Handbook of Local Area Networks, 1998 Edition:LAN Basics Click Here! Search the site:   ITLibrary ITKnowledge EXPERT SEARCH Programming Languages Databases Security Web Services Network Services Middleware Components Operating Systems User Interfaces Groupware & Collaboration Content Management Productivity Applications Hardware Fun & Games EarthWeb sites Crossnodes Datamation Developer.com DICE EarthWeb.com EarthWeb Direct ERP Hub Gamelan GoCertify.com HTMLGoodies Intranet Journal IT Knowledge IT Library JavaGoodies JARS JavaScripts.com open source IT RoadCoders Y2K Info Previous Table of Contents Next IEEE 802.1 Media Access Control Bridge Standards The IEEE proposed a number of standards for internetworking LANs. The IEEE 802.1 proposal calls for a media access control sublayer service common to all LANs and a network layer routing scheme for the more general case of LAN-to-any-subnetwork communications. The model identifies three LAN interconnection topologies—LAN-to-LAN, LAN-to-X.25 wide area network (WAN), and LAN-to-X.25 WAN-to-LAN. The X.25 packet-level protocol is the basis of the network layer protocol designed to provide the connection-oriented network service. The 802.1 working group’s Internetworking Functional Requirements document serves as a framework for linking networks using a bridge architecture. This proposal identifies several important functions that bridge elements must perform. They are addressing, buffering, error handling, flow control, protocol conversion segmentation and reassembly, and congestion control. The primary architectures for LAN-to-LAN and LAN-to-WAN internetworking architectures can be categorized as one of two generic classes. The first is the use of transport protocol class 4, operating over the connectionless network protocol, based on logical link control 1. This configuration supports connectionless network service and connection-oriented transport service. The second is the use of transport protocol class 1 over the X.25 packet-level protocol, which operates over logical link controls 1 or 2 to create a connection-oriented transport service. The X.25 packet-level protocol supports the following functions when implemented on a LAN: •  Addressing. The address field must be able to incorporate the network service access point information. •  Connection setup and release. These provide the data packet sequenc-ing to facilitate the connection setup and release. •  Explicit flow control. Buffer overrun and loss of data are prevented by an explicit flow control. •  Transfer of expedited data. This allows up to 32 octets of high-priority data to be transferred in an interrupt packet that is not subject flow control procedures. •  Error control. This provides for error detection and, optionally, recovery at the packet layer. •  Reset. This enables procedural error recovery. •  Q bit. This is used to qualify a packet as user data for a terminal or as control information for a packet assembler/dissembler, which could support asynchronous, bisynchronous, or synchronous data link control terminals. •  Provision of OSI connection-oriented network service. This allows internetworking with other OSI subnetworks. •  Error recovery at higher level. The proposed interconnection model is based on a number of assumptions, namely, that frames stayed in order on the links, that cyclic redundancy checks never experienced undetected errors, that the link remains functional, and that data link control modules never failed or malfunctioned. In reality, transmission errors will go undetected, links will malfunction, and there will be hardware and software failures at the nodes. As a result, the higher-level protocol must compensate for the drawbacks of the low-level protocols. One solution is to provide an adequate safeguard mechanism at the network layer so that it can provide error-free packet pipes from source to destination node. The advantages are that the error-recovery mechanism can take advantage of the mechanisms used in the network layer for routing and flow control. Because a packet must travel through several subnets to reach its destination, the end-to-end recovery can be implemented at the transport layer within the external sites. However, if the subnetworks have to alter the size of the packet, it is preferable to provide end-to-end acknowledgment on a message rather than packet basis, making error recovery more appropriate at the transport layer. Such applications as financial transactions, distributed data bases, and file transfers require lower error rates. In such cases, error-recovery mechanisms can be added at network, transport, or higher layers, or they can be combined to provide a complete structure. Address translation can be performed by a bridge or a gateway. Both perform the interconnection function but at different levels of the OSI reference model. Bridges interconnect LANs at the data link layer, whereas the gateways achieve that at the network transport, or application layer. Gateways are more suitable for connecting heterogenous architectures. INTERCONNECTION OF IEEE 802 LANS Two algorithms for interconnecting LANs through a bridge have been standardized. The first, the spanning tree bridge, originated at Digital Equipment Corp. (DEC) and is defined in ISO/IEC Standard 10038-1993. It requires that the bridge maintain a data base of the addresses of stations on the LANs. The bridge monitors all frames that travel on the two LANs. It then learns the relative locations of the stations and puts this information into its data base. As each frame arrives at the bridge, it searches the data base for the destination address contained in the frame. The forwarding of the frame to the other LAN depends on the result of the search. The second algorithm, called the source-routing transparent bridge, was developed at IBM. It is based on routing information provided by the source station, which is used by the bridge to make the switching decision. As an example, a frame may contain an ordered list of addresses of the LAN segments through which the frame must pass before reaching its destination point. This algorithm makes the job of the bridge very simple, but each station must be aware of the network status. Previous Table of Contents Next Use of this site is subject certain Terms & Conditions. Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details.