Design of an adaptive Kalman equalizer and its performance over fading multipath channels

Abstract

The development of an adaptive equalizer for mobile radio using the Kalman algorithm is addressed, including an assessment of its performance by means of computer simulation and hardware experimentation. The adaptive equalizer consists of a fractionally spaced decision feedback equalizing element and a tap-coefficient-update element. Tap coefficients are updated based on the Kalman filter algorithm. Details for the parameters are given, such as the number of taps, the quantization number of bits, and the forgetting factor. Hardware of the equalizer consists of a digital signal processor and a few finite impulse response digital filters implemented on some LSI chips where the tap coefficients are adaptively changeable. A bit error ratio (BER) of less than 10/sup -2/ is achieved, and remarkable improvement is observed compared with the unequalized performance. It is shown that the adaptive equalizer is useful for fading multipath channels.<>