Apparatus for and method of robust packet detection and frequency offset estimation

Patent metadata

Field	Value
Publication number	US-9184971-B2
Application number	US-73909407-A
Country	US
Kind code	B2
Filing date	Apr 23, 2007
Priority date	Apr 25, 2006
Publication date	Nov 10, 2015
Grant date	Nov 10, 2015

Abstract

A novel and useful apparatus for and method of packet detection and carrier frequency offset estimation. The packet detection mechanism is robust to channels and sustains reasonable miss-detect (and false alarm) rates at low SNR values. The mechanism uses a modified combined cross correlation and delay and correlate scheme. A delay and correlate scheme is used in order to handle the effects of multipath while swapping integration and multiplication to increase cross-correlation factors resulting in improved sensitivity in low SNR conditions. Correlation is divided into multiple chains to generate a plurality of partial correlations to observe short patterns in the spread sequence resulting in improved performance in long multipath channels.

First claim

What is claimed is:

1. A method of packet detection for use in a digital receiver, said method comprising the steps of:
delaying an input sample stream through a plurality of delay elements;
calculating a plurality of partial correlations of said plurality of delayed input samples with reference signal samples;
calculating an energy branch signal from said input sample stream;
performing nonlinear operations on said plurality of partial correlations to obtain a correlation branch signal; and
operating decision logic on said energy branch and correlation branch signals to obtain a detection decision.

2. The method according to claim 1 , wherein each said delay element comprises a single input sample delay.

3. The method according to claim 1 , wherein each said delay element comprises an N sample delay wherein N denotes the signal period.

4. The method according to claim 1 , wherein said energy branch signal is generated from the output of said plurality of delay elements.

5. The method according to claim 1 , wherein said reference signal samples comprise flipped complex conjugated reference signal samples.

6. The method according to claim 1 , wherein said step of performing nonlinear operations comprises digitally filtering the results of nonlinear processing of said partial correlations.

7. The method according to claim 1 , wherein said step of performing nonlinear operations comprises the step of multiplication of a plurality of said partial correlations thereby effectively performing a delay and correlation operation.

8. The method according to claim 1 , wherein said step of generating said energy branch signal comprises the step of applying a square magnitude operation to said input sample stream.

9. The method according to claim 1 , wherein said step of calculating said plurality of partial correlations comprises summing a plurality of delayed input samples multiplied by corresponding reference signal samples.

10. The method according to claim 1 , wherein said step of calculating a plurality of partial correlations comprises the step of summing samples having the same expected value independently of any channel distortion.

11. A method of packet detection for use in a digital receiver, said method comprising the steps of:
delaying an input sample stream through a plurality of delay elements;
calculating a plurality of partial correlations of said plurality of delayed input samples with reference signal samples;
calculating an energy branch signal from said input sample stream;
performing nonlinear operations on said plurality of partial correlations to obtain a correlation branch signal wherein said step of performing nonlinear operations comprises the steps of:
generating a square magnitude of each of said partial correlations;
summing said square magnitudes to generate a correlation branch signal therefrom;
generating said detection decision as a function of said correlation branch signal and said energy branch signal and

operating decision logic on said energy branch and correlation branch signals to obtain a detection decision.

12. A method of packet detection for use in a digital receiver, said method comprising the steps of:
delaying an input sample stream through a plurality of delay elements to generate a plurality of delayed signal samples thereby;
multiplying said first plurality of delayed signal samples with reference signal samples to generate a first plurality of products thereby;
summing said plurality of products to yield a plurality of summed values thereby;
multiplying pairs of said summed values together to generate a second plurality of products, whereby one sum is multiplied by the conjugate of another sum;
calculating an energy branch signal from said input sample stream; and
performing nonlinear detection as a function of said second plurality of products and said energy branch signal to obtain a detection decision.

13. The method according to claim 12 , wherein each said delay element comprises an N sample delay wherein N denotes the reference signal period.

14. The method according to claim 12 , wherein said reference signal samples comprise conjugated flipped versions of the delayed reference signal samples.

15. The method according to claim 12 , wherein said step of generating said energy branch signal comprises the step of applying a square magnitude operation to said input sample stream.

16. The method according to claim 12 , wherein said step of performing nonlinear detection comprises the steps of:
generating a square magnitude of each of said second plurality of products;
summing said square magnitudes to generate a correlation branch signal therefrom;
generating said output detection signal as a function of said correlation branch signal and said energy branch signal.

17. The method according to claim 12 , said step of performing nonlinear detection further comprises the step of digitally filtering said correlation branch signal and said energy branch signal before said step of generating said output detection signal.

18. The method according to claim 12 , said step of performing nonlinear detection comprises the step of asserting said detection signal active if said energy branch signal multiplied by a threshold is less than said correlation branch signal.

19. The method according to claim 12 , wherein said step of performing nonlinear detection comprises the steps of:
taking the real part of each of said second plurality of products;
summing said real parts to generate a correlation branch signal therefrom;
generating said output detection signal as a function of said correlation branch signal and said energy branch signal.

20. The method according to claim 12 , said step of performing nonlinear detection further comprises the step of digitally filtering said correlation branch signal and said energy branch signal before said step of generating said output detection signal.

21. The method according to claim 12 , said step of performing nonlinear detection comprises the step of asserting said detection signal active if said energy branch signal multiplied by a threshold is less than said correlation branch signal.

Assignees

• Lerner Gregory
• Tal Nir
• Wolberg Dan
• Singh Manoneet
• Azenkot Yehuda
• Texas Instruments Inc

Inventors

• Lerner Gregory
• Tal Nir
• Wolberg Dan
• Singh Manoneet
• Azenkot Yehuda

Classifications

• H04L2027/0065

  Frequency error detectors (H04L2027/0067 takes precedence) · CPC title

• H04L2027/0046

  Open loops · CPC title

• H04L7/042

  Detectors therefor, e.g. correlators, state machines (digital correlators in general G06F17/15) · CPC title

• H04L7/08

  the synchronisation signals recurring cyclically · CPC title

• H04L27/2656 Primary

  Frame synchronisation, e.g. packet synchronisation, time division duplex [TDD] switching point detection or subframe synchronisation · CPC title