Sensor Integration by Joint PDF Construction using the Exponential Family

We investigate the problem of sensor integration to combine all the available information in a multi-sensor setting from a statistical standpoint. Specifically, we propose a novel method of constructing the joint probability density function (pdf) of the measurements from all the sensors based on the exponential family and small signal assumption. The constructed pdf only requires knowledge of the joint pdf under a reference hypothesis and, hence, is useful in many practical cases. Examples and simulation results show that our method requires less information compared with existing methods but attains comparable detection/classification performance.