Circuit for configuring data and energy parameters in a multi-channel communications system

• US 6,094,459 A
• Filed: 12/16/1997
• Issued: 07/25/2000
• Est. Priority Date: 12/16/1997
• Status: Expired due to Term

• Current Assignee: Realtek Semiconductor Corp.
• Original Assignee: Integrated Telecom Express, Inc.
• Inventors: Liu, Ming-Kang, Chen, Chunta, Kao, Chiihsin
• Primary Examiner(s): Chin, Stephen
• Assistant Examiner(s): GHAYOUR, MOHAMMAD H
• Application Number:
  US08/991,444
• Time in Patent Office: 952 Days
• Field of Search: 375/295, 375/260, 375/219, 375/222, 375/225, 379/93.08, 370/235, 370/231
• US Class Current: 375/295
• CPC Class Codes: H04L 5/0044 allocation of payload

Abstract

A circuit for optimizing the bit and energy configurations of data sub-channels in a multi-channel data transmission signal is disclosed. The circuit preferably includes a high speed memory coupled to a digital signal processor running an executable routine for analyzing sub-channel signal-to-noise characteristics, determining theoretical bit capacity loadings, evaluating power performance margins, and optimizing sub-channel configurations based on a series of iterative calculations intended to maximize the overall system power performance margin for any given target data rate. The circuit keeps track of the number of non-zero bit sub-channels (N_CH) from iteration to iteration, which results in a more accurate loading of the sub-channels. If adjustments to bit loadings are required to achieve a particular initial target rate, the circuit first adjusts those sub-channels which will have the least effect on the overall system performance margin, so that compliance with system requirements is better achieved. In addition, an "iteration criteria" count can be incorporated to ensure that the initialization of sub-channel loadings can be effectuated (at least to a very close order) in a predetermined and controlled fashion. The resulting bit/energy loadings can be adjusted to be fully compliant with applicable Discrete Multi-Tone (DMT) implementations of Asymmetric Digital Subscriber Loop (ADSL) protocols.

Claims

1. A circuit for use in a high speed multi-channel transmission system, which system is intended to transmit data at a system data rate R using N sub-channels, said circuit comprising:

a sub-channel parameter memory for storing K signal-to-noise associated with K sub-channels, where K≦

N; and

a processing unit for determining data capacities of each of the K sub-channels based on an evaluation of the following parameters;

i) the K signal-to-noise values; and

ii) said system data rate R; and

iii) a number N_ch of the K sub-channels having a non-zero bit capacity; and

wherein the processing unit determines the data capacities in one or more iterations of a routine executed by the processing unit, and N_ch is calculated during each iteration.

2. The circuit of claim 1, wherein said processing unit is a signal processor, and further including a microcode routine which is executed by said signal processor to determine the data capacities.

3. The circuit of claim 1, wherein the sub-channel parameter memory stores minimum (b_min (i)) and maximum (b_max (i)) bit loadings for each of the K sub-channels.

4. The circuit of claim 3, wherein the sub-channel parameter memory stores power modification factors e_min (i) and e_max (i) for each of the K sub-channels.

5. The circuit of claim 4, wherein the processing unit determines N_ch based on an evaluation of b_min (i), b_max (i), e_min (i) and e_max (i).

6. The circuit of claim 4, wherein b_min (i), b_max (i), e_min (i) and e_max (i) are adjusted dynamically as needed to meet specific system constraints.

7. The circuit of claim 1, wherein the processing unit also determines the data capacities based on a consideration of maximum transmission power P for such system.

8. The circuit of claim 7, wherein the processing unit also considers power adjustment values for the sub-channels in determining the data capacities, the power adjustment values corresponding to:

space="preserve" listing-type="equation">(e.sub.max (i)-e.sub.min (i))/(b.sub.max (i)-b.sub.min (i))

where b_min (i) and b_max (i) are calculated minimum and maximum bit loadings respectively for each of the K sub-channels, and e_min (i) and e_max (i) are sub-channel power modification factors associated with such minimum and maximum bit loadings.

9. The circuit of claim 8, wherein a value of transmission power to be used in said system is adjusted based on an evaluation of a sorted set of the power adjustment values.

10. The circuit of claim 1, wherein the processing unit initializes the system to achieve said system data rate R by generating a data rate interval consisting of two different values {B₁, B₂ } during the iterations which values bound (B₁ <

B<

B₂) a target value B, and where B=R/R_symbol, where R_symbol is a constant.

11. The circuit of claim 10, wherein the two different values B₁, B₂ are generated based on an estimated value of a power margin (γ

_m) for said system.

12. The circuit of claim 11, wherein a power margin γ

_min is associated with B₁, and a second power margin γ

_max is associated with B₂, and the processing unit achieves said system data rate R based on computing an iterative power margin γ

_m =√

γ

_min *√

γ

_max.

13. The circuit of claim 1, wherein a bit error rate γ

_b for each sub-channel can be varied.

14. The circuit of claim 1, wherein the processing unit dynamically adapts data capacities of each of the K sub-channels in response to changes of said system data rate R or transmission characteristics of a channel coupled to said system.

15. The circuit of claim 1 wherein the processing unit dynamically adapts the system to achieve said system data rate R by generating an interval consisting of two different value {B_min, B_max } which values bound (B_min <

B_target <

B_max) a target bit rate B_target.

16. The circuit of claim 15, wherein the two different values B_min, B_max are generated based on an estimated value of a power margin (γ

_m) for said system.

17. The circuit of claim 16, wherein a power margin γ

_min is associated with B_min, and a second power margin γ

_max is associated with B_max, and the processing unit achieves said data rate based on computing an iterative power margin γ

_m =√

γ

_min *√

γ

_max using a binary iteration procedure.

18. The circuit of claim 1 wherein the processing unit dynamically adapts the system to achieve said system data rate R by computing a difference Δ

B between an achieved bit rate B_total and a target bit rate B_target, and when Δ

B is not zero, an adjusted data capacity is computed for the sub-channels based on an evaluation and comparison of power margin variation to vary the data capacity for each of such sub-channels.

19. The circuit of claim 18, wherein when Δ

B is >

0, a power margin factor F₊₁ (i) relating to an excess power margin of sub-channel is computed, and data capacity can be added to the sub-channels in order beginning with the sub-channels with larger power margin factors until Δ

B is equal to zero.

20. The circuit of claim 19, wherein when Δ

B is <

0, a power margin factor F_-1 (i) relating to a power margin in such sub-channel is computed, and data capacity is removed from the sub-channels in order beginning with the sub-channels with smaller power margins until Δ

B is equal to zero.

21. The circuit of claim 20 wherein an energy loading is computed for the sub-channels based on the adjusted data capacity for such sub-channel to comply with system power transmission requirements.

22. A circuit for use in a high speed multi-channel transmission system, which system is intended to transmit data at a target data rate R using K sub-channels, said circuit comprising:

a sub-channel parameter memory for storing K signal-to-noise values associated with K sub-channels, where K<

N; and

a processing unit for determining data capacities of each of the K sub-channels based on an evaluation of the following parameters;

i) the K signal-to-noise values; and

ii) said target data rate R; and

iii) an iteration criteria count M; and

wherein the processing unit can determine the data capacities in M or fewer iterations of a configuration routine.

23. The circuit of claim 22, wherein the data processing unit calculates and uses a number N_ch of the K sub-channels having a non-zero bit capacity during the iterations to determine the data capacities.

24. The circuit of claim 22, wherein said processing unit is a signal processor, and further wherein the configuration routine is implemented in microcode which is executed by said signal processor to determine the data capacities.

25. The circuit of claim 22, wherein the processing unit also determines the data capacities based on a consideration of maximum transmission power P for such system.

26. The circuit of claim 22, wherein the processing unit also considers power adjustment values for the sub-channels in determining the data capacities, the power adjustment values corresponding to:

space="preserve" listing-type="equation">(e.sub.max (i)-e.sub.min (i))/(b.sub.max (i)-b.sub.min (i))

where b_min (i) and b_max (i) are calculated minimum and maximum bit loadings respectively for each of the K sub-channels, and e_min (i) and e_max (i) are sub-channel power modification factors associated with such minimum and maximum bit loadings.

27. The circuit of claim 26, wherein b_min (i), b_max (i), e_min (i) and e_max (i) are adjusted dynamically as needed to meet specific system constraints.

28. The circuit of claim 22, wherein the processing unit initializes the system to achieve said target data rate R by generating an interval consisting of two different values {B_min, B_max } during the M or fewer iterations which values bound (B_min <

B<sub>target<

Bmax</sub>) of a target value B_target, where B_target =R/R_symbol, and where R_symbol is a constant.

29. The circuit of claim 28, wherein the two different values B_min, B_max are generated based on an estimated value of a power margin (γ

_m) for said system.

30. The circuit of claim 29, wherein a power margin γ

_min is associated with B_min, and a second power margin γ

_max is associated with B_max, and the processing unit achieves said target data rate based on computing an iterative power margin γ

'"'"'_m =√

γ

_min *√

γ

_max by using a binary iteration method.

31. The circuit of claim 22 wherein the processing unit dynamically adapts the system to achieve said target data rate R by computing a difference Δ

B between an achieved bit data rate B_total and a target bit rate B_target, and when Δ

B is not zero, an adjusted data capacity is computed for the sub-channels based on an evaluation and comparison of power margin variation to vary the data capacity for each of such sub-channels.

32. The circuit of claim 31, wherein when Δ

B is >

0, a power margin factor F₊₁ (i) relating to an excess power margin of sub-channel is computed, and data capacity can be added to the sub-channels in order beginning with the sub-channels with larger power margin factors until Δ

B is equal to zero.

33. The circuit of claim 32, wherein when Δ

B is <

0, a power margin factor F_-1 (i) relating to a power margin in such sub-channel is computed, and data capacity is removed from the sub-channels in order beginning with the sub-channels with smaller power margins until Δ

B is equal to zero.

34. The circuit of claim 33 wherein an energy loading is computed for the sub-channels based on the adjusted data capacity for such sub-channel to comply with system power transmission requirements.

35. The circuit of claim 22, wherein a bit error rate γ

_b for each sub-channel can be varied.

36. A circuit for use in a high speed multi-channel transmission system, which system is intended to transmit data at a system data rate R using K sub-channels, and said system being coupled to a transmission channel having varying transmission characteristics, said circuit comprising:

a sub-channel parameter memory for storing signal to noise values associated with some or all of said sub-channels; and

a processing unit which can perform a configuration procedure to configure said system to operate at said system data rate R; and

wherein the configuration procedure includes operations executed by the processing unit in one or more iterations, and during each iteration a sub-channel usability determination is made of which of the K sub-channels should be disabled, such determination being based on a consideration of a calculated minimum (b_min) and maximum (b_max) bit capacity of said sub-channels; and

further wherein the processing unit allocates bit loadings only to the sub-channels which are not disabled during each iteration.

37. The circuit of claim 36, wherein the sub-channel usability determination is based on a consideration of a power modification factor e_min associated with b_min and a power modification factor e_max associated with b_max, and wherein the power modification factors correspond to an adjustment constant usable for determining a power transmission value for each of the sub-channels.

38. The circuit of claim 36, wherein the processing unit disables those sub-channels which would require, in order to achieve a particular bit capacity, power modifications larger than a predetermined threshold.

39. The circuit of claim 38, wherein the processing unit also disables those sub-channels for which b_min and b_max are both less than a predetermined value.

40. The circuit of claim 36, wherein the processing unit dynamically adapts data capacities of each of the K sub-channels in response to changes of said system data rate R or said transmission characteristics.

41. The circuit of claim 40 wherein the processing unit dynamically adapts the system to achieve said system data rate R by generating an interval consisting of two different values {B_min, B_max } which values bound (B_min <

B_target <

B_max) a target bit rate B_target and where B_target =R/R_symbol, where R_symbol is a constant.

42. The circuit of claim 41, wherein the two different values B_min, B_max are generated based on an estimated value of a power margin (γ

_m) for said system.

43. The circuit of claim 42, wherein a power margin γ

_min is associated with B_min, and a second power margin γ

_max is associated with B_max, and the processing unit achieves said system data rate based on computing an iterative power margin γ

_m =√

γ

_min *√

γ

_max using a binary iteration routine.

44. The circuit of claim 36 wherein the processing unit dynamically adapts the system to achieve said system data rate R by computing a difference Δ

B between an achieved bit rate B_total and a target bit rate B_target, and when Δ

B is not zero, an adjusted data capacity is computed for the sub-channels based on an evaluation and comparison of power margin variation to vary the data capacity for each of such sub-channels.

45. The circuit of claim 44, wherein when Δ

B is >

0, a power margin factor F₊₁ (i) relating to an excess power margin of sub-channel is computed, and data capacity can be added to the sub-channels in order beginning with the sub-channels with larger power margin factors until Δ

B is equal to zero.

46. The circuit of claim 45, wherein when Δ

B is <

0, a power margin factor F_-1 (i) relating to a power margin in such sub-channel is computed, and data capacity is removed from the sub-channels in order beginning with the sub-channels with smaller power margins until Δ

B is equal to zero.

47. The circuit of claim 46 wherein an energy loading is computed for the sub-channels based on the adjusted data capacity for such sub-channel to comply with system power transmission requirements.

48. A circuit for use in a high speed multi-channel transmission system, which system is intended to transmit data at a system target data rate R using K sub-channels and with an overall target output power value P through a channel having varying transmission characteristics, said circuit comprising:

a sub-channel parameter memory for storing signal to noise values associated with some or all of said sub-channels; and

a processing unit which can perform a configuration procedure to configure said system to operate at said system target data rate R; and

wherein the configuration procedure includes operations executed by the processing unit in one or more iterations, and during each iteration;

i) a determination is made of a proposed minimum (b_min) and a proposed maximum (b_max) bit loading for the data capacities of sub-channels that are not disabled; and

ii) a determination is made of an output power value P'"'"' associated with such proposed loading; and

iii) if P'"'"'>

P, bit capacity is removed in sorted order from those sub-channels which have the greatest power differential associated with carrying either b_min or b_max until P'"'"'=P.

49. The circuit of claim 48, further wherein during each iteration an achieved bit rate B'"'"' is calculated and compared to a target bit rate B, and the configuration procedure is executed until B'"'"'=B and P'"'"'=P.

50. The circuit of claim 49, wherein during each iteration if B'"'"'≠

B, an adjustment is made to system power margin (γ

_m) so that in later iterations B'"'"' will converge to B.

51. The circuit of claim 48, wherein during each iteration, a number N_ch of non-disabled sub-channels is calculated and used for determining P'"'"'.

52. The circuit of claim 48, wherein said processing unit is a signal processor, and further wherein the configuration procedure is a microcode routine which is executed by said signal processor to determine the data capacities.

53. The circuit of claim 48, wherein the processing unit can dynamically adapt data capacities of each of the K sub-channels in response to changes of said target data rate R or said transmission characteristics.

54. The circuit of claim 53 wherein the processing unit dynamically adapts the system to achieve said system target data rate R by generating an interval consisting of two different bit rate values {B_min, B_max } which values bound (B_min <

B_target <

B_max) a target symbol data rate B_target and where B_target =R/R_symbol, where R_symbol is a constant.

55. The circuit of claim 54, wherein the two different bit values B_min, B_max are generated based on an estimated value of a power margin (γ

_m) for said system.

56. The circuit of claim 55, wherein a power margin γ

_min is associated with bit rate B_min and a second power margin γ

_max is associated with bit rate B_max, and the processing unit achieves said data rate based on computing an iterative power margin γ

_m =√

γ

_min *√

γ

_max using a binary iteration routine.

57. The circuit of claim 48 wherein the processing unit dynamically adapts the system to achieve said target data rate R by computing a difference Δ

B between an achieved bit rate B_total and a target bit rate B_target, and when Δ

B is not zero, an adjusted data capacity is computed for the sub-channels based on an evaluation and comparison of power margin variation associated with varying the data capacity for each of such sub-channels.

58. The circuit of claim 57, wherein when Δ

B is >

0, a power margin factor F₊₁ (i) relating to an excess power margin of sub-channel is computed, and data capacity can be added to the sub-channels in order beginning with the sub-channels with larger power margin factors until Δ

B is equal to zero.

59. The circuit of claim 58 wherein when Δ

B is <

0, a power margin factor F_-1 (i) relating to a power margin in such sub-channel is computed, and data capacity is removed from the sub-channels in order beginning with the sub-channels with smaller power margins until Δ

B is equal to zero.

60. The circuit of claim 59 wherein an energy loading is computed for the sub-channels based on the adjusted data capacity for such sub-channel to comply with system power transmission requirements.

61. A circuit for use in a high speed multi-channel transmission system, which system is intended to transmit data at a system data rate R through a channel having varying transmission characteristics, and said system using K sub-channels, a total transmission power P, and a power margin γ

_m, said circuit comprising;

a sub-channel parameter memory for storing signal to noise values with some or all of said sub-channels; and

a processing unit which can perform a configuration procedure to configure data capacities; and

wherein the configuration procedure includes operations executed by the processing unit to compute a rate B_total based on the following;

i) a determination of a proposed minimum (b_min) and a proposed maximum (b_max) bit loading for the data capacities of sub-channels that are not disabled; and

ii) a determination of an output power value P'"'"' and power modification factors, e_max and e_min associated with such bit loadings; and

iii) replacing b_max with b_min, and e_max with e_min, for one or more sub-channels, until P'"'"'≦

P;

whereby B_total =Σ

b_max for all the sub-channels, and is maximized for such values of P and γ

_m.

62. The circuit of claim 61, wherein the operation of replacing b_max with b_min and e_max with e_min is done in descending order according to a sorting of the sub-channels based on the following computation for each sub-channel:

(e_max -e_min)/(b_max -b_min).

63. The circuit of claim 62, wherein B_total can be adjusted to achieve a different rate B_target by estimating an adjustment Δ

γ

_m to said system performance margin γ

_m.

64. The circuit of claim 63, wherein a target rate B_target can be achieved by the processing unit by computing two values of B_total, namely, B_min and B_max, which bound B_target (B_min <

B_target <

B_max) and by the processing unit also computing a new B_total based on a system performance margin γ

_m '"'"' which is a function of γ

_min and γ

_max , where γ

_min and γ

_max are performance margins associated with B_min and B_max respectively.

65. The circuit of claim 64 wherein γ

_m '"'"'=√

γ

_min *√

γ

_max and γ

_m is iteratively computed until B_total converges to B_target using a binary iteration routine.

66. The circuit of claim 61, wherein during the configuration procedure, a number N_ch of non-disabled sub-channels is calculated.

67. The circuit of claim 61, wherein the processing unit can dynamically adapt data capacities of each of the K sub-channels in response to changes of said target data rate R or said transmission characteristics.

68. The circuit of claim 67 wherein the processing unit dynamically adapts the system to achieve said data rate R by generating an interval consisting of two different values {B_min, B_max } which values bound (B_min <

B_target <

B_max) a target value B_target and where B_target =R/R_symbol, where R_symbol is a constant.

69. The circuit of claim 68, wherein the two different values B_min, B_max are generated based on an estimated value of a power margin (γ

_m) for said system.

70. The circuit of claim 69, wherein a power margin γ

_min is associated with B_min, and a second power margin γ

_max is associated with B_max, and the processing unit achieves said system data rate based on computing an iterative power margin γ

_m =√

γ

_min *√

γ

_max using a binary iteration routine.

71. The circuit of claim 61 wherein the processing unit dynamically adapts the system to achieve said data rate R by computing a difference Δ

B between an achieved bit rate B_total and a target bit rate B_target, and when Δ

B is not zero, an adjusted data capacity is computed for the sub-channels based on an evaluation and comparison of power margin variation to vary the data capacity for each of such sub-channels.

72. The circuit of claim 71, wherein when Δ

B is >

0, a power margin factor F₊₁ (i) relating to an excess power margin of sub-channel is computed, and data capacity can be added to the sub-channels in order beginning with the sub-channels with larger power margin factors until Δ

B is equal to zero.

73. The circuit of claim 72, wherein when Δ

B is <

0, a power margin factor F_-1 (i) relating to a power margin in such sub-channel is computed, and data capacity is removed from the sub-channels in order beginning with the sub-channels with smaller power margins until Δ

B is equal to zero.

74. The circuit of claim 73 wherein an energy loading is computed for the sub-channels based on the adjusted data capacity for such sub-channel to comply with system power transmission requirements.

75. The circuit of claim 61, wherein said processing unit is a signal processor, and further wherein the configuration procedure is a microcode routine which is executed by said signal processor to determine the data capacities.

76. The circuit of claim 61, wherein a bit error rate γ

_b for each sub-channel can be varied.