 2015). With the significant development and deployment of WSNs, associating the sensed data with its physical locations becomes a crucial requirement for different applications such as object tracking, environment monitoring, healthcare, intrusion detection, and habitat monitoring among others (Chow et al., 2018). Indeed, accurate positional information of the nodes also helps acquiring the value of location tagged parameters such as pressure, temperature, humidity and geographic coordinates from a given site (Pandey et al., 2017).
The simplest technique to localize a sensor node is to equip it with a Global Positioning System (GPS). However, its high cost and increased energy consumption makes it difficult to install in every node (Belghith et al., 2008), (Belghith et al., 2009) (Bouhdid et al., 2017). To overcome this weakness, other techniques, called collaborative localization techniques, were proposed (Niculescu et al., 2001), (Oliveira et al., 2009), (Ding et al., 2012), (Gui et al., 2015), (Li et al., 2015), (Ahmadi et al., 2016), (Darakech et al., 2018), (Bouhdid et al., 2018). They rely on the idea that sensor nodes with unknown locations (un-localized nodes) are guided by one or more sensor nodes with already known locations (either from GPS or by direct manual placement) for position estimation. The latter are called anchors or beacons. Based on the received information, the un-localized sensor nodes can compute their coordinates using distance measuring techniques (ranging techniques) (Paul et al., 2017). However, it is noticed that increasing the anchors number would also increase the deployment cost of the network (Mahjri et al., 2016). Moreover, these anchor nodes will also deplete their energy very quickly, which will drastically reduce the network lifetime and affect the network performance. Consequently, localization methods are designed such that the number of anchor nodes are reduced and explore instead and further the cooperation between nodes to enhance the localization (position) accuracy. One of the main cooperative approaches is the recursive localization approach, such as the Recursive Position Estimation (RPE) (Albowicz et al., 2001), where a node estimates its location based on the position information of three reference nodes. Once its position is estimated, it broadcasts its own location information to assist other nodes in estimating their positions. RPE makes full use of the connectivity of the network, and requires few anchors with the obvious advantages of simple localization method and easy realization.