Connected Mobile Sensors for Self-Deployment

Abstract

A Mobile Sensor Networks (MSNs) is normally made up of a number of mobile sensors which can be placed in a region of interest (ROI) by people. Sensors communicate with each other through wireless links to perform the distributed sensing ability for covering a region. Through a specific algorithm, the sensors move automatically, and finally the sensor network achieves a large sensing coverage. Sensing coverage can be established and promoted by different algorithms, and a good algorithm can lead sensors to form the largest possible sensing area without any sensing holes (areas that cannot be detected or monitored). The coverage of a sensor network is defined as the total area of interest covered, minus the area of the sensing holes.

We introduce a novel algorithm called the Spanning Tree-based Greedy-Rotation-Back (STGRB). The traditional Greedy-Rotation-Greedy (GRG) algorithm uses a point as a specific start so it's not appropriate for practical circumstances. So in the STGRB, we first use a spanning tree algorithm to get the gravity center (the sensor that connects the most sensors or is physically close to most of the other sensors) of the network. In this way we eliminate the first condition of selecting a sensor as a start point and also we conserve the energy of the sensors via letting them move a shorter distance.