SMDS Switched Multimegabit Data Service

SMDS
Switched Multimegabit Data Service

Switched Multimegabit Data Service (SMDS) is a high-speed, packet-switched, datagram-based WAN networking technology used for communication over public data networks (PDNs). SMDS can use fiber- or copper-based media and supports speeds of 1.544 Mbps over Digital Signal level 1 (DS-1) transmission facilities, or 44.736 Mbps over Digital Signal level 3 (DS-3) transmission facilities. In addition, SMDS data units are large enough to encapsulate entire IEEE 802.3, IEEE 802.5, and Fiber-Distributed Data Interface (FDDI) frames. This chapter summarizes the operational elements of the SMDS environment and outlines the underlying protocol. A discussion of related technologies, such as Distributed Queue Dual Bus (DQDB) is also provided. The chapter closes with discussions of SMDS access classes and cell formats.

SMDS Network Components

SMDS networks feature several underlying entities to provide high-speed data service. These include customer premises equipment (CPE), carrier equipment, and the subscriber network interface (SNI). CPE is terminal equipment typically owned and maintained by the customer. CPE includes end devices, such as terminals and personal computers, and intermediate nodes, such as routers, modems, and multiplexers. Intermediate nodes, however, sometimes are provided by the SMDS carrier. Carrier equipment generally consists of high-speed WAN switches that must conform to certain network equipment specifications, such as those outlined by Bell Communications Research (Bellcore). These specifications define network operations, the interface between a local carrier network and a long-distance carrier network, and the interface between two switches inside a single carrier network.

The SNI is the interface between CPE and carrier equipment. This interface is the point at which the customer network ends and the carrier network begins. The function of the SNI is to render the technology and operation of the carrier SMDS network transparent to the customer. Figure 14-1 illustrates the relationship between these three components of an SMDS network.

Figure 14-1: The SNI provides an interface between the CPE and the carrier equipment in SMDS.

SMDS Interface Protocol (SIP)

The SMDS Interface Protocol (SIP) is used for communications between CPE and SMDS carrier equipment. SIP provides connectionless service across the subscriber-network interface (SNI), allowing the CPE to access the SMDS network. SIP is based on the IEEE 802.6 Distributed Queue Dual Bus (DQDB) standard for cell relay across metropolitan-area networks (MANs). The DQDB was chosen as the basis for SIP because it is an open standard that supports all the SMDS service features. In addition, DQDB was designed for compatibility with current carrier transmission standards, and it is aligned with emerging standards for Broadband ISDN (BISDN), which will allow it to interoperate with broadband video and voice services. Figure 14-2 illustrates where SIP is used in an SMDS network.

Figure 14-2: SIP provides connectionless service between the CPE and carrier equipment.

SIP consists of three levels. SIP Level 3 operates at the Media Access Control (MAC) sublayer of the data link layer of the OSI reference model. SIP Level 2 operates at the MAC sublayer of the data link layer. SIP Level 1 operates at the physical layer of the OSI reference model. Figure 14-3 illustrates how SIP maps to the OSI reference model, including the IEEE data link sublayers.

Figure 14-3: SIP provides services associated with the physical and data link layers of the OSI model.

SIP Level 3 begins operation when user information is passed to SIP Level 3 in the form of SMDS s ervice data units (SDUs). SMDS SDUs then are encapsulated in a SIP Level 3 header and trailer. The resulting frame is called a Level 3 protocol data unit (PDU). SIP Level 3 PDUs then are subsequently passed to SIP Level 2.

SIP Level 2, which operates at the Media Access Control (MAC) sublayer of the data Level layer, begins operating when it receives SIP Level 3 PDUs. The PDUs then are segmented into uniformly sized (53-octet) Level 2 PDUs, called cells. The cells are passed to SIP Level 1 for placement on the physical medium.

SIP Level 1 operates at the physical layer and provides the physical-link protocol that operates at DS-1 or DS-3 rates between CPE devices and the network. SIP Level 1 consists of the transmission system and Physical Layer Convergency Protocol (PLCP) sublayers. The transmission system sublayer defines the characteristics and method of attachment to a DS-1 or DS-3 transmission link. The PLCP specifies how SIP Level 2 cells are to be arranged relative to the DS-1 or DS-3 frame. PLCP also defines other management information.

Distributed Queue Dual Bus (DQDB)

The Distributed Queue Dual Bus (DQDB) is a data link layer communication protocol designed for use in metropolitan-area networks (MANs). DQDB specifies a network topology composed of two unidirectional logical buses that interconnect multiple systems. It is defined in the IEEE 802.6 DQDB standard.

An access DQDB describes just the operation of the DQDB protocol (in SMDS, SIP) across a user-network interface (in SMDS, across the SNI). Such operation is distinguished from the operation of a DQDB protocol in any other environment (for example, between carrier equipment within the SMDS PDN).

The access DQDB is composed of the basic SMDS network components:

Carrier equipment---A switch in the SMDS network operates as one station on the bus.

CPE---One or more CPE devices operate as stations on the bus.

SNI---The SNI acts as the interface between the CPE and the carrier equipment.

Figure 14-4 depicts a basic access DQDB, with two CPE devices and one switch (carrier equipment) attached to the dual bus.

Figure 14-4: A basic access DQDB may consist of an end node, router, and a switch.

An SMDS access DQDB typically is arranged in a single-CPE configuration or a multi-CPE configuration.

A single-CPE access DQDB configuration consists of one switch in the carrier SMDS network and one CPE station at the subscriber site. Single-CPE DQDB configurations create a two-node DQDB subnetwork. Communication occurs only between the switch and the one CPE device across the SNI. No contention is on the bus because no other CPE devices attempt to access it.

A multi-CPE configuration consists of one switch in the carrier SMDS network and a number of interconnected CPE devices at the subscriber site (all belonging to the same subscriber). In multi-CPE configurations, local communication between CPE devices is possible. Some local communication will be visible to the switch serving the SNI, and some will not.

Contention for the bus by multiple devices requires the use of the DQDB distributed queuing algorithm, which makes implementing a multi-CPE configuration more complicated than implementing a single-CPE configuration.

SMDS Access Classes

SMDS access classes enable SMDS networks to accommodate a broad range of traffic requirements and equipment capabilities. Access classes constrain CPE devices to a sustained or average rate of data transfer by establishing a maximum sustained information transfer rate and a maximum allowed degree of traffic burstiness. (Burstiness in this context is the propensity of a network to experience sudden increases in bandwidth demand.) SMDS access classes sometimes are implemented using a credit-management scheme. In this case, a credit-management algorithm creates and tracks a credit balance for each customer interface. As packets are sent into the network, the credit balance is decremented. New credits are allocated periodically, up to an established maximum. Credit management is used only on DS-3 rate SMDS interfaces, not on DS-1 rate interfaces.

Five access classes are supported for DS-3-rate access (corresponding to sustained information rates). Data rates supported are 4, 10, 16, 25, and 34 Mbps.

SMDS Addressing Overview

SMDS protocol data units (PDUs) carry both a source and a destination address. SMDS addresses are 10-digit values resembling conventional telephone numbers.

The SMDS addressing implementation offers group addressing and security features.

SMDS group addresses allow a single address to refer to multiple CPE stations, which specify the group address in the Destination Address field of the PDU. The network makes multiple copies of the PDU, which are delivered to all members of the group. Group addresses reduce the amount of network resources required for distributing routing information, resolving addresses, and dynamically discovering network resources. SMDS group addressing is analogous to multicasting on LANs.

SMDS implements two security features: source address validation and address screening. Source address validation ensures that the PDU source address is legitimately assigned to the SNI from which it originated. Source address validation prevents address spoofing, in which illegal traffic assumes the source address of a legitimate device. Address screening allows a subscriber to establish a private virtual network that excludes unwanted traffic. If an address is disallowed, the data unit is not delivered.