1 / 29

Perfect Reconstruction Two-Channel FIR Filter Banks

Perfect Reconstruction Two-Channel FIR Filter Banks. A perfect reconstruction two-channel FIR filter bank with linear-phase FIR filters can be designed if the power-complementary requirement between the two analysis filters and is not imposed.

benjamin
Download Presentation

Perfect Reconstruction Two-Channel FIR Filter Banks

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Perfect Reconstruction Two-Channel FIR Filter Banks • A perfect reconstruction two-channel FIR filter bank with linear-phase FIR filters can be designed if the power-complementary requirement between the two analysis filters and is not imposed Copyright © 2001, S. K. Mitra

  2. Perfect Reconstruction Two-Channel FIR Filter Banks • To develop the pertinent design equations we observe that the input-output relation of the 2-channel QMF bank can be expressed in matrix form as Copyright © 2001, S. K. Mitra

  3. Perfect Reconstruction Two-Channel FIR Filter Banks • From the previous equation we obtain • Combining the two matrix equations we get Copyright © 2001, S. K. Mitra

  4. Perfect Reconstruction Two-Channel FIR Filter Banks where are called themodulation matrices Copyright © 2001, S. K. Mitra

  5. Perfect Reconstruction Two-Channel FIR Filter Banks • Now for perfect reconstruction we must have and correspondingly • Substituting these relations in the equation we observe that the PR condition is satisfied if Copyright © 2001, S. K. Mitra

  6. Perfect Reconstruction Two-Channel FIR Filter Banks • Thus, knowing the analysis filters and , the synthesis filters and are determined from • After some algebra we arrive at Copyright © 2001, S. K. Mitra

  7. Perfect Reconstruction Two-Channel FIR Filter Banks where and is an odd positive integer Copyright © 2001, S. K. Mitra

  8. Perfect Reconstruction Two-Channel FIR Filter Banks • For FIR analysis filters and , the synthesis filters and will also be FIR filters if wherecis a real number andkis a positive integer • In this case Copyright © 2001, S. K. Mitra

  9. Orthogonal Filter Banks • Let be an FIR filter of odd orderNsatisfying the power-symmetric condition • Choose • Then Copyright © 2001, S. K. Mitra

  10. Orthogonal Filter Banks • Comparing the last equation with we observe that andk = N • Using in with we get Copyright © 2001, S. K. Mitra

  11. Orthogonal Filter Banks • Note:If is a causal FIR filter, the other three filters are also causal FIR filters • Moreover, • Thus, for a real coefficient transfer function if is a lowpass filter, then is a highpass filter • In addition, Copyright © 2001, S. K. Mitra

  12. Orthogonal Filter Banks • A perfect reconstruction power-symmetric filter bank is also called anorthogonal filter bank • The filter design problem reduces to the design of a power-symmetric lowpass filter • To this end, we can design a an even-order whose spectral factorization yields Copyright © 2001, S. K. Mitra

  13. Orthogonal Filter Banks • Now, the power-symmetric condition implies thatF(z) be a zero-phase half-band lowpass filter with a non-negative frequency response • Such a half-band filter can be obtained by adding a constant term K to a zero-phase half-band filter Q(z) such that Copyright © 2001, S. K. Mitra

  14. Orthogonal Filter Banks • Summarizing, the steps for the design of a real-coefficient power-symmetric lowpass filter are: • (1)Design a zero-phase real-coefficient FIR half-band lowpass filterQ(z) of order 2NwithNan odd positive integer: Copyright © 2001, S. K. Mitra

  15. Orthogonal Filter Banks • (2) Letddenote the peak stopband ripple of • DefineF(z) = Q(z) + dwhich guaranteesthat for allw • Note:Ifq[n] denotes the impulse responseofQ(z),then the impulse responsef [n] ofF(z) is given by • (3) Determine the spectral factor ofF(z) Copyright © 2001, S. K. Mitra

  16. Orthogonal Filter Banks • Example -Consider the FIR filter whereR(z) is a polynomial in of degreewithNodd • F(z) can be made a half-band filter bychoosingR(z) appropriately • This class of half-band filters has been called thebinomialormaxflat filter Copyright © 2001, S. K. Mitra

  17. Orthogonal Filter Banks • The filterF(z) has a frequency response that is maximally flat atw = 0 and atw = p • ForN = 3, resulting in which is seen to be a symmetric polynomial with 4 zeros located at , a zero at , and a zero at Copyright © 2001, S. K. Mitra

  18. Orthogonal Filter Banks • The minimum-phase spectral factor is therefore the lowpass analysis filter • The corresponding highpass filter is given by Copyright © 2001, S. K. Mitra

  19. Orthogonal Filter Banks • The two synthesis filters are given by • Magnitude responses of the two analysis filters are shown on the next slide Copyright © 2001, S. K. Mitra

  20. Orthogonal Filter Banks • Comments:(1) The order ofF(z) is of theform 4K+2,whereKis a positive integer • Order of isN = 2K+1,which is odd as required Copyright © 2001, S. K. Mitra

  21. Orthogonal Filter Banks • (2)Zeros ofF(z) appear with mirror-image symmetry in thez-plane with the zeros on the unit circle being of evenmultiplicity • Any appropriate half of these zeros can be grouped to form the spectral factor • For example, a minimum-phase can be formed by grouping all the zeros inside the unit circle along with half of the zeros on the unit circle Copyright © 2001, S. K. Mitra

  22. Orthogonal Filter Banks • Likewise, a maximum-phase can be formed by grouping all the zeros outside the unit circle along with half of the zeros on the unit circle • However, it is not possible to form a spectral factor with linear phase • (3)The stopband edge frequency is the same for F(z) and Copyright © 2001, S. K. Mitra

  23. Orthogonal Filter Banks • (4)If the desired minimum stopband attenuation of is dB, then the minimum stopband attenuation ofF(z) is dB • Example -Design a lowpass real-coefficient power-symmetric filter with the following specifications: , and dB Copyright © 2001, S. K. Mitra

  24. Orthogonal Filter Banks • The specifications of the corresponding zero-phase half-band filterF(z) are therefore: and dB • The desired stopband ripple is thus which is also the passband ripple • The passband edge is at • Using the functionremezordwe firstestimate the order ofF(z) and then using thefunctionremezdesignQ(z) Copyright © 2001, S. K. Mitra

  25. Orthogonal Filter Banks • The order ofF(z) is found to be 14 implyingthat the order of is 7 which is odd as desired • To determine the coefficients ofF(z) we adderr(the maximum stopband ripple) to thecentral coefficientq[7] • Next, using the functionrootswe determinethe roots ofF(z) which should theretically exhibit mirror-image symmetry with double roots on the unit circle Copyright © 2001, S. K. Mitra

  26. Orthogonal Filter Banks • However, the algorithm s numerically quite sensitive and it is found that a slightly larger value thanerrshould be added to ensuredouble zeros ofF(z) on the unit circle • Choosing the roots inside the unit circle along with one set of unit circle roots we get the minimum-phase spectral factor Copyright © 2001, S. K. Mitra

  27. Orthogonal Filter Banks • The zero locations ofF(z) and are shown below Copyright © 2001, S. K. Mitra

  28. Orthogonal Filter Banks • The gain responses of the two analysis filters are shown below Copyright © 2001, S. K. Mitra

  29. Orthogonal Filter Banks • Separate realizations of the two filters and would require 2(N+1) multipliers and 2Ntwo-input adders • However, a computationally efficient realization requiringN+1 multipliers and 2Ntwo-input adders can be developed by exploiting the relation Copyright © 2001, S. K. Mitra

More Related