Decentralized task assignment in camera networks

The problem of task assignment naturally arises in multiagent multitask systems, where an optimal matching of agents and tasks is sought. This problem is a combinatorial optimization problem that can be solved in a centralized fashion, given the knowledge over all agent states and available tasks and the presence of a coordination center. Nonetheless, a decentralized solution would be beneficial in systems where communication among all agents may be limited and where the system efficiency and robustness cannot rely on the performance of a single central unit. A typical example of such a system is a videosurveillance network. In this work, we address the problem of decentralized task assignment in this context, first by formalizing the problem and presenting a general model framework, and then by proposing a decentralized solution inspired by the Stable Marriage Problem, that presents interesting performances in terms of optimality over a number of defined metrics, being supported by co...

• Type: Journal
• Year: 2010
• Where: CDC
• Authors: Angelo Cenedese, Federico Cerruti, Mirko Fabbro, Chiara Masiero, Luca Schenato
• Keywords: CDC 2010, Combinatorial Optimization Problem, Control Systems, Decentralized Solution, Task Assignment