ATM Interworking Requires Range of Test Facilities

ATM has established itself as a reliable WAN technology. But according to a recent end-user study, it is increasingly making inroads as a transport technology for LAN traffic.1

The study also revealed that there are some misconceptions regarding the role ATM plays in LAN. “First, it has been presumed that ATM LANs are only being installed as new infrastructure—not to upgrade existing Ethernet or token ring LANs. Second, many people think ATM is only used in LANs that have special voice or video requirements,” said Kathryn Korostoff, president of Sage Research, a Massachusetts-based market research firm.

“Our research shows that both of these assumptions are out of touch with reality,” Ms. Korostoff continued. “While they may be true in some cases, the vast majority of the respondents said they used ATM to upgrade existing LANs, installing it primarily for its ease of scaling and performance benefits.”

The report found that more than 70% of the users surveyed are implementing ATM primarily as an upgrade to an existing LAN, such as, for campus, desktop, and vertical riser LAN applications or to connect servers. More than 60% report they are upgrading existing Ethernet networks and that using ATM in the LAN does not preclude employing Ethernet switches. The study also found that, in the average organization that installed ATM, more than 80% of ATM traffic is data traffic, not video and voice traffic.

The study indicates many organizations began using ATM in their campus environments and now implement it across a variety of networks. Fourteen percent use ATM at the desktop, 24% implement it between workgroups, and 28% have ATM in the WAN. Forty-four percent use ATM in the riser backbone with another 44% connect network servers with ATM. The campus backbone accounts for 85% of ATM installations in this survey.1

This information confirms that, at the present time, ATM does not exist alone, but is most frequently used in conjunction with legacy network facilities. This implies that there not only is a need for dedicated ATM test equipment, but also a possibly even greater demand for multifunctional test sets.

Dedicated and Multifunctional Instruments

Since the ATM Forum and its vendor members have been shifting their focus from basic ATM design and specification issues to integrating ATM with existing infrastructures, so have test-equipment suppliers. Dedicated, task-specific ATM analyzers and instruments are being joined by multipurpose test equipment whose functions are expanded to include ATM-related tests.

Both dedicated and multipurpose instruments have their place in testing ATM. “Dedicated test equipment is most valuable in the development lab and the early deployment of equipment or services, where attention often is focused on the functionality of one of the many communication layers,” said John Engels, product marketing manager at Telecommunications Techniques (TTC). “But as ATM matures, there is a growing need for multipurpose analyzers. Service technicians test ATM in addition to their physical layer-related maintenance duties. They need tools to isolate problems to a particular layer, whether it is the physical or the ATM layer.

The increasing use of a variety of services over ATM also reinforces the need for multipurpose test equipment. For instance, if LAN data is running over an ATM network, the service provider must make sure that both ATM and the LAN work properly.

But more than one class of multipurpose test equipment or one with a rather extensive set of capabilities may be required. This is because a variety of services may have to be accommodated at a given location, and test needs differ for backbone, edge, and signaling testing.

Backbone, Edge, and Signaling Tests

In the ATM installation phase as well as maintenance, different test-equipment capabilities may be dictated by which part of a network we are considering. Specifically, in the core backbone network such as the long-distance public transmission area, concerns center on the underlying transmission-system functionality to a much greater degree than at the more benign edge-to-customer premises environment.

Typical test considerations pertaining to the backbone, edge, and signaling are outlined by Gregan Crawford, broadband test specialist at Hewlett-Packard:

In the Backbone

ATM often is carried over large SONET/SDH networks which, in turn, may have their own quality of service problems. The transmission environment may be hostile and subject to disruption or fading, requiring APS; repeaters and multiplexers give rise to problems with jitter and wander, and there can be problems with network clock synchronization. All these are physical-layer problems which can impact the performance of the ATM being transported. As a result, test equipment for core network installation/maintenance must be just as capable of verifying proper performance of the SONET/SDH transmission systems as testing the ATM service it is carrying.

Many transmission systems incorporate built-in OAM for fault and performance management. Standards exist for OAM at the physical layer and the ATM layer, and may soon be finalized for the ATM Adaptation Layer, particularly for AAL-2. Test equipment should contain suitable facilities to make use of these processes and help pinpoint problems.

At the Edge

At the edge of the public network as well as at the customer premises, higher- layer testing becomes important. At these locations, the need for physical-layer testing is lessened since there is no APS to contend with, jitter is better controlled, and the transmission rates are lower.

Edge equipment at the public network site usually must handle the higher-layer processes, specifically those associated with interworking. For example, one of the biggest ATM users today is frame relay. Frame relay carried via ATM is a higher layer process as far as ATM is concerned. Other common interworking functions include SMDS over ATM, circuit emulation such as DS-3 over ATM, and voice telephony over ATM. Test equipment must verify the correct behavior of all these processes.

At customer premises, protocol testing becomes essential. Verification of LAN-related interworking processes requires the capabilities to test for compliance with the ATM Forum’s LAN Emulation and multiprotocol over ATM specifications, as well as to test performance per the IETF’s Classical IP over ATM specifications.

A variety of hardware interfaces exists at customer premises. Although SONET interfaces often are found operating over multimode fiber at 155 Mb/s or 622 Mb/s, ATM interfaces designed to use LAN-type unshielded twisted pair cables also exist.

Signaling

An ATM protocol analyzer with suitable test software is needed to troubleshoot signaling problems. The analyzer should have the capability to emulate the user—and possibly the network—side of the signaling state machine. Test requirements include verifying UNI signaling (ATM Forum and/or ITU-T), the ATM Forum’s ILMI responsible for address registration, and possibly the ATM Forum’s PNNI.

Multiple Interfaces and Services

The major benefits of ATM include its capability to simultaneously carry a range of services (voice, video, and data) and its transmission-speed flexibility. Data rates may range from 25 Mb/s to greater than 622 Mb/s. Of course, the transmission media differs, such as copper being used for lower speeds and fiber for higher data rates.

“For speeds of up to 25.6 Mb/s, ATM test equipment must accommodate a copper infrastructure commonly used in a LAN network, one that offers CAT-5 UTP wiring to connect to the RJ interface,” said Joe Zeto, product marketing manger at Tekelec. “Copper also is used for speeds of up to 44.736 Mb/s but with BNC connections.

“At higher ATM speeds, traffic is carried at OC-1 (51.8 Mb/s), OC-3 (155 Mb/s), and on up to OC-12 (622 Mb/s) through multimode or single-mode fibers. The physical interfaces that support these speeds normally are SC and ST connectors,” Mr. Zeto concluded.

ATM test equipment often must be used at several network nodes, each possibly operating at a different rate and using diverse physical transmission media. To accommodate the various speeds and physical terminations, most of today’s test equipment features modular implementations. Specific test-set facilities and capabilities that are important when selecting ATM test equipment were outlined by Mr. Engels of TTC:

Modular Interfaces—Modular interfaces are important so that you can upgrade your test sets as the networks and equipment mature. Test sets bought for DS3 and OC-3 today may have to support OC-12 in the near future. New types of ATM interfaces, such as ADSL, should be available to help test and deploy new services and technology.

Rate-Independent Features—Test-set features and setups should be relatively consistent regardless of the interface. You should not be forced to learn and use new displays simply because the transmission rates are different. Test setups used for one interface should be equally applicable to other interfaces so you can spend your time testing, not setting up the test set.

Monitoring Capability—Test equipment should support both in-service and out-of-service testing. For out-of-service testing, the test set must monitor attenuated signals from monitor ports or optical splitters. If no standard monitoring points exist, the test set should support in-line analysis, typically called through mode. This capability allows you to place the test set in the circuit and pass the data through the set while it analyzes that data.

Past and Future of ATM

For several reasons, the deployment of ATM has progressed at a much slower pace than originally anticipated. As one reason, Mr. Crawford of Hewlett-Packard cited the apparent instability of standards and interoperability agreements, a situation which led to the establishment of the Anchorage Accord.

“The Anchorage Accord was a declaration by the ATM Forum that the core set of ATM specifications, such as Traffic Management 4.0, Signaling 4.0, and ILMI 4.0, was accepted as stable,” explained Mr. Crawford. “The forum decided that the core would not undergo any major revisions. Only corrections would be made when mistakes are discovered.”

Cost- and complexity-related causes for the slow acceptance of ATM were enumerated by TTC’s Mr. Engels:

Cost/Value—ATM generally has been more expensive, due to its recent development, complexity, and the slow emergence of standard ICs to deal with ATM. Since ATM equipment relies more on custom hardware, it costs more than older technologies such as Ethernet or SONET. However, recent announcements from many chip manufacturers that have developed ATM chip sets will help provide lower-cost ATM solutions in the future.

Complexity—ATM is technically complex and simultaneously must be integrated with legacy technologies. As a result, it takes a substantial effort to provide the legacy support and the features promised by the ATM community.

“The delayed deployment of ATM has obviously slowed the market for ATM test sets, particularly for field equipment,” added Mr. Engels. “But a catch-up is imminent. The market share for lab test equipment has been, and is, far greater for ATM than for other technologies. And over time more field service test sets will be required.”

“The strength of ATM lies in its flexibility and scaling capabilities—almost any service can be carried over ATM,” emphasized HP’s Mr. Crawford. “As a result, there will be a growing acceptance of ATM.” Consequently, more test-equipment suppliers will develop new ATM-targeted test equipment or extend their product offerings to service the ATM market with add-on modules or ATM-targeted options.

References

1. End-User Study on Asynchronous Transfer Mode (ATM) Deployment in the Local Area Network (LAN) conducted by the ATM Forum in association with Sage Research. Presented at NetWorld+Interop Atlanta in 1997 and reported by Business Wire via Individual Inc.

Acknowledgments

The following companies provided information for this feature:

General Signal Networks (609) 234-7900

Hewlett-Packard (800) 452-4844

RADCOM (201) 529-2020

Tekelec (818) 880-5656

Telecommunications Techniques (800) 638-2049

Sidebar

Glossary of Terms

AAL-2:


ATM Adaptation Layer, Type 2


APS:


Automatic Protection Switching


ATM:


Asynchronous Transfer Mode


ADSL:


Asymmetrical Digital Subscriber Line


CAT-5:


Category 5 Cabling Performance Level


DS3:


Digital Service, level 3, signal speed of 44.736 Mb/s


IETF:


Internet Engineering Task Force


ILMI:


Interim Link Management Interface


IP:


Internet Protocol


ITU-T:


International Telecommunications Union Telecommunications


LAN:


Local Area Network


OAM:


Operations, Administration and Maintenance


OC-1:


Optical Carrier Level 1 (basic rate, 51.840 Mb/s)


OC-3:


Optical Carrier level-3 (3 times a SONET optical signal of OC-1)


OC-12:


Optical Carrier level-12 (SONET Channel of 622.08 Mb/s)


PNNI:


Private Network-Network Interface


RJ:


Registered Jacks (registered with the FCC)


SC, ST:


Types of Optical Fiber Connectors


SMDS:


Switched Megabit Data Service


SONET/SDH:


Synchronous Optical NETwork/Synchronous Digital Hierarchy


UNI:


User-Network Interface


UTP-3:


Unshielded Twisted Pair


WAN:


Wide Area Network


 

 

Copyright 1998 Nelson Publishing Inc.

April 1998




Sponsored Recommendations

Comments

To join the conversation, and become an exclusive member of Electronic Design, create an account today!