Hybrid filter bank analog/digital converter

• US 5,568,142 A
• Filed: 10/20/1994
• Issued: 10/22/1996
• Est. Priority Date: 10/20/1994
• Status: Expired due to Fees

• Current Assignee: Massachusetts Institute of Technology
• Original Assignee: Massachusetts Institute of Technology
• Inventors: Velazquez, Scott R., Nguyen, Truong O., Broadstone, Steven R.
• Primary Examiner(s): Gaffin, Jeffrey A.
• Assistant Examiner(s): KOST, JASON L
• Application Number:
  US08/326,474
• Time in Patent Office: 733 Days
• Field of Search: 341/155, 341/144, 341/141, 341/131, 341/126, 381/37
• US Class Current: 341/126
• CPC Class Codes: H03H 17/0266 Filter banks
• CPC Class Codes: H03M 1/121 Interleaved, i.e. using multiple converters or converter parts for one channel

Abstract

A hybrid filter bank analog-to-digital converter includes continuous-time analysis filters and discrete-time synthesis filters. The continuous-time analysis filters partition a continuous-time wideband input signal into continuous-time subband signals. An analog-to-digital converter bank quantizes the subband signals at a low data rate. A bank of upsamplers increases the data rate of the quantized subband signals. A bank of discrete-time synthesis filters processes the upsampled subband signals, generating signals which are the discrete-time approximation of the continuous-time subband signals. The subband signals may be recombined into a discrete-time wideband signal which is the discrete-time approximation of the continuous-time wideband input signal. The linearity errors, analog-to-digital converter mismatches and quantization noise are not compounded between the frequency bands, thereby increasing resolution. The system may also include compensators for compensating for non-linear behavior in the analog-to-digital converters and may include application-specific processors for further processing of the compensated subband signals. An alternative embodiment employs discrete-time analysis filters and continuous-time synthesis filters for digital-to-analog conversion.

Claims

1. A system for converting a signal between continuous-time and discrete-time comprising:

an analysis filter bank for partitioning a wideband input signal into first subband signals;

a converter bank of analog/digital converters for converting the first subband signals to second subband signals;

a synthesis filter bank for reconstructing the second subband signals such that third subband signals are generated which are the converted approximation of the first subband signals; and

wherein one of the analysis bank and synthesis bank comprises continuous-time filters and the other comprises discrete-time filters.

2. The system of claim 1 further comprising an adder for recombining the third subband signals into a wideband output signal which is the converted approximation of the wideband input signal.

3. The system of claim 1 wherein the analysis filters and the synthesis filters are determined by an optimization technique which iteratively adjusts the analysis and synthesis filters such that the system distortion function converges on a perfect delay and such that the system aliasing functions converge on zero.

4. The system of claim 1 implemented in a maximally decimated architecture.

5. The system of claim 1 wherein polyphase decomposition is employed such that filtering is accomplished at the lowest data rate.

6. The system of claim 1 wherein the bank of analog/digital converters includes a subset of analog/digital converters operating at a higher resolution than the remaining analog/digital converters.

7. The system of claim 1 wherein the bandwidth of the subbands is non-uniform.

8. The system of claim 1 wherein the first subband signals are oversampled by the analog/digital converters for increasing resolution.

9. The system of claim 1 implemented in a tree-structured architecture.

10. The system of claim 1 implemented in a wavelet architecture.

11. The system of claim 1 wherein the analysis filter bank comprises continuous-time filters, the converter bank comprises analog-to-digital converters, and the synthesis filter bank comprises discrete-time filters.

12. The system of claim 1 wherein the analysis filter bank comprises discrete-time filters, the converter bank comprises digital-to-analog converters, and the synthesis filter bank comprises continuous-time filters.

13. The system of claim 1 further comprising a mixing circuit for mixing the first subband signals down to baseband or up to higher frequencies.

14. The system of claim 1 wherein the discrete-time filters compensate for non-ideal analog behavior in the continuous-time filters.

15. A system for converting a continuous-time signal to a discrete-time signal comprising:

a bank of continuous-time analysis filters for partitioning a continuous-time wideband input signal into subband signals;

a bank of analog-to-digital converters for quantizing the subband signals;

a bank of rate changers for adjusting the data rate of the subband signals; and

a bank of discrete-time synthesis filters for reconstructing the subband signals such that signals are generated which are the discrete-time approximation of the continuous-time subband signals.

16. The system of claim 15 further comprising an adder for recombining the subband signals into a wideband output signal which is the discrete-time approximation of the continuous-time wideband input signal.

17. The system of claim 15 wherein the analysis filters and the synthesis filters are determined by an optimization technique which iteratively adjusts the analysis and synthesis filters such that the system distortion function converges on a perfect delay and such that the system aliasing functions converge on zero.

18. The system of claim 15 further comprising a bank of discrete-time compensators for compensating for non-ideal analog behavior in the analog-to-digital converters.

19. The system of claim 15 further comprising a bank of application specific processors for further processing of the subband signals.

20. The system of claim 19 wherein subband processing is performed before or after or both before and after the analog-to-digital converters quantize the subband signals.

21. The system of claim 15 implemented in a maximally decimated architecture.

22. The system of claim 15 wherein polyphase decomposition is employed such that synthesis filtering is accomplished at the lowest data rate.

23. The system of claim 15 wherein the bank of analog-to-digital converters includes a subset of analog-to-digital converters operating at a higher resolution than the remaining analog-to-digital converters.

24. The system of claim 15 wherein the bandwidth of the subbands is non-uniform.

25. The system of claim 15 wherein the subband signals are oversampled by the analog-to-digital converters for increasing resolution.

26. The system of claim 15 implemented in a tree-structured architecture.

27. The system of claim 15 implemented in a wavelet architecture.

28. The system of claim 15 further comprising a mixing circuit for mixing the continuous-time subband signals down to baseband.

29. A method for converting a signal between continuous-time and discrete-time comprising:

partitioning a wideband input signal into first subband signals at an analysis filter bank;

converting the first subband signals to second subband signals at a bank of analog/digital converters; and

reconstructing the second subband signals at a synthesis filter bank such that third subband signals are generated which are the converted approximation of the first subband signals and wherein one of the analysis bank and synthesis bank comprises continuous-time filters and the other comprises discrete-time filters.

30. The method of claim 29 further comprising the step of recombining the third subband signals into a wideband output signal which is the converted approximation of the wideband input signal.

31. The method of claim 29 further comprising the step of optimizing the analysis filters and the synthesis filters by iteratively adjusting the analysis and synthesis filters such that the system distortion function converges on a perfect delay and such that the system aliasing functions converge on zero.

32. The method of claim 29 wherein polyphase decomposition is employed such that analysis or synthesis filtering is accomplished at the lowest data rate.

33. The method of claim 29 further comprising the step of operating a subset of analog/digital converters at a higher resolution than the remaining analog/digital converters.

34. The method of claim 29 further comprising the step of oversampling the subband signals at the analog/digital converters for increasing resolution.

35. The method of claim 29 wherein the analysis filter bank comprises continuous-time filters, the analog/digital converter bank comprises analog-to-digital converters, and the synthesis filter bank comprises discrete-time filters.

36. The method of claim 29 wherein the analysis filter bank comprises discrete-time filters, the analog/digital converter bank comprises digital-to-analog converters, and the synthesis filter bank comprises continuous-time filters.

37. The method of claim 29 further comprising the step of mixing the continuous-time subbands down to baseband or up to higher frequencies.

38. A method for converting a continuous-time signal to a discrete-time signal comprising the steps of:

partitioning a continuous-time wideband input signal into subband signals at a bank of continuous-time analysis filters;

quantizing the subband signals at a bank of analog-to-digital converters;

adjusting the data rate of the subband signals at a bank of rate changers; and

reconstructing the subband signals at a bank of discrete-time synthesis filters for generating signals which are the discrete-time approximation of the continuous-time subband signals.

39. The method of claim 38 further comprising the step of recombining the subband signals into a wideband output signal which is the discrete-time approximation of the continuous-time wideband input signal.

40. The method of claim 38 further comprising the step of optimizing the analysis filters and the synthesis filters by iteratively adjusting the analysis and synthesis filters such that the system distortion function converges on a perfect delay and such that the system aliasing functions converge on zero.

41. The method of claim 38 further comprising the step of compensating for non-ideal analog behavior in the analog-to-digital converters at a bank of discrete-time compensators.

42. The method of claim 38 further comprising the step of processing the subband signals further at a bank of application specific processors.

43. The method of claim 42 wherein the step of processing is performed before or after or both before and after the analog-to-digital converters quantize the subband signals.

44. The method of claim 38 wherein the steps of adjusting the data rate and reconstructing the subband signals are employed in polyphase decomposition such that synthesis filtering is accomplished at the lowest data rate.

45. The method of claim 38 further comprising the step of operating a subset of analog-to-digital converters at a higher resolution than the remaining analog-to-digital converters.

46. The method of claim 38 further comprising the step of oversampling the subband signals at the analog-to-digital converters for increasing resolution.

47. The method of claim 38 further comprising the step of mixing the continuous-time subbands down to baseband.

48. In a data processing system, a method of generating a system for converting a signal between discrete-time and continuous-time comprising:

estimating a bank of continuous-time filters;

implementing the bank of continuous-time filters;

measuring the behavior of the continuous-time filter bank implementation;

estimating a bank of discrete-time filters to compensate for non-ideal behavior in the continuous-time filter bank implementation; and

adjusting the discrete-time filters such that the system distortion function converges on a perfect delay and such that the system aliasing function converges on zero.

49. The method of claim 48 wherein the conversion is from continuous-time to discrete-time.