Generated by DeepSeek V3.2| Asynchronous Transfer Mode | |
|---|---|
![]() | |
| Name | Asynchronous Transfer Mode |
| Developer | International Telecommunication Union, ATM Forum |
| Introduced | Early 1990s |
| Industry | Telecommunications, networking |
Asynchronous Transfer Mode. Asynchronous Transfer Mode (ATM) is a high-performance, cell-switching and multiplexing technology designed to unify telecommunications and computer networks. It was developed in the late 1980s and early 1990s as a core component of the Broadband Integrated Services Digital Network (B-ISDN) standard, promising to efficiently carry diverse traffic types like voice, video, and data. Although it achieved significant deployment in carrier backbones and certain enterprise settings, its complexity and the rise of simpler alternatives like Internet Protocol over Ethernet led to its decline as a universal solution.
The fundamental goal was to create a single, integrated network infrastructure capable of supporting the burgeoning demands of multimedia applications. Pioneered by the International Telecommunication Union's Telecommunication Standardization Sector, the technology was heavily promoted by major players like Cisco Systems, Nortel, and Alcatel-Lucent. It gained formal standardization through the collaborative efforts of the ATM Forum, an industry consortium. Key early applications were seen in the backbones of major Internet service providers, such as MCI Communications and Sprint Corporation, and within the public switched telephone network for trunking.
The core innovation was the use of fixed-length, 53-byte cells, comprising a 5-byte header and a 48-byte payload. This small, consistent size was chosen to minimize jitter and latency, critical for real-time traffic like voice carried over the public switched telephone network. The header contained fields for Virtual Path Identifier and Virtual Channel Identifier which together established virtual circuits, along with control bits for Payload Type Identifier, Cell Loss Priority, and Header Error Control. Switching occurred in hardware based on these identifiers, enabling very high-speed forwarding in devices like the Cisco 7000 series.
The protocol stack is structured in layers analogous to the OSI model. The physical layer defines media types, including Synchronous Optical Networking and Synchronous Digital Hierarchy for wide-area networks, and UTP cable for local use. The ATM layer handles core cell switching and multiplexing functions. Above this, the ATM adaptation layer, with types like AAL1 for constant bit-rate services and AAL5 for data, segments higher-layer packets into cells. This architecture interfaced with legacy systems via Frame Relay and Switched Multimegabit Data Service interworking functions, and supported sophisticated traffic management through parameters like Peak Cell Rate and Sustainable Cell Rate.
It found major success as a scalable backbone technology within the infrastructure of telcos and ISPs, connecting major points of presence. Enterprises deployed it for high-speed campus backbone networks, often using LAN Emulation to bridge with existing Ethernet segments. It was also the transport of choice for many early Digital Subscriber Line access multiplexers, forming the core network between the DSLAM and the Internet. Furthermore, it was mandated for the User-Network Interface and Network-to-Network Interface in early implementations of MPLS.
Compared to variable-length packet technologies like Internet Protocol over Ethernet, it offered superior, guaranteed Quality of Service through its connection-oriented virtual circuits and traffic contracts. This made it initially superior for isochronous traffic compared to shared-media Token Ring or early Fast Ethernet. However, it was often contrasted with the simpler, connectionless model of IP routing, which eventually dominated. While competing with SMDS and Frame Relay for data services, its complexity and cost were higher than these more focused technologies, and it was ultimately supplanted by the widespread adoption of Gigabit Ethernet and advanced IP QoS mechanisms like DiffServ.
The rapid evolution and cost reduction of Internet Protocol and Ethernet technologies, driven by the explosive growth of the World Wide Web, rendered the complex, cell-based architecture less economical. The development of MPLS, which borrowed the virtual circuit concept but operated over IP, provided a more flexible alternative for carrier networks. Key intellectual property and concepts, such as traffic engineering and the notion of Quality of Service, were directly inherited by MPLS and later software-defined networking initiatives like those from the Open Networking Foundation. Its influence persists in the foundational infrastructure of some legacy telecommunications networks and as a historical benchmark for integrated services network design.
Category:Network protocols Category:Telecommunications standards Category:Data transmission Category:ITU-T recommendations