A Model of Autonomous Motion in Ad Hoc Networks to Maximise Area Coverage

Abstract

Ad hoc networks are self-configuring networks of mobile nodes, connected by wireless links. Suppose each mobile node can make observations within a circular area of radius r obs centred on its own location. The area coverage of the network is defined as the total area observed by the mobile nodes. We investigate a distributed scalable method based on local interactions with minimal sensing and low computational cost whereby the nodes move autonomously (self-deployment) in order to maximise the coverage of the network, while at the same time ensuring that the mobile nodes do not move so far away from each other (thus trivially maximising the coverage) that they become disconnected. Certain nodes may be instructed to move to specified locations. These guide nodes induce a correlated movement of groups of nodes which follow the guide nodes and establish maximal coverage in the specified locations. Simulation results demonstrate the coverage achieved by a group of 100 nodes when moving on an unbounded plane (optional guide nodes induce a collective motion to areas of interest) and when moving on a bounded plane with barriers or hills.

DACYCLEM: A decentralized algorithm for maximizing coverage and lifetime in a mobile wireless sensor network

“…The forces could also be used as a method to generate a social behavior among a group of robots [40]. Some methods even propose to avoid obstacles by adding a repulsive forces to them [47,4]. This model leads the robots to use bypass strategies as they deploy themselves.…”

“…When they are equipped with sensors they form a system that can be considered as a mobile wireless sensor network (M-WSN). Current researches in the field are facing different issues: mobility, energy saving, topology control, connectivity maintenance, obstacle detection and avoidance or fault tolerance, to name a few [1,2,3,4,5]. The problem addressed in this paper is made up several sub-problems among these.…”

A Mobility Model Based on Improved Artificial Potential Fields for Swarms of UAVs

“…This task is of increasing complexity if the environment in which the drones operate are not known prior to the mission (and contain obstacles) and thus need to be calculated on board, in real-time. Some methods have been extensively studied to solve this problem: APF methods introduced by Khatib [16], that we detail in the following section; virtual forces [11], [13]; Genetic Algorithms [14]; chaotic processes [23]; fuzzy logic [9]; Particule Swarm Optimization (PSO) or Reinforcement Learning [8].…”

“…Finally, we can quote Virtual Forces methods, that make it possible to maintain formations in constrained environments or to fulfill coverage missions [11]. Virtual Forces are also adapted for systems that rely on leaders and followers [13]. Still, methods based on virtual forces have problems similar to those of APF-based methods.…”

A model of self deployment to maximise area coverage in sensor networks

“…Controlled mobility can be modelled by means of artificial physics (see [12,13] and the references therein) where node movements are determined by attractive and/or repulsive forces amongst the nodes and between the nodes and the objects to be approached or avoided. The aggregated forces that a node is subject to are determined from local information such as the positions of adjacent nodes.…”