Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Network Resource Identification Algorithm

IP.com Disclosure Number: IPCOM000074406D
Original Publication Date: 1971-Apr-01
Included in the Prior Art Database: 2005-Feb-23
Document File: 4 page(s) / 57K

Publishing Venue

IBM

Related People

Duffie, CA: AUTHOR [+2]

Abstract

Described is a method of establishing resource (e.g., line, terminal component, etc.) identification such that the network configuration is transparent to the Central processor. The design goal in defining this algorithm is as follows: 1) The central processor shall maintain a directory of binary coded resource identification codes (I.D.) for each resource in the system, yet need not know to which communications controller any resource is attached. 2) Each communications controller shall only be responsible for the system resources that are direct]y attached to it. 3) There shall exist a key transformation algorithm that, once applied by a given communications controller to a binary I.D.

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 46% of the total text.

Page 1 of 4

Network Resource Identification Algorithm

Described is a method of establishing resource (e.g., line, terminal component, etc.) identification such that the network configuration is transparent to the Central processor. The design goal in defining this algorithm is as follows:
1) The central processor shall maintain a directory of

binary coded resource identification codes (I.D.) for

each resource in the system, yet need not know to which

communications controller any resource is attached.
2) Each communications controller shall only be responsible

for the system resources that are direct]y attached to it.
3) There shall exist a key transformation algorithm that,

once applied by a given communications controller to a

binary I.D., will give an indication as to whether one

of its attached resources is being addressed, or if not,

over which attached communications line the transaction

should be forwarded such that it will eventually reach

the appropriate resource.

The one constraint imposed by the resource identification algorithm is that it only applies to fan networks. That is to say, for each communications controller and resource in the network, there can be one and only one direct path back to the central processor.

The identification technique is as follows: Each addressable resource in the system is identified to the central processor by a unique binary code. This code is assigned when the system is configured and is configuration dependant to a slight degree. That is, if a resource is moved from one communications controller to another, its binary code I.D. will change. Or, if a new controller is added to the network, the binary code I.D.'s for all resources attached to the controller that immediately precedes the new one in the path back to the central processor will change. The method of assigning resource I.D.'s is discussed later.

Each addressable resource attached to a given communications controller is identified by that controller through a binary index relative to zero. This is convenient because the index gives rapid access to an internal vector table which in turn accesses the appropriate internal resource control block.

Each communications controller that possesses lines to one or more remote communications controllers will perform the following key transformation as it receives a transaction with a binary resource I.D. code. The binary code I.D. is divided by a value one greater than the number of lines the controller extends to remote controllers, e.g., a controller with two remotes attached to it would divide by three. This value is hereafter referred to as the transform divisor. If the result of the division yields a zero remainder, then the controller can deduce that the quotient is the binary index into its internal vector table. If the result of the division yields a nonzero remainder, then the controller can deduce that the resource is not locally attached, and it must therefore pass the quotient as the

1

P...