Adaptive equalization of a digital communications channel with a reduced-order equalizer

Extended Kalman filters have been suggested for adaptive equalization of digital communication channels. They are designed from a state space structure that describes the dynamics of both the transmission of the digital signals as well as the properties of the channel. When the channel is modeled as a finite impulse response filter with many taps, the extended Kalman filter would be unacceptable for high speed operation due to the required number of calculations. In this paper we provide a reduced-order alternative, which can require fewer calculations because of its ability to estimate states of interest, while ignoring other states and at the same time giving performance that can be almost as good as the full-order extended Kalman filter.<<ETX>>

[1]  C. Sims,et al.  Linear discrete reduced order filtering , 1978, 1978 IEEE Conference on Decision and Control including the 17th Symposium on Adaptive Processes.

[2]  Krishan Nagpal,et al.  Linearized Reduced Order Filtering , 1987, 1987 American Control Conference.

[3]  Steve McLaughlin,et al.  Adaptive equalisation via Kalman filtering techniques , 1991 .

[4]  H. Kaufman,et al.  The Kalman Filter for the Equalization of a Digital Communications Channel , 1971 .

[5]  K. Nagpal,et al.  Reduced-order estimation Part 2. Smoothing , 1987 .

[6]  S. Qureshi,et al.  Adaptive equalization , 1982, Proceedings of the IEEE.

[7]  Craig S. Sims,et al.  Reduced-order estimation Part 1. Filtering , 1987 .

[8]  F. Lewis.  Optimal Estimation: With an Introduction to Stochastic Control Theory , 1986 .