Sensor Localization in Wireless Sensor Networks Using Cultural Algorithm

1. Introduction

A wireless sensor network (WSN) consists of a group of small sensor nodes that can sense, gather information from the environment and process it collectively. Typical applications of WSNs include wildlife monitoring, traffic surveillance, disaster event detection, crop and pollution monitoring, and target detection and tracking (Mohamed et al., 2018). In most applications of WSNs, sensor nodes are deployed in a random manner using aircraft or robots. The information about the locations of sensor nodes plays a pivotal role in WSN applications. It is critical in rescue operations in natural calamities, such as forest fires, landslides and floods.

Localization is a process that creates location awareness in sensor nodes. Location information is crucial in locating the geographical location of the event, or in tracking applications, such as location of the movement of animals or enemies in a battlefield or to deploy a troop to help the people trapped in a disastrous situation or detecting the location of enemy in the battlefield. Many network-based applications require location information. Some of the examples are routing algorithms, identifying network boundaries, information querying for on sensor deployment and network topology. Due to increasing application potential of WSNs, developers have been deploying large-scale sensor networks in which localization gets quite challenging. The most commonly used method to create location awareness in ad hoc mobile networks is the use of global positioning system (GPS). This method is however not feasible in WSNs due to cost and energy constraints. The GPS cannot be accessed by the sensor nodes if they are deployed underground, underwater or in thick forests (Paul & Sato, 2017). A common approach to localization is to deploy a few sensor nodes that know the coordinates of their locations. These nodes are referred as anchors or beacons and the remaining nodes in the network are called unknown, target or dumb nodes. The dumb nodes localize themselves using the information about the coordinates of the anchor nodes. This is known as anchor-assisted localization.

Localization methods are broadly classified as range-free or range-based. Range-free localization methods use the connectivity information, such as hop count and peer nodes’ locations for location estimation (Singh, Sharma et al., 2015). In range-based localization methods, the distances between anchors and the dumb node are estimated using metrics, such as angle of arrival of signals, time difference in the arrival of signals, time of arrival, difference in received signal strength (RSS) and communication range (Mekelleche & Haffaf, 2017). The estimated distances and the anchor locations are used to estimate the locations of a dumb sensor nodes. Trilateration, multilateration, triangulation and bounding box are the geometric methods are used in range-based localization.

The estimation of distance in a range-based localization may not be accurate due to the environmental noise. This affects the accuracy in location estimation and results in localization error (Kulkarni et al., 2016). The major challenges in range-based sensor localization are as given below (Alrajeh et al., 2013):

• 1.

  To design a localization algorithm that localizes maximum number of nodes with high accuracy and in less computing time;

• 2.

  To minimize the effect of noise in location estimation;

• 3.

  To perform localization using minimum number of anchors;

• 4.

  To minimize energy consumption in localization process;

• 5.

  To perform frequent localization for mobile nodes and developing an optimal path for mobile beacon.