Low-complexity noncoherent fusion rules for wireless sensor networks monitoring multiple events

Abstract

Parallel triple-layer wireless sensor network (WSN) assisted by frequency-hopping (FH) and M-ary frequency-shift keying (MFSK) modulation, abbreviated FH/MFSK WSN, is proposed to monitor multiple source events (SEs) each having multiple states. The SEs are simultaneously observed by a few local sensor nodes (LSNs), which convey their decisions to a fusion center (FC) where the SEs' states are classified (detected) based on noncoherent fusion rules. In this paper, four low-complexity noncoherent fusion rules are studied. They are the conventional benchmark of equal-gain combining (EGC) and three proposed noncoherent fusion rules, namely erasure-supported EGC (ES-EGC), EGC-assisted N-order iterative interference cancellation (IIC), and ES-EGC-assisted N-order IIC. The complexity of these fusion rules is analyzed and the performance of the FH/MFSK WSN employing these fusion rules is investigated and compared when wireless channels from LSNs to the FC experience Rayleigh fading.