Network utility model is, for now, a nice dream

Peel back the onion skin-like layers of complexity behind what it will take to transform distributed LAN/WANs into a utility-like service and you reveal a daunting if not impossible task.

What the world wants is one seamless, content-rich, highly reliable, high-performance network upon which to flow all manner of communications traffic. While that may take the form of a totally reconstructed network of networks to replace the multitude of independently owned and operated voice, video and data separate systems out there, it’s more likely to be a sort of layered-on infrastructure that provides highly managed aggregation, giving customers the appearance of a utility service.

Beyond that, there are also strong indications that most businesses, if given the choice, would opt to get out of the complex and frustrating business of network building and instead lease or rent enterprise communication services. On paper, it’s all a great idea, but just as Rome wasn’t built in a day, likewise the model of a network utility won’t be built overnight – if at all.

It will take many years to construct and refine a singular communications network that extends everywhere, much like an electric power grid. Certainly from a business perspective, a data communications utility would be extremely desirable.

Data communication networks today drive almost every business process imaginable and are routinely used by many, many millions of people each day. To illustrate just how important data communications has become, simply ask any business to rank the importance of its data communications network vs. voice. Most would say the two are at least equally vital. In fact, recent IDC research revealed more companies in Canada saying their data communications networks are actually more important today than are voice services.

It’s taken decades to evolve and advance power grids to the ubiquitous, near unbreakable current forms. How long will it take to do the same for integrated voice, data and video communications services?

And with the current push towards converging voice, data and eventually video onto a single network infrastructure, the challenge of creating the network utility becomes even more difficult. Consider, too, that the current technology expected to drive all communication services is Internet Protocol or IP – a connectionless service, meaning extremely difficult to manage.

The glue that will assemble utility construction, as it is with all types of communication service infrastructures, is management. Management is the enabler of network performance and reliability, by ensuring problems are corrected before or as soon as possible after they occur. We all know that in theory you can over-provision a network to the point where you might not need to worry about performance. But such an option can be ridiculously expensive, totally impractical and it certainly won’t solve your monitoring and security issues.

Bandwidth over-provisioning was the fundamental approach used to build circuit-switched voice networks – where you essentially create circuits of 64Kbps to provision a communication service that typically only requires an average of 8Kbps on that circuit. But when bandwidth-intensive data came along, the topology proved highly inefficient.

Effective management is the way to make communications networks high-performance, reliable and, more importantly, transparent to the user. But as any network administrator will tell you management is a task easier said than done. Current systems used to manage circuit-switched voice traffic provide a high degree of richness in terms of function and service options, service quality and especially service pricing – you can price by usage and meter for the amount of time services are used. But these are, all to often, monsters and money pits.

Carrier voice network management systems have been extremely expensive to implement, take a great deal of time to evolve to maturity, remain highly proprietary systems that require high degrees of customisation, and are difficult to effect changes to. Undoubtedly, communication service providers don’t want to repeat this scenario in building utility-like services, but will they have a choice?

Also from a management perspective, as a business customer, what guarantee will I have that a communications utility provider will deliver the level of performance and reliability I demand? Not necessarily through Service Level Agreements (SLAs), which all too often put the onus on a customer to determine whether or not I received the quality and type of service I’ve paid for.

What’s really needed are Service Level Guarantees (SLGs), which place the onus the other way – on providers to ensure what’s been paid for has also been received. SLGs require service assurance management tools that provide this rich type of monitoring and the capability to regulate network performance and ensure SLGs are being met.

Consider, too, that as a customer, do I want to deal with a communications service provider utility in the same way that I currently deal with my hydroelectric or gas utilities? Customers remain faceless entities to these bureaucratic monoliths, and are often held hostage by the lack of service alternatives or the wherewithal on the part of utility companies to deal with customers in a more personalized way. Arguably, this fact may be one of the greatest inhibitors to the successful adoption of communication services as utilities.

The point here is that the utility concept is a nice dream. Certainly it’s highly desirable and there’s little doubt that the business world is particularly ready for this evolution. But as is always the case in networking, management remains a huge challenge, and in the case of building the communications power grid, many of the tools and solutions that would enable such a utility either aren’t yet available or have yet to be invented.