MODELLING AND SIMULATION OF A SIMPLIFIED DEMODULATION AND DETECTION METHOD FOR ZERO-IF FSK SIGNALS IN WIRELESS COMMUNICATIONS.

M-ary baseband FSK modulated signals, which utilize M-set different frequencies, can be modelled as a set of complex rotating phasors. The received modulated signals can be detected by counting the speed and the direction of the complex phasor rotation. Conventional methods such as zero-crossing detection used this baseband FSK signal property. The detection approach presented in this paper is to count the number of quadrant crossing instead of zero crossing in a symbol period to demodulate FSK signals. Counting quadrant crossings acts as a noise filtering and improves the detection performance. The performance is improved further by using decision rules at fuzzy boundary and soft decision detection. The system modelling and simulation methods are illustrated in the paper. The performance comparison and improvement are shown by a 4-ary FSK signal example on wireless Rayleigh fading channels. [ABSTRACT FROM AUTHOR]