A few weeks ago, I wrote a column about the impact of wireless LAN system architecture on performance and why the debate on this issue won't be settled any time soon. This week I want to look at the impact of system and solution architecture on the success of wireless wide-area networks (WWANs).
The challenge here is more than one of scale alone. WWANs are the province of network operators and carriers, and their requirements for system performance and other features and capabilities far exceed those of any enterprise. A good example here is in what are known as Operational Support Systems (OSS), which handle literally hundreds of tasks, including billing, provisioning, trouble ticket, problem tracking and resolution, and depending upon the OSS vendor and product, a lot more.
My point, though, is the same: The radio is only a tiny part of the wireless solution value chain, and the kinds of services we as users have access to, their pricing, and the overall success of a particular carrier offering, will primarily be a function of the non-radio elements of the network -- and these will also largely determine the satisfaction that we, the customers, have with the mix and quality of services provided on carrier wireless networks.
The key to a good solution lies in a piece of carrier equipment called a gateway. Formally, a gateway exists at Layer 4 of the infamous ISO/OSI model, which, of course, is never really implemented in nice, neat layers. With just a little architectural flexibility (one of the benefits, we might argue, of open systems), gateways are in a key position to define and provision a vast array of services. While gateways of various forms have existed in WWAN architectures for some time, I want to focus this week on WiMax, and for a couple of reasons.
First, mobile WiMax is a high-profile topic at the moment, with installations proceeding around the world, and market momentum on the side of those investing big bucks in these networks. But perhaps more importantly for my purposes here, WiMAX system architecture, specified by the WiMAX Forum, actually places gateways in a very prominent and even pivotal role. As it turns out, WiMAX is based on three highly defined interfaces that allow a high degree of interoperable "mix and match" configurability, just like in enterprise networks, where we've come to take such for granted. The R1 interface is between the radio and the subscriber. R6 is the interface between the ASN (Access Service Network) gateway and the radio network, and R3 is the interface between the ASN gateway and external networks.
OK, that's two mentions of ASN gateways (ASN-GW), so I suppose this would be a good place to talk about what they do and why the gateway is more important than the radio to the success of a WiMAX solution. There's no absolute set of functions for a gateway, but its interesting location in the network, with visibility into all traffic between subscribers (via the radio networks) and the information resources those subscribers want access to (mostly the Web, but also including enterprise access, all manner of media, and, well, everything) means that the gateway is a great place to capture billing information -- a traditional gateway function. But it's also a great location to define and enforce policies, such as class of service and quality of service, filter and throttle traffic for security, integrity, and capacity management, provide location-based services and drop-in content, and generally assure the performance of both the wireless network and the applications that users will connect to via those radio links.
The WiMAX architecture allows a high degree of mix 'n' match, IP-based customization, meaning that carriers can pick the best gateway and then easily configure it to work with a potentially broad range of base stations, from macrocells to femtocells and everything in between. Think of the gateway as a big Layer-4-to-7 switch with its fingers into every aspect of network operations and subscriber traffic, and you've got the idea. Now, we users will never see the gateway directly, and, apart from those of you reading this column, most people will never know that ASN gateways even exist. But they are, I hope you will agree, the most critical buying decision that a carrier will make, largely determining how happy we're going to be with the services we're buying, and defining the next generation of services that will separate us even further from our cash.
Sure, the radio is important, but it's nothing without the rest of the network. Enough said.
Craig J. Mathias is a principal with Farpoint Group, an advisory firm specializing in wireless networking and mobile computing. He can be reached at firstname.lastname@example.org.