Skip to content Skip to navigation

Connexions

You are here: Home » Content » Data Quantization in IIR Filters

Navigation

Content Actions
  • Download module PDF
  • Add to ...
    Add the module to:
    • My Favorites
    • A lens
    • An external social bookmarking service
    • My Favorites (What is 'My Favorites'?)
      'My Favorites' is a special kind of lens which you can use to bookmark modules and collections directly in Connexions. 'My Favorites' can only be seen by you, and collections saved in 'My Favorites' can remember the last module you were on. You need a Connexions account to use 'My Favorites'.
    • A lens (What is a lens?)

      Definition of a lens

      Lenses

      A lens is a custom view of Connexions content. You can think of it as a fancy kind of list that will let you see Connexions through the eyes of organizations and people you trust.

      What is in a lens?

      Lens makers point to Connexions materials (modules and collections), creating a guide that includes their own comments and descriptive tags about the content.

      Who can create a lens?

      Any individual Connexions member, a community, or a respected organization.

      What are tags? tag icon

      Tags are descriptors added by lens makers to help label content, attaching a vocabulary that is meaningful in the context of the lens.

    • External bookmarks
  • E-mail the author
  • Rate this module (How does the rating system work?)

    Rating system

    Ratings

    Ratings allow you to judge the quality of modules. If other users have ranked the module then its average rating is displayed below. Ratings are calculated on a scale from one star (Poor) to five stars (Excellent).

    How to rate a module

    Hover over the star that corresponds to the rating you wish to assign. Click on the star to add your rating. Your rating should be based on the quality of the content. You must have an account and be logged in to rate content.

    		(0 ratings)

Recently Viewed

This feature requires Javascript to be enabled.

Data Quantization in IIR Filters

Module by: Douglas L. Jones

Summary: Quantization errors in an IIR filter structure are filtered by portions of the structure before reaching the output. The total quantization noise variance at the output is the sum of the variances of the individual filtered quantizer noises as seen at the filter output.

Note: Your browser may not currently support MathML. See our browser support page for additional details. You can always view the correct math in the PDF version.

Finite-precision effects are much more of a concern with IIR filters than with FIR filters, since the effects are more difficult to analyze and minimize, coefficient quantization errors can cause the filters to become unstable, and disastrous things like large-scale limit cycles can occur.

Roundoff noise analysis in IIR filters

Suppose there are several quantization points in an IIR filter structure. By our simplifying assumptions about quantization error and Parseval's theorem, the quantization noise variance σ y,i 2 σ y,i 2 at the output of the filter from the iith quantizer is

σ y,i 2=12π∫-ππ| H i w|2 S n i wdw= σ n i 22π∫-ππ| H i w|2dw= σ n i 2∑n=-∞∞ h i 2n σ y,i 2 1 2 w H i w 2 S S n i w σ n i 2 2 w H i w 2 σ n i 2 n h i n 2 (1)
where σ n i 2 σ n i 2 is the variance of the quantization error at the iith quantizer, S n i w S S n i w is the power spectral density of that quantization error, and H i w H H i w is the transfer function from the iith quantizer to the output point. Thus for PP independent quantizers in the structure, the total quantization noise variance is σ y 2=12π∑i=1P σ n i 2∫-ππ| H i w|2dw σ y 2 1 2 i P 1 σ n i 2 w H i w 2 Note that in general, each H i w H i w , and thus the variance at the output due to each quantizer, is different; for example, the system as seen by a quantizer at the input to the first delay state in the Direct-Form II IIR filter structure to the output, call it n 4 n 4 , is
Figure 1
Figure 1 (fig1QuantIIR.png)
with a transfer function H 4 z=z-21+ a 1 z-1+ a 2 z-2 H 4 z z -2 1 a 1 z a 2 z -2 which can be evaluated at z=ⅇⅈw z w to obtain the frequency response.

A general approach to find H i w H i w is to write state equations for the equivalent structure as seen by n i n i , and to determine the transfer function according to Hz=CzI−A-1B+d H z C z I A B d .

Figure 2
Figure 2 (fig2QuantIIR.png)

Exercise 1

The above figure illustrates the quantization points in a typical implementation of a Direct-Form II IIR second-order section. What is the total variance of the output error due to all of the quantizers in the system?

By making the assumption that each Q i Q i represents a noise source that is white, independent of the other sources, and additive,

Figure 3
Figure 3 (fig3QuantIIR.png)
the variance at the output is the sum of the variances at the output due to each noise source: σ y 2=∑i=14 σ y , i 2 σ y 2 i 1 4 σ y , i 2 The variance due to each noise source at the output can be determined from 12π∫-ππ| H i w|2 S n i wdw 1 2 w H i w 2 S n i w ; note that S n i w= σ n i 2 S n i w σ n i 2 by our assumptions, and H i w H i w is the transfer function from the noise source to the output.

Comments, questions, feedback, criticisms?

Send feedback