Browse Prior Art Database

Microprogrammed Filter

IP.com Disclosure Number: IPCOM000088253D
Original Publication Date: 1977-May-01
Included in the Prior Art Database: 2005-Mar-04
Document File: 3 page(s) / 42K

Publishing Venue

IBM

Related People

Carsalade, H: AUTHOR [+3]

Abstract

A digital filter is a device which builds up each sample of the filtered signal by performing computations. Such a filter may, therefore, be designed as a specialized microprocessor.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 54% of the total text.

Page 1 of 3

Microprogrammed Filter

A digital filter is a device which builds up each sample of the filtered signal by performing computations. Such a filter may, therefore, be designed as a specialized microprocessor.

In practice, it may often be convenient to share the same device between several users, i.e., multiplex the filter for several input signals. However, this tends to make the programmer's work more complicated.

The present filter has been designed for simplified programming by conveniently using the so-called relative addressing techniques.

For an N-tap filter, the n/th/ output sample Y(n) is related to values of N-input samples through equation (1)

(Image Omitted)

where a(i) and x(n-1)(for i=0,1,...N-1) are, respectively, filter coefficients and input samples.

If there are K-input signals to share the same filter, this one should compute sequentially: and then start again for computing y/1/(n+1), y/2/(n+1),..., y/K/(n+1). New input samples x(n+1) should thus be fed into the filter, while the samples x(n-N-1) should be fed out. A data memory with associated addressing means will be used for that purpose.

The filter proposed here includes: A DATA MEM with input and output buffer registers (CREG and DREG); an accumulator with shifting means (ACCU, SHIFT); and addressing and control means stored in PROGRAM MEM (SUPERVISOR, ARITH, DYN. ADDRESS). The user's programming work is simplified by providing, in ARITH, a microprogram in which each instruction relating to an operation on a filter coefficient contains displacement information. And, in addition, the SUPERVISOR is provided with a memory storing the initial DATA MEM address assigned to the first sample of each of the K input signals,
i.e., x/1/(o), x/2/(o), ... x/k/(o). An adder ADD...