Adaptive thresholding algorithm for the noise due to unknown symbols in correlation based channel impulse response (CIR) estimate

1. A method for adjusting the tap weights of an equalizer comprising:

estimating an intermediate impulse response of a channel, wherein the intermediate impulse response comprises a plurality of multipath spikes and a plurality of non-deterministic noise components at locations throughout the channel;

applying a variable level threshold function to the intermediate impulse response across at least a portion of the channel in order to provide a final impulse response of the channel, wherein the variable level threshold function comprises a plurality of components, wherein each of the components comprises a statically predetermined shape spanning plural indices of the intermediate impulse response, wherein the variable level threshold function has the effect of removing the noise components of the channel having values less than the variable level threshold function at locations within the channel corresponding to the noise components, and wherein the variable level threshold function is characterized by a level that varies across the at least portion of the channel;

determining the tap weights from the final impulse response; and,

applying the tap weights to the equalizer.

2. The method of claim 1 wherein the applying of a variable level threshold function to the intermediate impulse response comprises:

forming an intermediate threshold function in response to each of the multipath spikes; and,

combining the intermediate threshold functions to form the variable level threshold function.

3. The method of claim 1 wherein the estimating of the intermediate impulse response comprises correlating a signal received over the channel with a reference to produce a correlation having a number of terms.

4. The method of claim 3 wherein the applying of a variable level threshold function comprises applying a variable level threshold function having a linear relationship with the number of terms in the correlation and a square root relationship with data in the correlation.

5. The method of claim 3 wherein the applying of a variable level threshold function to the intermediate impulse response comprises:

forming an intermediate threshold function in response to each of the multipath spikes; and,

combining the intermediate threshold functions to form the variable level threshold function.

6. The method of claim 1 wherein the applying of a variable level threshold function comprises applying a variable level threshold function based on an expectation of squared transmitted data.

7. The method of claim 1 wherein the applying of a variable level threshold function comprises applying a variable level threshold function based on an expectation of squared transmitted data and an index k of entries in the intermediate impulse response.

8. The method of claim 7 wherein the applying of a variable level threshold function based on an expectation of squared transmitted data and an index k of entries in the intermediate impulse response comprises applying a variable level threshold function according to the following equation:

E{ñ 2 [k]}=Cσ d 2 N ( k )

wherein σ d 2 comprises the variance for all transmitted data, wherein C comprises a constant, wherein E{ñ 2 [k]} comprises the expectation of squared transmitted data, wherein N(k) is given as follows:

N ( k )=− k −( L chan −1)≦ k< 0

N ( k )=0 0≦ k ≦( P−L corr )

N ( k )= k −( P−L corr ) ( P−L corr )< k ≦( L chan −1)

wherein L chan comprises a length of the channel, wherein L corr comprises a length of a correlation between a signal received over the channel and a reference, and wherein P is a constant corresponding to the length of a known training sequence.

9. The method of claim 8 wherein P comprises 728.

10. The method of claim 8 wherein the constant C is related to data in the reference.

11. The method of claim 10 wherein p comprises 728 and the reference comprises a known training sequence.

12. The method of claim 1 wherein each of the components is scaled according to a magnitude of a corresponding one of the multipath spikes.

13. A method for estimating the impulse response of a channel comprising:

estimating an initial impulse response of the channel according to a correlation of a received signal with a training sequence;

removing deterministic noise from the initial channel impulse response to produce an intermediate channel impulse response, wherein the deterministic noise comprises noise resulting substantially from autocorrelation of the training sequence;

locating a plurality of multipath spikes in the intermediate channel impulse response, wherein each of the multipath spikes has a unique index in the intermediate channel impulse response;

locating a threshold component at each of the indices of the multipath spikes, wherein each of the threshold components extends over a plurality of indices of the intermediate channel impulse response;

combining the threshold components by index to produce a final threshold; and,

applying the final threshold to the intermediate channel impulse response so as to null out noise components from the intermediate channel impulse response that have values less than the final threshold and so as to produce a final channel impulse response.

14. The method of claim 13 further comprising deriving tap weights for an equalizer from the final channel impulse response.

15. The method of claim 13 wherein the combining of the threshold components includes scaling each of the threshold components according to a magnitude of a corresponding one of the multipath spikes.

16. The method of claim 15 further comprising deriving tap weights for an equalizer from the final channel impulse response.