Introduction to Cell Switching and ATM

Cell switching is a network communication method that combines features of both circuit switching and packet switching. It is a variant of packet switching where the data packets, called cells, have a fixed, small size. This approach was designed to create a versatile network capable of efficiently handling different types of traffic (from time-sensitive voice and video to bursty computer data).

The quintessential technology that implements cell switching is ATM (Asynchronous Transfer Mode).

Why Use Small, Fixed-Size Cells?

The choice of small, fixed-length cells was a deliberate engineering decision aimed at solving the primary challenge of integrated networks: how to guarantee quality for real-time services while efficiently handling data.

• Predictable Delay (Latency): In a traditional packet network with variable packet sizes, a small, urgent voice packet might get queued behind a massive file-transfer packet. This creates unpredictable, long delays (jitter), which is disastrous for voice quality. With small cells, the maximum waiting time is short and predictable.
• Hardware Switching: The fixed and simple structure of cells makes it possible to process them very quickly in hardware. Switches do not need complex software to determine the beginning and end of each packet, enabling very high transmission speeds, often reaching hundreds of Mb/s or more.
• Single Technology for All Services: ATM was designed to be the backbone for the Broadband ISDN (B-ISDN), a single network for all services. Cells were the perfect compromise: deterministic enough for voice, flexible enough for data.

Anatomy of an ATM Cell

An ATM cell has a total length of 53 bytes, a size that was the result of a major international compromise. This is divided into a 5-byte header for control and a 48-byte payload for the user data.

Header Fields Explained

• GFC (Generic Flow Control): A 4-bit field present only at the User-Network Interface (UNI). It was intended for managing traffic flow, but was rarely used.
• VPI/VCI (Virtual Path/Channel Identifier): These two fields form the cell's address. ATM uses a two-level hierarchy: multiple Virtual Channels (VCs, identified by VCI) can be bundled into a single Virtual Path (VP, identified by VPI). This allows switches to route a whole bundle of connections by looking only at the VPI, simplifying network management.
• PT (Payload Type): Indicates the type of data in the payload field (for example, whether it contains user data or special network management data, OAM cells).
• CLP (Cell Loss Priority): A single bit that sets the cell's priority. If $CLP=1$, the cell is considered low-priority and may be discarded by the network during times of congestion. If $CLP=0$, the network will try its best to deliver it.
• HEC (Header Error Control): An 8-bit checksum calculated only for the 5-byte header. This is critical because a corrupted VPI/VCI would cause the cell to be misrouted. The HEC allows the network to correct single-bit header errors or discard cells with uncorrectable header errors. Error control for the 48-byte payload is the responsibility of higher layers.

ATM's Legacy and Influence

Although ATM did not become the universal network technology as envisioned by B-ISDN, its concepts were profoundly influential and solved many of the problems present in earlier networks.

• Failure in the LAN: ATM lost the battle for the local area network to Ethernet, which was simpler, cheaper, and evolved quickly enough to meet enterprise needs.
• Success in the Core: For a long time, ATM was a dominant technology in the backbones of telecom operators, effectively aggregating IP, voice, and Frame Relay traffic onto a single high-speed infrastructure.
• Intellectual Successor: The core idea of ATM (using short, fixed-length labels (VPI/VCI) to make fast hardware-based switching decisions) is the direct precursor to MPLS (Multi-Protocol Label Switching), which is the foundation of most modern carrier and ISP networks today.