Active Information Fusion

Efficient Sensor Selection Via Submodularity and Patitioning

 Given a sensor selection criterion, it is still necessary to have effienct sensor selection algorithms. We propose efficient sensor selection algorithms by exploiting both the characteristics of the utility function and the probabilistic dependency among the sensors for two typical cases: the budget-limit case and the optimal-tradeoff case. The budget-limit case involves selecting a sensor set that provides the maximum information gain within a budget limit. The optimal-tradeoff case involves selecting a sensor set that optimizes the tradeoff between information gain and cost. We first prove that mutual information is a submodular function in a relaxed condition, which helps the proposed algorithms to have performance guarantees. For the budget-limit case, we introduce a greedy approach that has a constant factor of (1-1/e) guarantee to the optimal performance. To further improve the computation, we introduce a partitioning procedure that exploits the probabilistic independence among the sensors. For the optimal-tradeoff case, a submodular-supermodular procedure is embedded with the proposed sensor selection algorithm to choose the sensor set that achieves the optimal tradeoff between the benefit and the cost.

Introduction 

Summary

Demos

Publications

BN Resources