Timing synchronization in a receiver employing orthogonal frequency division multiplexing

• US 5,953,311 A
• Filed: 02/18/1997
• Issued: 09/14/1999
• Est. Priority Date: 02/18/1997
• Status: Expired due to Term

• Current Assignee: Chartoleaux KG LLC (Intellectual Ventures LLC)
• Original Assignee: Discovision Associates (Pioneer Corporation)
• Inventors: Davies, David Huw, Claydon, Anthony Peter John, Parker, Jonathan
• Primary Examiner(s): Olms, Douglas W.
• Assistant Examiner(s): VINCENT, DAVID ROBERT
• Application Number:
  US08/802,328
• Time in Patent Office: 938 Days
• Field of Search: 370/203, 370/206, 370/208, 370/210, 370/350, 370/503, 370/504
• US Class Current: 370/210
• CPC Class Codes: H04L 27/265 Fourier transform demodulators, e.g. fast Fourier transform [FFT] or discrete Fourier transform [DFT] demodulators H04L27/26524 takes precedence
• CPC Class Codes: H04L 27/2662 Symbol synchronisation
• CPC Class Codes: H04L 27/2676 Blind, i.e. without using known symbols

Abstract

A method and apparatus are disclosed for determining the boundaries of guard intervals of data symbols being received in a coded orthogonal frequency division multiplexed signal. Temporal samples separated by an interval of an active interval of a data symbol are associated in pairs, and difference signals obtained. The dispersion of a first comparison block of difference signals is determined, and compared to the dispersion of a second comparison block of difference signals, the second comparison block being displaced from the first comparison block by n samples. An F ratio is calculated for the dispersions of the two blocks. F ratios are iterated in a succession of comparison blocks, and a signal representing the F ratios subjected to peak detection. The peaks represent the boundaries of the symbol'"'"'s guard interval. This information is utilized in synchronizing an FFT window for subsequent signal reconstruction.

Claims

1. A method of determining the boundaries of symbols in a stream of data being transmitted through a channel and having a symbol period T_s, wherein the symbols comprise an active interval and a guard interval, said guard interval being a replication of a portion of said active interval, the method comprising the steps of:

sampling first portions of said stream, wherein said first portions are separated by said active interval to form a first comparison block of data;

measuring a first dispersion of said first comparison block;

sampling second portions of said stream, wherein said second portions are separated by said active interval to form a second comparison block of data;

measuring a second dispersion of said second comparison block; and

calculating a statistic of said first dispersion and said second dispersion; and

synchronizing said stream based on said statistic, for signal reconstruction by a fast Fourier transform.

2. The method according to claim 1, wherein said statistic comprises an F ratio.

3. The method according to claim 1, wherein said symbols are complex, and said samples consist of a selected one of a real component and an imaginary component thereof.

4. A method of decoding a stream of data symbols having a symbol period T_s, wherein the symbols comprise an active interval and a guard interval, said guard interval being a replication of a portion of said active interval, the method comprising the steps of:

sampling said stream at a sampling interval during an evaluation interval of said symbol period T_s ;

associating samples in a plurality of pairs, each said pair having a first member and a second member, each said first member being separated in said stream from said second member by said active interval, wherein said first members and said second members lie in a first block and a second block respectively, and said first block and said second block are no larger than said guard interval;

for each said pair, determining a function of said first member and said second member of each said pair, and outputting a first output signal that is representative of said function;

measuring a dispersion of said first output signal;

outputting a second output signal representative of said dispersion;

displacing said first block and said second block relative said stream, and repeating said steps of associating samples, determining a function, measuring a dispersion and outputting a third output signal;

comparing said second output signal with said third output signal and outputting a fourth output signal that is representative of said comparison; and

detecting a predetermined criterion in said fourth output signal, wherein said criterion indicates boundaries of said guard interval and said active interval; and

synchronizing an FFT window using said predetermined criterion for signal reconstruction.

5. The method according to claim 4, further comprising adjusting a sign of said first output signal to yield a unipolar signal.

6. The method according to claim 4, wherein said step of sampling said stream comprises sampling an imaginary component thereof.

7. The method according to claim 4, wherein said step of sampling said stream further comprises determining a modulus of said samples.

8. The method according to claim 4, wherein said criterion is a predetermined magnitude.

9. The method according to claim 4, wherein said criterion is a maximum or a minimum value.

10. The method according to claim 4, wherein following said step of displacing said first block and said second block relative said stream, said first block is noncontiguous with said displaced first block and said second block is noncontiguous with said displaced second block.

11. The method according to claim 4, wherein in a first operating mode said evaluation interval has a first value, and in a second operating mode, said evaluation interval has a second value.

12. The method according to claim 11, further comprising the steps of, in said second operating mode:

memorizing the dispersion of said location of said guard interval in successive symbols; and

adjusting said second value of said evaluation interval in response to said memorized dispersion.

13. The method according to claim 4, further comprising the step of following said step of detecting a predetermined criterion in said fourth output signal, accepting a set of samples for an active interval before detecting a boundary of a subsequent guard interval.

14. A method of decoding a stream of data symbols having a symbol period T_s, wherein the symbols comprise an active interval and a guard interval, said guard interval being a cyclic extension of a portion of said active interval, the method comprising the steps of:

sampling said stream at a sampling interval during an evaluation interval of said symbol period T_s ;

associating samples in a plurality of pairs, each said pair having a first member and a second member, each said first member being separated in said stream from said second member by said active interval, wherein said first members and said second members lie in a first block and a second block respectively, and said first block and said second block are no larger than said guard interval;

for each said pair, determining a function of said first member and said second member of each said pair, and outputting a first output signal that is representative of said function;

measuring a dispersion of said first output signal;

outputting a second output signal representative of said dispersion;

displacing said first block and said second block relative said stream, and repeating said steps of associating samples, determining a function, measuring a dispersion and outputting a third output signal;

comparing said second output signal with said third output signal and outputting a fourth output signal that is representative of said comparison; and

detecting a predetermined criterion in said fourth output signal, wherein said criterion indicates boundaries of said guard interval and said active interval; and

synchronizing an FFT window using said predetermined criterion for signal reconstruction.

15. The method according to claim 14, wherein said step of sampling said stream comprises sampling a real component thereof.

16. The method according to claim 14, wherein said step of sampling said stream comprises sampling an imaginary component thereof.

17. The method according to claim 14, further comprising adjusting a sign of said first output signal to yield a unipolar signal.

18. The method according to claim 14, wherein said criterion is a predetermined magnitude.

19. The method according to claim 14, wherein said criterion is a maximum or a minimum value.

20. The method according to claim 14, wherein in a first operating mode said evaluation interval has a first value, and in a second operating mode, said evaluation interval has a second value.

21. The method according to claim 20, further comprising the steps of, in said second operating mode:

memorizing the dispersion of said location of said guard interval in successive symbols; and

adjusting said second value of said evaluation interval in response to said memorized dispersion.

22. The method according to claim 14, wherein said step of detecting a predetermined criterion in said fourth output signal further comprises performing a test of statistical significance.

23. The method according to claim 14, further comprising the step of following said step of detecting a predetermined criterion in said fourth output signal, accepting a set of samples for an active interval before detecting a boundary of a subsequent guard interval.

24. A method of decoding a stream of data symbols having a symbol period T_s, wherein the symbols comprise an active interval and a guard interval, said guard interval being a cyclic extension of a terminal portion of said active interval, the method comprising the steps of:

sampling said stream at a sampling interval for an evaluation interval of said symbol period T_s ;

associating samples in a plurality of pairs, each said pair having a first member and a second member, each said first member being separated in said stream from said second member by said active interval, wherein said first members and said second members lie in a first block and a second block respectively, and said first block and said second block are no larger than said guard interval;

for each said pair, determining a difference of said first member and said second member of each said pair, and outputting a first output signal that is representative of said difference;

measuring the dispersion of the magnitudes of said first output signal;

outputting a second output signal representative of said dispersion;

displacing said first block and said second block relative said stream, and repeating said steps of associating samples, determining a difference, measuring a dispersion and outputting a third output signal;

determining an F ratio of a magnitude of said second output signal and a magnitude of said third output signal and outputting a fourth output signal that is representative of said F ratio; and

detecting a predetermined magnitude in said fourth output signal, wherein said magnitude indicates boundaries of said guard interval and said active interval; and

synchronizing an FFT window using said predetermined criterion for signal reconstruction.

25. The method according to claim 24, wherein said step of sampling said stream comprises sampling a real component thereof.

26. The method according to claim 24, wherein said step of sampling said stream comprises sampling an imaginary component thereof.

27. The method according to claim 24, further comprising adjusting a sign of said first output signal to yield a unipolar signal.

28. The method according to claim 24, wherein in a first operating mode said evaluation interval has a first value, and in a second operating mode, said evaluation interval has a second value.

29. The method according to claim 28, further comprising the steps of, in said second operating mode:

memorizing the dispersion of said location of said guard interval in successive symbols; and

adjusting said second value of said evaluation interval in response to said memorized dispersion.

30. The method according to claim 24, wherein said step of determining an F ratio comprises evaluating a logarithm of the numerator of said F ratio and evaluating a logarithm of the denominator of said F ratio.

31. A method of timing synchronization for demodulating a signal that is modulated according to frequency division multiplexing, the signal comprising a stream of data symbols having a symbol period T_s, wherein the symbols comprise an active interval and a guard interval, said guard interval being a replication of a portion of said active interval, the method comprising the steps of:

temporally sampling said stream at a sampling interval for at least said symbol period T_s ;

associating samples in a plurality of pairs, each said pair having a first member and a second member, each said first member being separated in said stream from said second member by said active interval, wherein said first members and said second members lie in a first block and a second block respectively, and said first block and said second block are no larger than said guard interval;

for each said pair, determining a difference of said first member and said second member of each said pair, and outputting a first output signal that is representative of said difference;

measuring the dispersion of the magnitudes of said first output signal;

outputting a second output signal representative of said dispersion;

displacing said first block and said second block relative said stream, and repeating said steps of associating samples, determining a difference, measuring a dispersion and outputting a third output signal;

determining an F ratio of a magnitude of said second output signal and a magnitude of said third output signal and outputting a fourth output signal that is representative of said F ratio; and

detecting a predetermined magnitude in said fourth output signal, wherein said magnitude indicates boundaries of said guard interval and said active interval; and

synchronizing an FFT window using said predetermined criterion for signal reconstruction.

32. The method according to claim 31, wherein said step of sampling said stream comprises sampling a real component thereof.

33. The method according to claim 31, wherein said step of sampling said stream comprises sampling an imaginary component thereof.

34. The method according to claim 31, wherein said step of sampling said stream further comprises determining a modulus of said samples.

35. The method according to claim 31, wherein in a first operating mode said evaluation interval has a first value, and in a second operating mode, said evaluation interval has a second value.

36. The method according to claim 35, further comprising the steps of, in said second operating mode:

memorizing the dispersion of said location of said guard interval in successive symbols; and

adjusting said second value of said evaluation interval in response to said memorized dispersion.

37. The method according to claim 31, further comprising the step of following said step of detecting a predetermined criterion in said fourth output signal, accepting a set of samples for an active interval before detecting a boundary of a subsequent guard interval.

38. The method according to claim 31, wherein said step of determining an F ratio comprises evaluating a logarithm of the numerator of said F ratio and evaluating a logarithm of the denominator of said F ratio.

39. A demodulator for a frequency division multiplexed signal, the signal comprising a stream of data symbols having a symbol period T_s, wherein the symbols comprise an active interval and a guard interval, said guard interval being a replication of a portion of said active interval, wherein the improvement comprises a timing synchronization circuit for locating a boundary of said guard interval, the circuit comprising:

a first delay circuit, providing a delay for a first predetermined interval, and accepting a sampled input signal;

a first subtracter, for subtracting an output of said first delay circuit from said input signal;

a first circuit coupled to said first subtracter for producing a signal having an unipolar magnitude that is proportional to an input thereof;

a second delay circuit coupled to an output of said first circuit;

an adder/subtracter circuit, having a first input coupled said output of said first circuit, a second input connected to an output of said second delay circuit, and a third input;

a register for accumulating the output of said adder/subtracter circuit, wherein said register is connected to said third input;

a lookup memory, addressed according to said output of said adder/subtracter circuit, for accessing a logarithm of said output of said adder;

a third delay circuit, providing a delay for a second predetermined interval, and having an input connected to said lookup memory;

a second subtracter having a first input connected to said lookup memory, and a second input connected to an output of said third delay circuit; and

a comparison circuit, connected to an output of said second subtracter, for comparison thereof with a threshold value.

40. The demodulator according to claim 39, wherein said first delay circuit comprises:

a first FIFO for storing a first predetermined number of samples of said input signal;

a second FIFO for storing a second predetermined number of samples of said input signal; and

a selector for selecting one of said first FIFO and said second FIFO.

41. The demodulator according to claim 39, wherein said first circuit comprises a circuit for computing a modulus of a complex signal.

42. The demodulator according to claim 39, further comprising a control circuit for selecting predetermined samples of said input signal for processing in said timing synchronization circuit.