Feasibility of Adaptive Transmission for Multicarrier and Frequency Hopping Mobile Radio Systems

We address feasibility of adaptive transmission for rapidly varying fading channels encountered in mobile communication systems. Due to rapid channel variations and the feedback delay, implementation of adaptive transmission in these channels requires accurate prediction of future channel conditions. We describe the design and performance of adaptive transmission methods enabled by the Long Range Fading Prediction algorithms (LRP) for Orthogonal Frequency Division Multiplexing (OFDM) and slow Frequency-Hopping Spread Spectrum (FH/SS) systems. The LRP algorithms are tested with a modified Jakes model and our realistic physical model. To enable channel loading for adaptive OFDM (AOFDM), robust adaptive prediction methods that track channel parameter variations are developed and compared. Feasibility of AOFDM for rapidly varying mobile radio channels is demonstrated, with bit rates approaching those for channels with ideal knowledge of the channel state information (CSI). For FH/SS systems that employ coherent detection, we utilize correlated fading to predict future CSI for the upcoming frequency based on channel observations of other frequencies. This LRP method is utilized in adaptive modulation to mitigate the effects of fading. Analysis and simulation results show that while prediction is not as reliable for FH/SS as for AOFDM, significant performance gains are achieved relative to non-adaptive methods.