The analogy I’m drawing here is between an operator’s IN platform and an enterprise’s / web service provider’s application server.
The Intelligent Network (IN) enables both fixed and mobile operators to differentiate their proposition by providing value-added services in addition to the standard voice and messaging telecom services such as PSTN, ISDN and GSM services on mobile phones, e.g. 800 numbers, prepaid, etc. In IN, the ‘intelligence’ is provided by a server called the SCP (Service Control Points). There are other nodes, which provide triggers and databases, but I’ll keep it focused on the core function. IN is based on the Signaling System #7 (SS7) protocol between telephony switch and other network nodes in the network operator. The main driver of IN was previously all new features and services were implemented directly in the switch. This made creating VAS (Value Added Services) slow and expensive. So IN enabled a degree of separation, however, as we can see by the woefully slow rate of telephony innovation compared to the web, it only partially achieved this objective. Today, IN is ham-stung by proprietary vendor implementations, which continue to ‘lock-in’ operators to specific suppliers and limit service innovation.
An application server hosts an API (Application Program Interface) to expose business logic and business processes for use by applications. The term can refer to: services that are made available by the server, or the software framework used to host the services such as WebLogic, WebSphere, JBoss, and GlassFish to name just a few. The main benefits of an AS are in lowering the Total Cost of Ownership (TCO) by:
- Centralizing the business logic making upgrades, config changes, etc. easier;
- Easier security as data and application logic must pass through this point;
- Improving performance of applications in heavy use environments; and
- Making application/service creation easy: the server does the hard programming, so developers can focus on business logic.
The AS can be thought of as the enterprise/web service provider’s SCP. Given the rate of innovation demonstrated by the web service providers compared to telco; and the need for operators to adopt systems that enable it to take advantage of the ecosystems being generated on the web; moving to an AS from their proprietary SCP would functionally make sense. Then there’s always the statement that telco is special/unique, it requires low latency, high availability, high throughput; which means enterprise systems are not addressable. This was true perhaps 5 years ago; however, today telcos’ requirements are no longer the ‘gold-standard;’ financial services, military applications, and healthcare have more stringent requirements in availability (zero downtime), latency (sub millisecond), and scalability. But adopting a proprietary AS still risks some of the lock-in risks of the legacy IN.
A driver behind open source is the software development costs in organizations are between 10-15% of total costs. It’s the business model that drives the cost and value. When IBM was evaluating Linux, they estimated it took a development investment of about $400m to create a competitive OS; they estimated that Linux would receive about $800m in development investment through being open source! Double what IBM could afford! Socially it enables a more effective use of our development resources. So an open source AS, focused on telco enables a larger development community to innovate on both the platform and applications.
Open source in the enterprise is now mainstream:
“By 2010, Global IT organizations will use open-source products in 80% of infrastructure-focused software investments and 25% of business software investments.” Source Gartner.
“58% of IT execs reported that they now use Open Source for mission critical applications, 79% now use open source in application infrastructure, 62% view open source software as capable of delivering significant business payback, and 80% viewed factors other than cost such as open standards support, use of code, and avoiding lock-in.” Source Forrester/Unisys.
The question on stability has long since passed; in many respects open source now provides a more stable platform, given the broader and more diverse development investment. We’re even seeing some large operators declare open source as their preferred software model. As in the enterprise, open source is going to happen for an operator’s IN as it lowers the total costs of ownership by up to 90%. M1 in Singapore is leading the way with its adoption of an open source IN from hSeind. The web service providers have clearly demonstrated service innovation on an AS is much faster than on a closed, proprietary IN. The current economic environment may prompt more rapid adoption by operators as they focus on operational costs; and operators have already successfully adopted open source for their BOSS (Business and Operational Support System).
So if an operator is looking at an NGIN (Next Generation IN), don’t continue in the current proprietary rut: consider an open source AS and finally compete with the web service providers.