The Road to an All IP Telephony Network Is Paved with SS7

What if the migration from SS7 (Signaling System 7) in telecom networks didn't happen overnight? What if new telecom applications still need to be aware of the world built on the older technologies in order to succeed?

Predictions of the Public Switched Telephone Network's (PSTN's) quick demise were premature, as were similar predictions made about every technology from mainframe computers to Ethernet to floppy disks when a successor emerged. New technology inevitably ushers out the old, but in the overwhelming majority of cases it's a gradual exit and not the bum's rush into the back alley. The inevitable shift in telecommunications from PSTN protocols to Internet Protocol (IP) is already following the evolutionary pattern, but it has a few twists that make it different from, for example, the financial service industry's transition away from mainframes to server farms. Those differences have serious implications for equipment-buying decisions over the next few years as companies build out their IP infrastructures.

Although the telecom industry has already started going down an evolutionary path similar to other industries, there are factors pushing it down the path faster than its counterparts. Industries that use information technology (IT) to support their business can afford to take their time adopting new technologies. An auto company's customers, for example, don't know if the company is using the latest information technology because it doesn't show up in the end product -- the vehicles. In the telecommunications service industry, however, IT is the end product. Service providers are already struggling to meet their customers' demands for advanced services like presence-based routing, on-demand video, gaming, and so forth that require IP infrastructures. That makes the prerogative to change more urgent than technology evolutions in other industries have been.

The fundamental reality of telecom's transition to IP is that IP must work hand-in-glove with SS7 signaling protocols during the transitional period. IP is the future, but the PSTN network core still harbors most of the important databases and application functionality, and most of the world's communication devices still use SS7 signaling for call management. Service providers that plow resources into IP networks without regard to SS7 will cut themselves off from most of their potential market, reaching only that very small percentage of customers who have mobile phones, PDAs, landlines, and other devices that support IP signaling protocols.

So what's a change-minded service provider to do during what will probably be a twenty-year transition? Those that succeed will follow a path comprised of three basic stages:

Operate parallel networks that support SS7 and IP signaling, using gateway interworking

Run SS7 protocols over IP networks

Convert to IP protocols such as SIP (Session Initiation Protocol)

It's obvious from the first two points that the path to IP runs right through SS7.

Two for the Price of One: Consider the popular Internet phone service Skype. Skype started off as a strictly IP service, with users making calls from PC to PC. When the company was ready to grow, however, it soon found its IP-only customer base was too small. The growth market still used TDM (Time Division Multiplexed) networks and SS7 signaling as their primary transport and call management mechanisms, respectively. Skype implemented gateways to combine SS7 and VoIP traffic in a single network and, fueled by a huge new customer base, Skype blossomed into the Internet's most successful VoIP service.

Skype's architecture is prototypical for what all service providers will have to do during the transition to IP. Providers can't abandon their PSTN investments without massive financial risk. Even if they could, it's not technically savvy. Technically possible, yes, but not technically savvy. Mature, highly scalable application functionality in the SS7 network core handles basic call processing functions like local number portability, address translations, and 911 database lookups. Recreating this code base in IP over a short timeframe would be expensive and prone to errors.

The most effective short-term solution is to deploy signal-processing gateways at PSTN and IP network edges so calls can traverse both networks. A 911 lookup goes to the PSTN core, while a video streaming request goes to the IP servers. This approach gives service providers the flexibility to expand their service portfolios to meet current demand, fully depreciate their PSTN investments, and replace their SS7 code bases with IP code over enough time to avoid major mistakes.

The Other End of the Evolutionary Path: IP's lower cost and superior flexibility are a potent combination for service providers trying to slake the public's thirst for new voice, text, and video services. No one with any credibility doubts that service providers are on an evolutionary course to transfer the functionality in their PSTN network cores to IP. Service providers, however, must implement IP infrastructures to offer these services without isolating them in islands that are accessible only to the small percentage of customers with devices that support IP signaling protocols. SS7 is the boss of the signaling world, confirmed by the number of devices and networks still using it. The gap between SS7 and IP signaling narrows every day, but IP still lacks SS7 signaling's functionality and reach. SS7, for its part, can't handle the advanced services like video and data streams that IP supports. Parallel SS7-IP networks with interworking gateways are the solution. They are the foundation of a successful transition to all-IP networks.

What if the signaling world is changing? What if you cannot predict which way it will go? The key to success is the sure knowledge that the road to the IP future is paved with SS7.

Steve Davis is a principal engineer at Mount Laurel, N.J.-based Ulticom (www.ulticom.com).