Issue No.01 - January/February (2009 vol.24)
Kagan Tumer , Oregon State University
Adrian Agogino , University of California, Santa Cruz
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MIS.2009.10
Air traffic management is a fundamental challenge facing the US Federal Aviation Administration (FAA) and, owing to the difficulties in improving the infrastructure, a problem in need of an algorithmic solution. A proposed multiagent approach assigns an agent to a navigational fix (a specific location in 2D space) and derives specific agent rewards that shape the agent actions. These agents aim to maximize their own local rewards, but their doing so leads to coordinated behavior that reduces global congestion.
multiagent coordination, multiagent learning, air traffic management
Kagan Tumer, Adrian Agogino, "Improving Air Traffic Management with a Learning Multiagent System", IEEE Intelligent Systems, vol.24, no. 1, pp. 18-21, January/February 2009, doi:10.1109/MIS.2009.10