Estimation of noise variance combined with narrow-band interference detection
View Patent ↗Method and apparatus are provided for estimating noise variance using a long-term data aided algorithm and an interference variance using a short-term data aided algorithm. Using these estimations, an interference hypothesis may be determined. Some embodiments compute the variance for a decision directed noise sample, convert the variance for the decision directed noise to a true noise variance per packet, and convert the variance for the decision directed noise to a true noise variance per symbol. The interference hypothesis may be based on the noise variance estimations per symbol and the noise variance estimation per packet. Some embodiments determine the presence of noise based on a comparison of the noise variance per packet, the noise variance per symbol, and each hypothesis and compute the long-term noise variance. Using the long-term noise variance in place of the variance for a decision directed noise sample for a subsequent determination of interference.
1. A method of generating an interference hypothesis indicating the presence of a predetermined level of interference in a communication signal having at least one channel estimation symbol, the method comprising:
a) estimating a true noise variance using along-term data aided algorithm;
b) estimating an interference variance using a short-term data aided algorithm; and
c) determining an interference hypothesis based on comparing the true noise variance estimate and the interference variance estimate by:
i) of the at least one channel estimation symbol;
ii) verifying that the computed received signal power for a channel estimation symbol is within a predetermined threshold of another computed received signal power for another channel estimation symbol;
iii) comparing the true noise variance estimate to the interference variance estimate; and
iv) setting an interference hypothesis to indicate that interference is present if the computed received signal power for the channel estimation symbol is within the predetermined threshold and the true noise variance estimate is within a predetermined threshold of the interference variance estimate.
2. The method of claim 1 , further comprising setting an initial indicator to indicate that interference is not present when at least one determined interference hypothesis indicates that interference is not present.
3. The method of claim 1 , further comprising setting an interference indicator to indicate interference is present when two channel estimation symbols each contain one or more hypotheses indicating that interference is present.
4. The method of claim 1 , further comprising changing an indicator from an interference state to a noninterference state based on a predetermined number of noninterference hypotheses.
5. The method of claim 1 , further comprising changing an indicator from a noninterference state to an interference state based on a predetermined number of interference hypotheses.
6. A method of interference detection comprising:
a) estimating a true noise variance using a long-term data aided algorithm;
b) estimating an interference variance using a short-term data aided algorithm;
c) determining a first interference hypothesis based on comparing the true noise variance estimate and the interference variance estimate by:
i) computing a received signal power for at least one channel estimation symbol;
ii) verifying that the computed received signal power for a channel estimation symbol of the at least one channel estimation symbol is within a predetermined of another computed received signal power for another channel symbol;
iii) comparing the true noise variance estimate to the interference variance estimate; and
iv) setting an interference hypothesis to indicate that interference is present if the computed received signal cower for the channel estimation symbol is within the predetermined threshold and the true noise variance estimate is within a predetermined threshold:
d) determining a second interference hypothesis based on comparing true noise variance estimates and setting the second interference hypothesis to indicate that interference is present if an estimate of true noise variance using the short-term data aided algorithm is a predetermined threshold above the estimate of the true noise variance using the long-term data aided algorithm: and
e) setting an initial decision to indicate that interference is present if both of the interference hypotheses indicate that interference is present.
7. The method of claim 6 , further comprising setting an initial indicator to indicate that interference is not present when at least one determined interference hypothesis indicates that interference is not present.
8. The method of claim 6 , further comprising setting an interference indicator to indicate interference is present when two channel estimation symbols each contain one or more hypotheses indicating that interference is present.
9. The method of claim 6 , further comprising changing an indicator from an interference state to a noninterference state based on a predetermined number of noninterference hypotheses.
10. The method of claim 6 , further comprising changing an indicator from a noninterference state to an interference state based on a predetermined number of interference hypotheses.