Browse Prior Art Database

A Method for Crossing Two Communication Channels in Two Dimensions

IP.com Disclosure Number: IPCOM000022106D
Original Publication Date: 2004-Feb-25
Included in the Prior Art Database: 2004-Feb-25
Document File: 2 page(s) / 79K

Publishing Venue

IBM

Abstract

Channels of communication are in three dimensions free to pass over or under each other so as to cross over without intersecting; a consideration of their two-dimensional analogue shows however that to cross two channels requires that they intersect. It is therefore natural to ask whether a system of two intersecting channels can be implemented in such a way that each can successfully carry its respective data from source to sink, without the two input-streams interfering (specifically at the intersection), and without compromising the bandwidth of the channels; this is the problem addressed and solved by the present proposal.

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 2

A Method for Crossing Two Communication Channels in Two Dimensions

Channels of communication (e.g. wires along which are sent electrical pulses which encode a bit-stream) are in three spatial dimensions free to pass over or under each other so as to cross over without intersecting; a rudimentary consideration of their two-dimensional analogue (as seen when connecting components on one layer of a chip) shows however that, assuming wires of strictly positive thickness, to cross two channels requires that they intersect. It is therefore natural to ask whether a system of two such intersecting channels can in a suitable, two-dimensional mathematical model be implemented in such a way that each channel can successfully carry its respective data from source to sink, without the two input-streams interfering (specifically at the intersection), and without compromising the bandwidth of the channels; this is the problem addressed (and solved) by the present proposal.

    There is a known solution to a special case of this problem, presented in [*], that works exclusively with binary channels (i.e. those that carry one of two states). The present proposal, in contrast, works with m -ary channels (i.e. those that carry one of m states) for any arbitrary positive integer m . This latter solution can thus replace the former, but not conversely.

    The method of solution is to split each channel into two copies ('streams') of the original, and to cross each stream of one channel with each of the other. This, coupled with the observations that (a) at each crossing point half of each message can be recovered (e.g. by multiplexing) and (b) synchronization can be such that a different half-message survives passage through each of the two streams, forms the bulk of the solution.

    This method does not depend on the number of states that can be carried by the channels (unlike the solution of [*]); neither does it reserve any of these states as 'channel identifiers'; and neither is the channels' bandwidth reduced.

    The proposed scheme is presented as a cellular automaton (though different cells are allowed different transition functions); as such, space and time are modelled discretely (as cells and ticks respectively). The states in which the cells can be are precisely those that can be carried by the channels.

    The cells' layout is as...