Metric biasing for maximum likelihood sequence estimators
US6452984B1 · kind B1 · utility
Assignee
Inventors
Key dates
| Filing date | Feb 25, 1999 |
| Grant date | Sep 17, 2002 |
| Priority date | — |
| Expiry date | Feb 25, 2019 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH03M13/39
- WIPO fieldBasic communication processes
- WIPO sectorElectrical engineering
Abstract
A decoder for decoding convolutionally encoded precoding data is described. The precoding sequence consists of two subsequences t and b; the value of t corresponds to a convolutional encoder state at a particular time. The a priori probability Pr(ti) that the subsequence t has a particular value ti is generated. In the preferred embodiment, the metric function that is applied to a test sequence including ti is then biased by adding (&sgr;2/&agr;2) ln(Pr(ti)), where &agr;2 is the signal average energy per symbol of the transmitted data, &sgr;2 is the variance of the noise added by a channel through which the data is transmitted. The resulting metric is then used by an MLSE decoder such as a Viterbi decoder to decode the received data.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.