This thesis presents a new equalizer structure, the asymmetric decision feedback equalizer (ADFE). The mean square error (MSE) and bit error rate (BER) performance of the ADFE is analyzed and simulated. The ADFE is shown to have an identical MSE characterization, identical MSE performance, and similar BER performance, to the DFE. The BER of the ADFE is estimated by simulation, and by use of a finite discrete Markov process.
The ADFE is paired with a DFE to form the ADFE system, which equalizes communications between two transceivers with an asymmetric distribution of equalization complexity. The performance of this system is examined under various nonidealities including a mismatch in characterization signal-to-noise ratios (SNRs), timing error, adaptive training, finite precision arithmetic, and measured indoor wireless channels. Results stemming from the SNR mismatch indicate the potential for improved BER performance of the DFE and ADFE through a modification of MSE characterization.PDF file, 416 kBytes