Optimal prediction of the mobile-radio fading envelope

Long-range prediction of the mobile-radio fading envelope is an enabling. technology for many fading compensation approaches. Because the fading envelope is well modeled as a bandlimited process it has special predictability properties. In this paper, we find a linear predictor which is optimal in the mean-square sense when the predictor impulse response is energy constrained. This solution may be used to determine the minimum mean squared error of a prediction based upon past values that are corrupted with estimation errors.

[1]  William W. Edmonson,et al.  The predictability of continuous-time, bandlimited processes , 2000, IEEE Trans. Signal Process..

[2]  William W. Edmonson,et al.  Linear prediction of bandlimited processes with flat spectral densities , 2001, IEEE Trans. Signal Process..

[3]  R. Clarke A statistical theory of mobile-radio reception , 1968 .

[4]  Hans D. Hallen,et al.  Long-range prediction of fading signals , 2000, IEEE Signal Process. Mag..

[5]  R.J. Lyman,et al.  Decision-directed tracking of fading channels using linear prediction of the fading envelope , 1999, Conference Record of the Thirty-Third Asilomar Conference on Signals, Systems, and Computers (Cat. No.CH37020).