Performance of nonlinear adaptive SbS-MAP detector using soft-statistics for digital transmissions over HF channels

The effectiveness of the recently proposed adaptive Maximum A Posteriori (MAP) receiver in Cusani and Mattila is verified for digital HF links between mobile and/or fixed units. Following Cusani and Mattila, a suitable MAP algorithm is employed to compute Symbol-by-Symbol (SBS) the A Posteriori Probabilities (APP's) of the channel state; from these, the detected data stream is obtained. The same APP's are employed by an adaptive nonlinear Kalman-like filter to estimate recursively the time-varying channel with fast convergence and good tracking properties, overcoming the delay problem suffered by conventional adaptive MLSE-VA detectors. Computer simulations allow to compare the adaptive SBS-MAP with other receivers (DFE, MLSE) and verify its effectiveness in the HF environment.