... data in the nth inter val using thegain estimates in the previous interval. We assume that wehave all the signal samples available up to the nth interval and the channel gain estimates are available ... distributedrandom variables. A. 1. ML gain estimatesSince ηcn and ηsnsamples are Gaussian distributed randomvariables, so if we assume that all the data and gain samplesare known, the received signal ... ’89),pp.640–646, San Francisco, Calif, USA, May 1989.[11] A. Aghamohammadi, H. Meyr, and G. Ascheid, A newmethod for phase synchronization and automatic gain control of linearly modulated signals of frequency-flat...