Browse Prior Art Database

Dynamic Address Allocation in a Local Area Network

IP.com Disclosure Number: IPCOM000045915D
Original Publication Date: 1983-May-01
Included in the Prior Art Database: 2005-Feb-07
Document File: 3 page(s) / 41K

Publishing Venue

IBM

Related People

Janson, PA: AUTHOR [+2]

Abstract

A technique is disclosed for dynamically assigning unique addresses to stations or to their individual functions (storage, printing, etc.) in a local area network.

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

Dynamic Address Allocation in a Local Area Network

A technique is disclosed for dynamically assigning unique addresses to stations or to their individual functions (storage, printing, etc.) in a local area network.

To avoid a preassignment, e.g., in the factory, of addresses to stations, it is desirable to have a procedure for reliably assigning unique addresses to such stations when they are attached to a system and are started to operate. It must be guaranteed that such addresses remain unique for several years even if a system partially fails for some time.

The following technique is suggested: A new station or a new function (service) that is started in a network is assigned, upon request, a two-part address. The first part, e.g., two bytes, identifies the subnetwork to which the station is attached, and the second part, e.g., four bytes, designates the respective station within the subnetwork. A Directory Server (DS) is provided for each subnetwork for assigning addresses.

As long as a DS operates normally, consecutive addresses (i.e., second address parts) are taken from a 30-bit counter which is advanced after each address assignment.

After a failure of a DS, or when a DS is initialized, the respective address counter is not started from zero because it is necessary to avoid reissuance of addresses which may still be valid and in use in a portion of the system.

Instead, the new starting address is taken from the 30 lowest bits of a binary representation of the current general clock or time value. The DS has a clock counter, in addition to the address counter, which is stepped every second. After a failure, the clock value can be read from some external source.

To ensure absolute uniqueness of addresses, newly assigned addresses must not wrap around in the available address space (2/30/) while some of the earlier assigned addresses are still in use. Thus, the clo...