ATM Asynchronous Transfer Mode Networks

ATM (Asynchronous Transfer Mode Networks)

This report aims to discuss Asynchronous Transfer Mode, known simply as ATM technology, as it pertains to Networking in a detailed and coherent manner. "It is clear that Asynchronous Transfer Mode (ATM) technology will play a central role in the evolution of current workgroup, campus and enterprise networks. "The ATM protocol is flexible enough to be used not only for local and wide area network applications, but also possibly for backplane processor-to-processor interconnection. The adoption of ATM for LAN applications has altered the industry landscape and significantly speeded up the expected deployment of ATM services by WAN service providers." (Estes, 1993) ATM delivers important advantages over existing LAN and WAN technologies, including the promise of scalable bandwidths at unprecedented price and performance points and Quality of Service guarantees, which facilitate new classes of applications such as multimedia." (Alles, 1995) The objective of this work therefore is to give the details of research on the topic and thus provide specific technical insights on the Asynchronous Transfer Mode layers. These layers include the ATM Layer, AAL Convergence sub-layer, ATM Adaptation Layer or AAL, the AAL Model Segmentation and Reassembly sub-layer and the Higher Layer Protocols for control, management, and application. As noted, the work intends to focus in on the technical aspects such as networking functions, communication between layers, and the components of the Asynchronous Transfer Mode Networks technology. This objective entails that only technical information on ATM's is presented and also entails that no definitions of what ATM's are or comparisons to other architectures are presented. In addition, another area avoided was the advantages and disadvantages of ATM's.

The ATM technology is a very detail oriented and maybe the most complex technology ever created in our current technologically driven society. "While the structure of ATM cells and cell switching do facilitate the development of hardware intensive, high performance ATM switches, the deployment of ATM networks requires the overlay of a highly complex, software intensive, protocol infrastructure. This infrastructure is required to both allow individual ATM switches to be linked into a network, and for such networks to internetwork with the vast installed base of existing local and wide area networks." (Alles, 1995)

The underlying philosophy of ATM can be explained as a process that utilizes a four tiered system or layered protocol architecture. The first layer is the physical layer which defines the interface with transmission medias similar to the OSI reference model. Basically, the physical layer handles transmission rates by interfacing and detailing how ATM cells are converted into line signals. ATM cells are carried on SONET or synchronous digital hierarchy (SDH), T3/E3, TI/E1, or 9500 bps modems. It is important to note that ATM communication is independent of the physical transport so that separates it from other LAN technologies such as Ethernet protocols that specify a certain transmission medium. "The SONET technology that is used by telephone companies is considered as inadequate by data service providers. Data service providers should recognize that SONET remains the best system for providing ATM services." (Martin, 1995)

The next ATM layer handles ATM Cells which have a simple format of a 5 byte header and a 48-byte payload. Cell format headers hold key addressing information such as the ATM Cell Address as well as other vital information. The payload of the cell houses user data that will be transported over the network. Cells are forwarded serially and also propagated in a strict numeric sequence which helps to distinguish them throughout a network. When originated, payload length was created as a compromise between a long cell length of 64 bytes which would have most likely been more efficient if and when long frames of data had to be transmitted as compared to a short cell length of 32 bytes which would have minimized end-to-end processing delays and most likely would have been better for voice, video, and other delay-sensitive protocols. At 48 bytes, the cell payload length can easily accommodate two 24-byte IBP FastPackets.

To go into more detail for ATM cell formats, there are two main types of ATM cell headers: the user-network interface or UNI, and the network-to network interface or NNI. The user-network interface is an ATM specific or native-mode service interface for Wide Area Network or WAN technology. "Currently, many public network service providers are considering the deployment of public ATM networks, which will offer an ATM interconnect service across public UNI to private ATM systems. In the first instance, it is likely that the service offered across such networks will not be a pure ATM service, but will be ATM-based variants of such existing WAN technologies as Frame Relay or the Switched Multimegabit Data Service (SMDS)." (Alles, 1995) An ATM UNI is used to interface between cell-based customer premises equipment or a CPE. CPE's are things such as ATM hubs and routers that make up the foundation of any ATM WAN.

The network-to network cell would be used to define the interface between nodes throughout a network's switches or between two or more unique networks. "Although broadband ATM switches can deliver high-speed ATM user-to-network (UNI) bearer service directly to the desktop, today's market is driven by legacy data applications." (Phillips, 1995) The NNI can be the interface between a user's private ATM networks with a service provider's public ATM network.

(Alles, 1995)

"At another level, ATM has created a gold rush of start-ups seeking to mine the anticipated riches of the "next generation" networking market." (Edwards, 1994) The principal functions of the user-network interface and the network-to network interface is to categorize Virtual Path Identifiers or VPI's and Virtual Circuit Identifies or VCI's. These Virtual Path and Circuit Identifiers function as routing and switching identifiers for ATM cells. VPI clarifies the path or route that will be taken by an ATM cell and VCI clarifies the circuit or connection number that will be used while on that path. The VPI and VCI are consistently translated by every ATM switch and remain unique for single physical links.

As noted, the ATM Cell utilizes a 5-byte header for the UNI and NNI cell formats. Although the fields of these formats are similar, there is an exception for the UNI cell format which includes a 4-bit generic flow control or GFC to assist in the overall flow control for the UNI level. The exception is based on the fact that the NNI level flow control is inadvertently created through longer VPI's which permits for more virtual paths, 12 bits vs. 8 bits for UNI. Other fields for the header are the VCI which consists of 16 bits, the payload type which consists of 3 bits, the cell loss priority which is 1 bit and the header error correction of 8 bits.

The next layer of the ATM layers is the ATM Adaptation Layer or AAL. This layer functions as a source for accommodating the data from the various sources all which may have differing characteristics. In other words, the ATM Adaptation Layer's job is to adjust the services provided by the ATM layer into those services that may be needed at the higher layer levels. Examples of formats or services include circuit emulation, video, audio, frame relay, and more. The AAL collects data from applications and converts them into 48 byte segments that are sorted into the payload of the ATM cell while also defining the basic principles of sub-layering.