Asymptotic Equivalence Between the UML and the Extended SLN Symbol Timing Estimation on an Additive Colored Gaussian Noise Channel

It has recently been reported that the unconditional maximum likelihood (UML) symbol-timing estimation is asymptotically equivalent to the square-law nonlinearity (SLN) on an additive white Gaussian noise (AWGN) channel. By introducing a whitened matched filter (WMF), we propose an extended SLN and prove the asymptotical equivalence between the UML and the extended SLN on an additive colored Gaussian noise. Simulation results show the same behavior asymptotically when a mixture of an AWGN and a flicker noise is adopted as a noise source.