Semi-flocking algorithm for motion control of mobile sensors in large-scale surveillance systems

Area of Science:

• Computer Science
• Robotics
• Artificial Intelligence

Background:

• Sensor networks require self-organization for effective surveillance.
• Biologically inspired algorithms like Flocking and Anti-Flocking offer solutions for sensor control and coordination.
• Existing algorithms struggle to simultaneously achieve robust dynamic area coverage and target coverage due to conflicting objectives.

Purpose of the Study:

• To introduce Semi-Flocking, a novel biologically inspired algorithm for sensor networks.
• To address the limitations of Flocking and Anti-Flocking algorithms in maintaining simultaneous area and target coverage.
• To balance dynamic area coverage and target coverage through flock-sensor coordination.

Main Methods:

• Developed the Semi-Flocking algorithm, assigning sensors to targets while allowing others to explore.
• Implemented flock-sensor coordination mechanisms within the Semi-Flocking algorithm.
• Evaluated Semi-Flocking's performance against Flocking and Anti-Flocking using randomly moving targets and a pedestrian dataset.

Main Results:

• Semi-Flocking demonstrated superior performance in both area coverage and target coverage compared to Flocking and Anti-Flocking.
• The proposed algorithm achieved shorter target detection times.
• Semi-Flocking resulted in fewer undetected targets in both experimental scenarios.

Conclusions:

• Semi-Flocking effectively balances dynamic area and target coverage in surveillance sensor networks.
• The algorithm's biologically inspired approach with flock-sensor coordination offers significant advantages.
• Semi-Flocking presents a promising advancement for sensor network surveillance applications.