Frame boundary detection
View Patent ↗A method of WLAN frame detection in a received signal, wherein the frame comprises first and second training sequences and the method comprises auto-correlating the signal with a delayed version of itself to establish a first frame boundary estimate based on behavior of the autocorrelation result due to the inclusion of the first training sequence in the frame, cross-correlating the signal with a copy of the second training sequence at a range of time offsets in order to generate a first cross-correlation profile, classifying the first cross-correlation profile into one of a number of categories, establishing a second frame boundary estimate from the first cross-correlation profile in a manner dependent upon the category assigned to the first cross-correlation profile and determining a refined frame boundary estimate on the basis of a consideration of the first and second frame boundary estimates. Apparatus for performing the method is also described.
1. A method of WLAN frame detection in a received signal, wherein the frame comprises first and second training sequences and the method comprises auto-correlating the signal with a delayed version of itself to establish a first frame boundary estimate based on behaviour of the autocorrelation result due to the inclusion of the first training sequence in the frame, cross-correlating the signal with a copy of the second training sequence at a range of time offsets in order to generate a first cross-correlation profile, classifying the first cross-correlation profile into one of at least three categories, establishing a second frame boundary estimate from the first cross-correlation profile in a manner dependent upon a category of the at least three categories into which the first cross-correlation profile is classified and determining a refined frame boundary estimate on the basis of the first and second frame boundary estimates.
2. A method according to claim 1 , wherein the frame comprises a third training sequence and the method further comprises cross-correlating the signal with a copy of the third training sequence at a range of time delays in order to generate a second cross-correlation profile and boosting the first cross-correlation profile prior to its classification by coherently adding the second cross-correlation profile into the first cross-correlation profile.
3. A method according to claim 1 , wherein determining the refined frame boundary estimate comprises selecting, as the refined frame boundary estimate, the first frame boundary estimate, the second frame boundary estimate or a weighted combination of the first and second frame boundary estimates.
4. A method according to claim 3 , wherein the selection of the refined frame boundary estimate depends on a peak magnitude in the auto-correlation result.
5. A method according to claim 3 , wherein the selection of the refined frame boundary estimate depends on a difference between the first and second frame boundary estimates.
6. A method according to claim 3 , wherein the selection of the refined frame boundary estimate depends on a peak magnitude in the first cross-correlation profile.
7. A method according to claim 1 , wherein one of the at least three categories is where the first cross-correlation profile has a single peak and for that category the second frame boundary estimate is established as the position of that peak.
8. A method according to claim 1 , wherein one of the at least three categories is where the first cross-correlation profile has just two peaks and for that category the second frame boundary estimate is established as the position of the later of those two peaks.
9. A method according to claim 8 , wherein the first cross-correlation profile is smoothed before establishing the second frame boundary estimate.
10. A method according to claim 1 , wherein one of the at least three categories is where the first cross-correlation profile has more than two peaks and for that category the second frame boundary estimate is established in dependence upon the positions of the three largest peaks.
11. A method according to claim 1 , wherein one of the at least three categories is a residual category for the case where the first cross-correlation profile fits no other category and for the residual category the second frame boundary estimate is established as the position of the maximum in the profile after smoothing.
12. A method of WLAN frame detection in a received signal, wherein the frame comprises first and second training sequences and the method comprises auto-correlating the signal with a delayed version of itself to establish a first frame boundary estimate based on behaviour of the autocorrelation result due to the inclusion of the first training sequence in the frame, cross-correlating the signal with a copy of the second training sequence at a range of time offsets in order to generate a first cross-correlation profile, classifying the first cross-correlation profile into one of a number of categories based on a number of peaks in the first cross-correlation profile, establishing a second frame boundary estimate from the first cross-correlation profile in a manner dependent upon the category into which the first cross-correlation profile is classified and determining a refined frame boundary estimate on the basis of the first and second frame boundary estimates.
13. A method according to claim 12 , wherein the frame comprises a third training sequence and the method further comprises cross-correlating the signal with a copy of the third training sequence at a range of time delays in order to generate a second cross-correlation profile and boosting the first cross-correlation profile prior to its classification by coherently adding the second cross-correlation profile into the first cross-correlation profile.
14. A method according to claim 12 , wherein determining the refined frame boundary estimate comprises selecting, as the refined frame boundary estimate, the first frame boundary estimate, the second frame boundary estimate or a weighted combination of the first and second frame boundary estimates.
15. A method according to claim 14 , wherein the selection of the refined frame boundary estimate depends on a peak magnitude in at least one of the auto-correlation result and the first cross-correlation profile.
16. A method according to claim 14 , wherein the selection of the refined frame boundary estimate depends on a difference between the first and second frame boundary estimates.
17. A method according to claim 12 , wherein one of the categories is where the first cross-correlation profile has a single peak and for that category the second frame boundary estimate is established as the position of that peak.
18. A method according to claim 12 , wherein one of the categories is where the first cross-correlation profile has just two peaks and for that category the second frame boundary estimate is established as the position of the later of those two peaks.
19. A method according to claim 12 , wherein one of the categories is where the first cross-correlation profile has more than two peaks and for that category the second frame boundary estimate is established in dependence upon the positions of the three largest peaks.
20. A method according to claim 12 , wherein one of the categories is a residual category for the case where the first cross-correlation profile fits no other category and for the residual category the second frame boundary estimate is established as the position of the maximum in the profile after smoothing.