Fractionally-Spaced Equalization: An Improved Digital Transversal Equalizer

Here we describe and demonstrate, via analysis and simulation, the performance improvement of voice-grade modems which use a Fractionally-Spaced Equalizer (FSE) instead of a conventional synchronous equalizer. The reason for this superior performance is that the fse adoptively realizes the optimum linear receiver; consequently it can effectively compensate for more severe delay distortion than the conventional adaptive equalizer, which suffers from aliasing effects. An additional advantage of the FSE is that data transmission can begin with an arbitrary sampling phase, since the equalizer synthesizes the correct delay during adaptation. We show that an FSE combined with a decision feedback section, which can mitigate the effect of severe amplitude distortion, can compensate for a wide range of linear distortion. At 9.6 kbit/s, the FSE provides a 2 to 3 dB gain in output signal-to-noise ratio, relative to the synchronous equalizer, over worst-case private-line channels. This translates to a theoretical improvement of approximately two orders of magnitude in bit error rate.