Theoretical BER evaluation for slow FH/NBFM systems in non-Gaussian-noise environment

Abstract

Conventionally frequency hopping spread spectrum systems are used in army tactical communication, where the noise environment is assumed to be Gaussian. Nowadays the system has its applications in commercial world also. The electromagnetic environment encountered by receiver systems in deployment conditions is often non-Gaussian in nature, owing to collocated EM systems comprising of own and enemy EM assets competing for the spectral resources resulting in co-channel, adjacent channel, wide as well as narrow band EMI situations of complex dynamics. The signal detectors designed for Gaussian statistics may suffer significant degradation when the actual statistics deviate from the Gaussian model. Both man-made and low frequency atmospheric interference environments are basically impulsive-noise, i.e. they have a highly structured form, characterized by significant probabilities of large interference levels. Under such environments the performance of the receiver systems, designed to perform optimally in Gaussian-noise, degrades significantly. The paper evaluates the performance of slow FH/NBFM communication systems in the presence of non-Gaussian noise. Theoretical calculation of BER is given for both, L-D detector and differential detector. The approach taken here is based on modeling the non-Gaussian noise as a two-component mixture of Gaussian distributions.