Distributed Antiflocking Algorithms for Dynamic Coverage of Mobile Sensor Networks

Abstract

Abstract-Mobile sensor networks (MSNs) are often used for monitoring large areas of interest (AoI) in remote and hostile environments which can be highly dynamic in nature. Due to the infrastructure cost, MSNs usually consist of limited number of sensor nodes. In order to cover large AoI, the mobile nodes have to move in an environment while monitoring the area dynamically. MSNs that are controlled by most of the previously proposed dynamic coverage algorithms either lack adaptability to dynamic environments or display poor coverage performances due to considerable overlapping of sensing coverage. As a new class of emergent motion control algorithms for MSNs, anti-flocking control algorithms enable MSNs to self-organize in an environment and provide impressive dynamic coverage performances. The anti-flocking algorithms are inspired by the solitary behavior of some animals who try to separate from their species in most of daily activities in order to maximize their own gains. In this paper, we propose two distributed anti-flocking algorithms for dynamic coverage of MSNs, one for obstacle free environments and the other one for obstacle dense environments. Both are based on the sensing history and local interactions among sensor nodes.

Path-Planning-Enabled Semiflocking Control for Multitarget Monitoring in Mobile Sensor Networks

“…Semi-flocking algorithms had been proposed in [15] and [16] which combine flocking [17], [18] and anti-flocking [19], [20] controls via mode switching. Initially, the nodes search the AoI to detect targets.…”

“…where c 1 is a positive constant, and q s is the next searching location of node i. Here, we apply the approach proposed in [20] to determine q s for node i. In that approach, each node records its last visiting time of cells in the AoI on its own information map.…”

“…In the following tests, the number of nodes is 24 and the number of targets in the terrain varies from 1 to 6. The following parameters remained fixed in all tests: a = b = 5 for φ(x), h = 0.2, c 1 = 0.85, c 2 = 2, c 3 = 2, θ k = 15 m, = 0.1, r s = 10 m, and r c = 18 m [15], [16], [20]. The maximum and minimum numbers of monitoring nodes required for each target are both 3.…”

“…Switch into the traveling state; 16: Carry out A* IL-based path planning; 17: Visit the pre-stored intersection points on the path one-after-another; 18: else 19: Switch into the monitoring state; 20:…”

Temnothorax albipennis migration inspired semi-flocking control for mobile sensor networks

“…Flocking-based algorithms have proven their capabilities in area coverage and target tracking applications 6,7 . In flockingcontrolled systems, behaviors of each individual agent are simple but interrelated.…”