Signal sensing and modulation classification using pervasive sensor networks

This paper discusses the use of asynchronous low-cost sensors in distributed locations for sensing and classifying weak wireless signals. This weak signal may not be identified by using a single sensor alone, but can be detected and classified by fusing multiple weak signals collected by sensor networks. The asynchronous signal copies have unwanted offsets in time, frequency, and phase due to the diversities in local oscillators and unknown communication channels. This work proposes a post-synchronization method to estimate and compensate for offsets in the fusion process without adjusting the sensor parameters. The properly combined signal from the distributed sensors achieves a higher processing gain for reliable signal exploitation.