Dismiss
InnovationQ will be updated on Sunday, Jan. 21, from 9am - 11am ET. You may experience brief service interruptions during that time.
Browse Prior Art Database

# Three Level to Two Level Digital Code Conversion With Noise Immunity

IP.com Disclosure Number: IPCOM000078153D
Original Publication Date: 1972-Nov-01
Included in the Prior Art Database: 2005-Feb-25
Document File: 3 page(s) / 21K

IBM

## Related People

Melas, CM: AUTHOR

## Abstract

This is a method and apparatus for increasing the noise immunity in the conversion of each data sample y originally encoded in a three-level code (a+b, b, b-a) for conversion into a two-level code (A,B), the noise immunity being achieved by increasing the threshold detection levels to +/- (a+b) from +/- (a+b over 2) and defining different detection criteria.

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 53% of the total text.

Page 1 of 3

Three Level to Two Level Digital Code Conversion With Noise Immunity

This is a method and apparatus for increasing the noise immunity in the conversion of each data sample y originally encoded in a three-level code (a+b, b, b-a) for conversion into a two-level code (A,B), the noise immunity being achieved by increasing the threshold detection levels to +/- (a+b) from +/- (a+b over 2) and defining different detection criteria. The method contemplates: Step 1: forming a first difference Delta y between successive signal samples y(n-1) and y(n) such that Delta y = y(n)-y(n-1);

Step 2: assigning the value A to y(n), if either Delta y >/- a+b or Delta y >/- b and y(n-1) was assigned value A;

Step 3: assigning the value B to y(n), if either Delta y </- b-a or Delta y </- b and y(n-1) was assigned the value B;

Step 4a: delaying the detection of y(n) for k sample times, in the event that either b-a < Delta y < b y(n-1) was assigned the value A or b < Delta y < a+b and y(n-1) was assigned the value B;

Step 4b: forming another first difference Dy over the next successive k-1 samples such that: Dy = (y(n+k)-y(n)) = (y(n+1)-y(n)) + (y(n+2)-y(n+1))

(y(n+k)-y(n+k1-1));

Step 4c: assigning the value A to y(n), if either Dy </- a+b or or Dy >/- b and the next k-1 samples were assigned the value A; Step 4d: assigning the value B to y(n), if either Dy </- b-a or or Dy </- b and the next k-1 samples were assigned the value B.

The detection and code conversion criteria are ordinarily satisfied by steps 1 through 3. It is only where the magnitude difference is "in between" the threshold levels that there is uncertainty as to what is the proper value. It is under this condition that the method and apparatus contemplate the "delayed detection" of steps 4a-d. should be apparent that one assigns a value A,B to y(n) after assigning values to subsequent signal samples. In this regard, it is analogous to interpolation of an intermediate value from knowledge of other values of a function.

All these coding schemes are three level and are assumed to be +2, 0, -2.

NRZ Binary one: +2 then 0 for consecutive ones.

Binary zero: -2, then 0 for consecutive zeros.

NRZI Binary one: +2 or -2
Binary zero: 0
Both these codes cannot have two consecutive +2 or -2 samples. Duobinary: Binary one: +2, or 0 when preceded by a...