Distributed nonuniform coverage with limited scalar measurements

Abstract

Abstract-In this paper we demonstrate a distributed coverage control method for a network of mobile agents moving in one dimension along a scalar information density field. The method requires each agent to take a finite number of measurements of the density field in the interval between its two neighbors and calculate its next position in order to drive the network nearer to the configuration for optimal coverage. We derive several results relating to the equilibrium properties of the sensor network and the convergence properties near fixed points. We illustrate with simulations of the algorithm.

Nonuniform Line Coverage From Noisy Scalar Measurements

“…Secondly, the general coverage control problem has been the subject of much recent interest within the control community; we refer the reader to the recent papers [6], [5], [17], [13], [14], [9], [2], [19], [10], [11], [3], [4], [12], [15], [16], [1], [20], [7] and the references therein. The nonuniform line coverage problem is the one-dimensional version of the general coverage control problem, and consequently, it provides a simplified setting to make advances in addressing outstanding questions in coverage control.…”

“…The necessity of knowing the cartogram transformation is relaxed in the case of nonuniform coverage control on the line in [12] where fully distributed nonuniform coverage protocols are derived. A follow-up work [7] considers the line coverage problem where only samples of ρ are available and explores the performance guarantees associated with some simple strategies. Some recent work [4], [8] considers strategies for coverage on the line when some proportion of the sensors are expected to randomly fail.…”