Training sequence for channel estimation in a data transmission system

• US 7,697,619 B2
• Filed: 06/09/2004
• Issued: 04/13/2010
• Est. Priority Date: 06/09/2004
• Status: Active Grant

• Current Assignee: Marvell Asia Pte Limited (Marvell Technology Group Limited)
• Original Assignee: Marvell World Trade Limited (Marvell Technology Group Limited)
• Inventors: Leung, Pak Hei Matthew, Nedic, Slobodan, Luk, King-Fuk, Tang, Eugene Yuk-Yin, Chow, Jacky S.
• Primary Examiner(s): Payne; David C
• Assistant Examiner(s): Flores; Leon
• Application Number:
  US10/865,133
• Publication Number:
  US 20050276338A1
• Time in Patent Office: 2,134 Days
• Field of Search: 375/260, 375/340, 370/210, 370/480, 379/93.01
• US Class Current: 375/260
• CPC Class Codes: H04L 25/0226 sounding signals per se
• CPC Class Codes: H04L 27/2613 Structure of the reference signals

Abstract

An improved training sequence for estimating a channel (e.g., channel impulse response) in the time domain is disclosed. The improved training sequence enables time-domain estimation and eliminates the need for correct frequency-domain channel estimation at every frequency. By utilizing the training sequence according to the invention, estimation of each of a plurality of time-domain channel taps can be decoupled from each other. This enables a channel estimate to be performed with not only a higher convergence speed but also lower complexity.

Claims

1. A data transmission system for using a training sequence to produce a channel estimate, the system comprising:

a receiver that receives a series of blocks of training samples, wherein each block of the series of blocks includes a predetermined number of training samples, and wherein for a given pair of the blocks of training samples, all samples in the given pair of the blocks of training samples are of identical value except for one particular sample in a first of the pair of blocks that is different in value from the identical values; and

channel estimation circuitry which uses the sample different in value to update individual taps of the channel estimate independently of other taps by computing a function of (a) a difference between (1) the sample different in value in the first block and (2) an other sample of the samples from a second of the pair of blocks of training samples, and (b) the channel estimate.

2. A system as recited in claim 1, wherein the series of the blocks of training samples are all time-domain training samples.

3. A system as recited in claim 1, wherein the training samples are generated and transmitted from a transmitter of the data transmission system.

4. A system as recited in claim 3, wherein the data transmission system is a multicarrier data transmission system.

5. A system as recited in claim 4, wherein the multicarrier data transmission system is digital subscriber line system.

6. A system as recited in claim 1, wherein at least the training samples that are identical in the given pairs of blocks have at least a pseudo-random characteristic.

7. A system as recited in claim 1, wherein at least the training samples that are identical in the given pairs of blocks are generated in at least a pseudo-random manner.

8. A system as recited in claim 1, wherein the blocks of the training samples include at least one of a cyclic prefix and a cyclic suffix.

9. A system as recited in claim 1, wherein each of the blocks of the training samples include N samples, and wherein the channel estimate is done using M taps, where N is greater than 2M.

10. A system as recited in claim 9, wherein a position of one particular sample within the blocks of the training symbols is after a (M−

1)th sample but before a (N−

M+1)th sample.

11. A system as recited in claim 10, wherein the series of the blocks of training samples are all time-domain training samples.

12. A system as recited in claim 11, wherein the data transmission system is a multicarrier data transmission system.

13. A system as recited in claim 12, wherein the multicarrier data transmission system is digital subscriber line system.

14. A system as recited in claim 13, wherein the training samples are generated and transmitted from a transmitter of the multicarrier data transmission system.

15. A system as recited in claim 10, wherein at least the training samples that are identical in the given pairs of blocks have at least a pseudo-random characteristic.

35. A system as recited in claim 1, wherein the function is a product operation.

36. A system as recited in claim 1, wherein the taps of the channel estimate are updated simultaneously in time and frequency domains.

16. A data transmission system for using a training sequence to produce a channel estimate, the system comprising:

a receiver that receives a series of blocks of training samples, wherein each block of the series of blocks includes a predetermined number of training samples, and wherein for a given group of the training samples, all the training samples in the given group are periodic with a periodicity of T, except for one sample of the training samples within the given group that is different in periodicity, where T is greater than or equal to the number of taps of a channel estimate; and

channel estimation circuitry which uses the training sample different in periodicity to update individual taps of the channel estimate independently of other taps by computing a function of (a) a difference between (I) the training sample different in periodicity in the given group of training samples and (2) an other training sample from the given group of training samples, and (b) the channel estimate.

17. A system as recited in claim 16, wherein the given group has a length of at least 3T.

18. A system as recited in claim 17, wherein the one of the training samples being different in the given group is positioned either (T−

1) samples after a start of the given group and (2T−

1) samples before a end of the given group or (2T−

1) samples after the start of the group and (T−

1) samples before the end of the given group.

19. A system as recited in claim 16, wherein the data transmission system is a multicarrier data transmission system.

20. A system as recited in claim 19, wherein the multicarrier data transmission system is digital subscriber line system.

21. A system as recited in claim 16, wherein those of the training samples in the group that are periodic have at least a pseudo-random characteristic.

37. A system as recited in claim 16, wherein the function is a product operation.

38. A system as recited in claim 16, wherein the different sample is in a first group of the training samples and the other sample is in a second group of the training samples.

39. A system as recited in claim 16, wherein the taps of the channel estimate are updated simultaneously in time and frequency domains.

22. A method for transmitting a training sequence for use in a multicarrier data transmission system to produce a channel estimate in channel estimation circuitry, said method comprising:

(a) obtaining a training sequence, the training sequence including a plurality of blocks of samples, wherein each block of the plurality of blocks includes a predetermined number of samples;

(b) selecting a first block of the blocks of samples from the training sequence;

(c) transmitting the first block;

(d) modifying the first block such that only one of the samples within the first block is changed;

(e) saving the modified first block into a buffer;

(f) selecting either a new block of the blocks of samples from the training sequence when not all of the training sequence has been transmitted, or a block of the blocks of samples from the buffer that stores modified blocks when that buffer is not empty;

(g) transmitting the selected block;

(h) when the selected block is the new block of the blocks of samples from the training sequence, modifying the selected block such that properties of the sample at a same location as in the modified first block is changed, wherein the changed sample is used to update individual taps of the channel estimate in the channel estimation circuitry independently of other taps by computing a function of (a) a difference between (1) the sample with changed properties from the new block and (2) an other sample from the training sequence, and (b) the channel estimate; and

(i) saving the modified selected block following said modifying (h) when the selected block is the new block.

23. A method as recited in claim 22, wherein said method further comprises:

(j) repeating said selecting (f) through said saving (i) until all the blocks of samples from the training sequence are transmitted.

40. A method as recited in claim 22, wherein the function is a product operation.

41. A method as recited in claim 22, wherein the other sample is in the modified first block.

42. A method as recited in claim 22, wherein the taps of the channel estimate are updated simultaneously in time and frequency domains.

24. A method for transmitting a training sequence for use in a multicarrier data transmission system to produce a channel estimate in channel estimation circuitry, said method comprising:

(a) obtaining a training sequence, the training sequence including a plurality of blocks of samples, wherein each block of the plurality of blocks includes a predetermined number of training samples;

(b) selecting a first block of the blocks of samples from the training sequence;

(c) transmitting the first block;

(d) modifying the first block such that properties of only one of the samples of the samples within the first block is changed;

(e transmitting the modified first block;

(f selecting a next block of the blocks of samples from the training sequence;

(g) transmitting the next block;

(h) modifying the next block such that properties of only one of the samples of the samples within the next block is changed, and wherein the changed sample is used to update individual taps of the channel estimate in the channel estimation circuitry independently of other taps by computing a function of (a) a difference between (1) the sample with changed properties from the modified next block and (2) an other sample from the modified first block, and (b) the channel estimate; and

(i) transmitting the modified next block.

25. A method as recited in claim 24, wherein the samples are time-domain samples.

26. A method as recited in claim 25, wherein the training sequence is at least pseudo-random.

27. A method as recited in claim 24, wherein said method further comprises:

(j) repeating said selecting (f) through said transmitting (i) until all the blocks of samples from the training sequence are transmitted.

43. A method as recited in claim 24, wherein the function is a product operation.

44. A method as recited in claim 24, wherein the taps of the channel estimate are updated simultaneously in time and frequency domains.

28. A method for transmitting a training sequence for use in a multicarrier data transmission system to produce a channel estimate in channel estimation circuitry, said method comprising:

(a) obtaining a group of samples from a training sequence, wherein the group of samples includes a predetermined number of samples, wherein the group of samples being periodic with a period of T, except for one sample different in periodicity, the number of samples in the group being at least 3T, wherein the different sample is used to update individual taps of the channel estimate in channel estimation circuitry independently of other taps by computing a function of (a) a difference between (I) the sample different in periodicity from the group of samples and (2) an other sample from the group of samples, and (b) the channel estimate; and

(b) transmitting the group of samples.

29. A method as recited in claim 28, wherein those of the training samples in the group that are periodic have at least a pseudo-random characteristic.

45. A method as recited in claim 28, wherein the function is a product operation.

46. A method as recited in claim 28, wherein the different sample is in a first group of samples and the other sample is in a second group of samples.

47. A method as recited in claim 28, wherein the taps of the channel estimate are updated simultaneously in time and frequency domains.

30. A computer readable medium including at least computer program code for transmitting a training sequence for use in a data transmission system to produce a channel estimate, said computer readable medium comprising:

computer program code for obtaining a training sequence, the training sequence including a plurality of blocks of samples, wherein each block of the plurality of blocks includes a predetermined number of samples;

computer program code for selecting a first block of the blocks of samples from the training sequence;

computer program code for transmitting the first block;

computer program code for modifying the first block such that properties of only one of the samples of the samples within the first block is changed, and wherein the sample with changed properties is used to update individual taps of the channel estimate independently of other taps by computing a function of (a) a difference between (1) the sample with changed properties from the first block and (2) an other sample from the training sequence, and (b) the channel estimate;

computer program code for transmitting the modified first block;

computer program code for selecting a second block of the blocks of samples from the training sequence;

computer program code for transmitting the second block;

computer program code for modifying the second block such that properties of the sample at a same location in the second block as the modified one of the samples within the first block is changed; and

computer program code for transmitting the modified second block.

31. A computer readable medium as recited in claim 30, wherein the training sequence is at least pseudo-random.

32. A computer readable medium as recited in claim 31, wherein the samples are time-domain samples.

48. A computer readable medium as recited in claim 30, wherein the function is a product operation.

49. A computer readable medium as recited in claim 30, wherein the other sample is in the modified first block samples, and (b) the channel estimate.

50. A computer readable medium as recited in claim 30, wherein the taps of the channel estimate are updated simultaneously in time and frequency domains.

33. A computer readable medium including at least computer program code for transmitting a training sequence for use in a data transmission system to produce a channel estimate, said method comprising:

computer program code for obtaining a group of samples from a training sequence, wherein the group of samples includes a predetermined number of samples, wherein the group of samples being periodic with a period of T, except for one sample different in periodicity, the number of samples in the group being at least 3T, and wherein the different sample is used to update individual taps of the channel estimate independently of other taps by computing a function of (a) a difference between (I) the sample different in periodicity from the group of samples and (2) an other sample from the group of samples, and (b) the channel estimate; and

computer program code for transmitting the group of samples.

34. A computer readable medium as recited in claim 33, wherein those of the training samples in the group that are periodic have at least a pseudo-random characteristic.

51. A computer readable medium as recited in claim 33, wherein the function is a product operation.

52. A computer readable medium as recited in claim 33, wherein the different sample is in a first group of samples and the other sample is in a second group of samples.

53. A computer readable medium as recited in claim 33, wherein the taps of the channel estimate are updated simultaneously in time and frequency domains.