SCAN: A Sensor Deployment Algorithm in Hazardous Regions via Mobile Actuators

In this paper, we propose two novel robot-assisted energy-efficient wireless mobile sensor deployment algorithms to monitor hazardous, uneven, and obstacle-prone areas where human intervention is prohibited. The algorithms are designed to exploit the operation in two simple steps. Initially, the robot deploys the mobile sensors in suitable positions by exploring the intended region using two different methods: 1) basic SCAN¹ and 2) opportunistic SCAN. In basic SCAN, the robot explores the intended area using simple scanning to install sensors. On the other hand, in opportunistic SCAN, the robot explores the intended area and installs the sensors whenever possible and stays in that region to cover as much as it can. Then, the installed sensors run coverage algorithm to provide information about the empty neighborhood to the running robot. Initially, the robot starts from some reference point and deploys sensors avoiding any obstacles or boundaries without the help of GPS. Combining these algorithms with the coverage algorithm guarantees maximum coverage for different obstacle distributions and for different regions of interest. Extensive simulation experiments are conducted to evaluate these algorithms. Opportunistic SCAN algorithm provides maximum coverage with minimum travel distance for the robot compared with the basic SCAN algorithm and the backtracking algorithm. Our algorithm also eliminates the deadlock issue for the robots that is raised when back pointer is removed.¹

Scalable Coordinate Accumulation Network.

I. Introduction

A wireless sensor network (WSN) consists of a large number of sensor devices with microchips and powered by limited energy source (small batteries) and these devices can communicate among them and with other authorized devices via wireless links. WSNs are used widely in numerous application areas including event monitoring, detection, control of environment, temperature sensing, humidity sensing, oil and gas exploration, toxic gas emission detection, traffic control, manufacturing and plant automation, and military surveillance etc [1]. Sensors can be both static and autonomous depending on the application domain [2]. Hybrid sensor networks [3]–[5] can be more energy efficient where static sensors are only responsible for sensing data from the environment and autonomous devices are responsible for the relocation [6]–[8] of those sensors, data aggregation [9], maintenance, etc. Here autonomous devices refer to the actuators or the robotic agents who are resource rich in terms of battery power, movement features etc. On the other hand, static wireless sensors are very tiny in shape and resource constraint. Thus, these small sensors deplete energy quickly if they have to perform movement, maintenance and repair by themselves, which is not feasible for real life application scenario. Different versions and configurations of hybrid sensor networks have been formulated to carry out the desired outcome with keeping mind those issues previously mentioned.