PROJECT TITLE :

Optimal quantization intervals in distributed detection

ABSTRACT:

We tend to optimize quantization intervals primarily based on Chernoff information (CI) and deflection criterion (DC) for distributed detection with multiple sensors and a fusion center (FC). The CI and DC of the general system are derived, and domestically optimal quantization of sensor observations is proved to attain international optimality for identical sensor observations. An optimal iterative partitioning algorithm is proposed for both nonidentically and identically distributed sensor reports. Results indicate that optimal a pair of-bit quantization offers detection performance that's close to what is achievable by collecting soft data from sensors.