Linear prediction of the rayleigh fading channel

This paper investigates the use of the linear MMSE prediction methods for the prognosis of non-stationary non-selective Rayleigh fading signals. It is shown that a properly chosen and configured prediction algorithm offers much better knowledge of the radio channel then the raw channel state estimate. The paper explores necessary conditions for successful prognosis and reveals methods for improving prediction effectiveness. Presented simulations show that the linear prediction is potentially a highly efficient tool for the narrowband signalling over the Rayleigh fading channel.