Track maintenance through cooperating agents

Maintenance of air tracks requires regular track updates to counter effects of aging. In multi-sensor networks, sensors need to cooperate to optimise the time allocation for track maintenance. Early updates lead to waste of time without improving the track quality, while late updates may result in track drop as the target has moved out of sight of the updating sensor. Earlier research has presented an agent-based scheduler to solve the NP-complete sensor scheduling problem. An element of this agent-based scheduler is the generation of track updates. This paper presents a method to translate the air picture into an agent-environment which monitors the quality of the track data and produces a track update request when necessary. This approach is economic in time as update requests are only generated when the quality of track data requires updating. Furthermore, the output of the track update generation interfaces directly with any agent-based scheduler.