Processing of signals either in time or in frequency domain is well established.Second order statistics(SOS)of Complex random signals are usually described by Covariance function(CF).In some cases, CF is not sufficient to completely describe SOS of complex random signals.For this purpose a new moment called Relation function(RF)is used.This thesis aims at processing of complex random signals using RF and Group delay functions(GDF).The concept of RF and its properties have been presented. Spectrum estimation of complex signals using RF and GDF(MGD &PGD)has been proposed.It is proved that pole or zero of a complex systems can be uniquely determined by RF approach.The concept of analytical signal spectrum estimation based on RF and GDF has been presented.It is shown that even at low SNR Spectrum of analytical signals can be clearly estimated with high resolution.The concept of RF has been extended to estimate the spectral components of 1st&2nd order dynamical systems.The concept of estimating wind profiles based on RF and GDF has been proposed.To check the superiority of RF results obtained based on the RF & GDF have been compared with that of CF.