Data communications principles
Data communications principles
Electral Dispersion Compensation for 10 Gbit/s Transmission Systems: Simulation Results
ONDM '98 Proceedings of the IFIP TC6 Second International Working Conference on Optical Network Design and Modeling
H∞ optimality of the LMS algorithm
IEEE Transactions on Signal Processing
| Hi-index | 0.00 |
The widely-used LMS algorithm for coefficient updates in adaptive (feedforward/decision-feedback) equalizers is found to be suboptimal for ASE-dominant systems but various coefficient-dithering approaches suffer from slow adaptation rate without guarantee of convergence. In view of the non-Gaussian nature of optical noise after the square-law optoelectronic conversion, we propose to apply the higher-order least-mean 2Nth-order (LMN) algorithms resulting in OSNR penalty which is 1.5-2 dB less than that of LMS. Furthermore, combined with adjustable slicer threshold control, the proposed equalizer structures are demonstrated through extensive Monte Carlo simulations to achieve better performance.