Multilingual Management

No man is an island. In a world of many cultures, many languages and many takes on reality - but of a few common basic needs - we must communicate and cooperate to survive. Without this, there is no chance for progress, whether social, economic or technological.

Telecommunications networks once came in one big box. Not literally, of course - but with no competition among local carriers, only two wireless carriers per market and just three major long-distance carriers, the industry executives of the age never had to think much about capabilities, differentiation or even the timing of equipment installations. All they really needed was a network, and if it came from one vendor - well, that just made things less confusing.

Avoiding conflict, however, gets you only so far. Under a different set of circumstances, the very things that helped you get started in the business can cripple you later on. Reliance on proprietary solutions potentially can close you off to progress.

For network operators, there was no great realization, just a slow dawning of the facts. As they tried to figure out how to run more competitive operations, it became apparent that single-supplier relationships kept them captive to long delivery schedules, inflated costs and other vendor quirks. Broadening their supplier base naturally would bring down prices and encourage vendors to speed up delivery schedules.

Network operators with single suppliers also knew that they potentially were missing out on product features that were available in some systems but not in others. This created the urge to dabble with systems from a second maker.

In addition, the "network in a box" theory kept carriers from getting the best of everything. In other words, the vendor with the best switch was not always the one with the best digital cross-connect system or the best add/drop multiplexer. It was possible that the best network was the sum of its parts.

The change in carrier thinking led to the establishment of the Sonet standard for implementing multivendor networks through optical interconnection. But Sonet offered little in the way of clues for managing multivendor networks. Now, as the rivalry between carriers heats up, better management of network elements has suddenly emerged as a key to increasing competitiveness, and the industry is trying to address the management void.

The new EMS Old element management view: Manage your element and shut up.

New element management view: Give peace a chance - manage your elements and their interfaces with other systems within a seamless, integrated network management realm.

"In the past, element management was entirely proprietary. The element management systems were crown jewels in the network because they offered the only way of controlling specific equipment," says Alain LeFebvre, marketing manager for OS products at Hewlett-Packard, which began offering its expansive OpenView Element Management Framework last year (Figure 1).

Sonet put the major network element manufacturers on the spot. It has forced them to open interfaces from their elements to those of other vendors. Now, network elements must be able to interface with those of other makes so that information can be shared between them. This posits a challenge to the element management systems (EMSs) of old, which have acted more or less as single-minded control panels with collection plates for periodic traffic statistics and alarms.

"EMSs have really provided only an adequate measure of value because management has not always been a high priority, and the people developing network elements were more preoccupied with the elements themselves," says Leonard Donnelly, vice president and general manager of sales and marketing at Euristix, which makes the Raceman EMSX element management toolkit (Figure 2).

However, in the new multivendor networks, the singular, closed-minded view of past element management strategies will not work. Culling information from these elements will not give carriers an accurate view of network traffic unless all elements that are communicating data can be collected and presented by an EMS.

"There can be a lot of information collected on a digital cross-connect system. You need to get to all that data and massage it," says Jeff Schmitz, Titan systems manager for the Tellabs Select OSS program in Tellabs' network solutions group.

To do so, the idea of an EMS must be redesigned. While it still must basically monitor and control specific network elements, it also must collect data across a field of many network elements of different types and makes. Also, the system must be integrated with additional features for software-based element function upgrades, redundancy and multiple-element time synchronization.

In addition, the new EMSs must provide ways for carrier personnel draw real-time management information and to control the systems for activating specific management functions or even changes in network element hardware.

New EMSs also must provide a sense of security. They are being called upon as the central access points to every network element, and therefore they must maintain a historical record of who has accessed particular network elements and when.

These added capabilities will convert the EMSs into active "reach-through" facilities that communicate with other systems rather than the one-dimensional systems of the past, says Donnelly.

This migration to open systems, the original goal of Sonet several years ago, can solve any issues that multivendor network complexity can create.

Moving north The move toward more openness in EMSs is indicative of general changes in network management philosophy. It used to be a secondary thought for both carriers and vendors. While EMSs may have been one-dimensional and isolated, the rest of the network was not exactly a glowing tribute to effective management. Network management strategies, if they could even be said to exist, were highly proprietary and sewn together from legacy network databases, field tests and network surveillance systems. Everything was reactionary; nothing was proactive.

"There was never much integration of management systems before. The openness of systems necessary for real end-to-end network management did not exist," says Keith Willetts, senior vice president of marketing at TCSI and president of the Network Management Forum.

The trend toward building multivendor networks, however, began changing that even before competition arrived.

The Telecommunications Management Network (TMN) framework, which set up four distinct levels of management processes, was the solution to the network element stew. Element management became the lowest level of the framework, where functions are more detailed and focused. Network management, which constitutes the web of EMSs and more general network management functions, was next. Service management was on top of that, followed by business management. Compliance with the so-called TMN pyramid, through the creation of TMN interfaces in all management systems and network elements, became an industry goal.

While TMN is a four-level architecture, the real physical impact for carriers and vendors is happening at the network management and element management layers. "Integration has to happen between these layers, and it has to happen in a universally understood way," says Jack Harrington, president and chief executive officer of ObjectStream.

At this point, the pressure is on vendors to develop EMSs that can interface with the network management layer to initiate a smooth communications flow throughout the TMN framework (Figure 3). "We're trying to integrate northbound from our EMS to upper level solutions," says Tellabs' Schmitz. Tellabs created its newly open Titan 5300 and 5500 DCS element management system with the help of Euristix, one of several relatively young companies emerging to address the need for translating between management layers.

Bring in the interpreter The translation issue is probably the chief hurdle for EMS vendors to overcome in their race to become more open, useful systems. Changes at the element management layer must now happen in a crowded multivendor element landscape and can be further complicated by the increasing number of management protocols, model languages and management application platforms.

There was a time when vendors and network operators had to worry only about Transaction Language 1 (TL1), the basic machine-to-machine language created with Sonet. But TMN brings its own transitionary interpreter, the Q-3 interface, which all operators with legacy systems need for standards compliance. And some systems use the simple network management protocol. Also, the industry has to deal with the emergence of new object environments for designing management applications. These include the open database connectivity standard, the common object request broker architecture and Java.

EMSs existing in a multivendor environment of a long-standing carrier network likely will need to understand and interface in all these formats.

With that detailed job at hand, the model used by Tellabs - venturing with Euristix to achieve EMS openness - is becoming the model of choice for network element manufacturers.

These software specialists - mostly small, young and entrepreneurial companies - by nature can act as the translators of the many management languages, standards and object programs currently in use because, for the most part, it is all they focus on. "The old world of element management is collapsing, and we're sticking ourselves in there," says Donnelly of Euristix.

ObjectStream, Lumos and, on the larger side, Vertel are other companies that have begun to carve a niche in EMS integration.

"For the functionality and the type of management applications carriers will want, the industry will be moving away from TL1 gradually and toward more functional object-oriented models. Someone has to perform that transition," says Conor Dowling, vice president of marketing at Lumos.

For many carriers, that transition will not be painless. Although a core of software specialists will be around to help, carriers will have to adopt the stance that successful service management, which depends on strong network management, actually starts with effective element management. The quicker they can move away from a model of boxed-in EMSs, the quicker they can test this theory.

"Now, an EMS should collect information for use across the network. There is no more separation between element and network management," says HP's LeFebvre.

"With element management in the past, people tried to force a one-platform view rather than something based on the reality and specific needs of managing multivendor networks," adds Fergal Lyons, senior product development manager at Euristix.

Of course, all this new thinking about the role of EMSs in multivendor networks does not mean that carriers need to go about completely redesigning their networks any time soon.

What they do need to do is pursue a degree of assurance that the network elements they install and the EMSs they implement will not continue to strand them on their own islands.

"Multivendor element management is becoming very critical, and some carriers tell us they're nervous about how their vendors are addressing the issue," says one source.

"Carriers are taking more control in management system buying decisions. They're not just settling for one-dimensional systems," adds Lyons.

That makes some of the major vendors very eager to start to market more comprehensive, in-depth management solutions.

Northern Telecom, besides being a major network element manufacturer, recently expanded its efforts in the world of management solutions. In response to growing network complexity and carrier concerns about managing multivendor environments, the company created a self-sufficient Integrated Network Management division to design end-to-end, fully integrated network management architectures.

"Carriers wanted us to be able to manage other vendors' network elements. It's as simple as that," says John Omura, business manager for Nortel's INM unit. "The need is there to integrate element management functions into an overall network management picture in the way the TMN hierarchy intended.

Cooperative spirit The recognition of multivendor networks as necessity for carriers has positioned vendors to take the technology jumps required to effectively redesign their EMSs. However, a new, more versatile age in element management also requires a new willingness among vendors to allow their network elements to become open systems. This willingness extends to joint interoperability and even software co-development, action items that competing vendors never would have cooperated on even a few years ago.

While the current state of vendor cooperation would not be described as a bastion of friendliness, it seems many vendors are realizing that dismissing opportunities to cooperate will only complicate implementation of their own systems.

"Vendor cooperation still ranges from good to bad," says Steve Nicolle, assistant vice president and general manager at Nortel's INM unit. "Attitudes are starting to open up, but there is still a lot of protectionism.

"Lucent, Fujitsu and Nortel are ahead of the curve. Many walls have been broken down," says Tellabs' Schmitz. Tellabs itself is viewed by many industry observers as an aggressive proponent of vendor interoperability. Most vendors now readily submit to collocated lab tests with little hesitation, something that never would have happened before unless Bellcore was involved.

"The industry is learning that some level of 'coop-itition' has to happen now because Sonet is already well into a phase of second vendor deployments," adds Nicolle.

The transition to open systems, in other words, is inevitable, but the displayed willingness to cooperate with competitors, demystify network elements and build more capable and better-integrated EMSs, will ease the transition.