Abstract
We present two different fractionally spaced (FS) equalisers based on subband methods, with the aim of reducing the computational complexity and increasing the convergence rate of a standard fullband FS equaliser. This is achieved by operating in decimated subbands; at a considerably lower update rate and by exploiting the prewhitening effect that a filter bank has on the considerable spectral dynamics of a signal received through a severely distorting channel. The two presented subband structures differ in their level of realising the feedforward and feedback part of the equaliser in the subband domain, with distinct impacts on the updating. Simulation results pinpoint the faster convergence at lower cost for the proposed subband equalisers.
| Original language | English |
|---|---|
| Pages | 567-570 |
| Number of pages | 4 |
| DOIs | |
| Publication status | Published - 2001 |
| Event | 11th IEEE Signal Processing Workshop on Statistical Signal Processing - , Singapore Duration: 6 Aug 2001 → 8 Aug 2001 |
Workshop
| Workshop | 11th IEEE Signal Processing Workshop on Statistical Signal Processing |
|---|---|
| Country/Territory | Singapore |
| Period | 6/08/01 → 8/08/01 |
Keywords
- adaptive equalizers
- computational complexity
- computer science
- convergence
- filters