Semi-Cooperative Spectrum Fusion (SCSF) for aerial reading of a correlated sensor field

We present new theoretical results on the semi-cooperative spectrum fusion (SCSF) method of data fusion for spatially correlated information in wireless sensors. Data fusion exploits correlated sensed information to save scarce sensor energy by reducing the communication energy costs. SCSF eliminates medium access control (MAC) and all data communication between sensors. In SCSF, a beacon is sent by the airborne collector through a directive antenna, and the illuminated sensors respond by simultaneously transmitting frequency modulated signals. The superposition of the sensor transmissions resembles a Doppler spectrum. The airborne collector (fusion center) applies a blind frequency estimator to the received superimposed signals for fusing the data. Estimation performance is evaluated in terms of the inverse normalized mean square error (INMSE) and is analyzed in terms of various system parameters such as sensor density, degree of correlation and modulation index. The analysis shows that the SCSF scheme performs well relative to a conventional fusion scheme that uses sample mean estimator for data fusion while not using MAC.