While Google, an application service provider lays fibre in the streets to become a communication infrastructure operator and while Amazon has come a long way from being a bookstore to a online computer rental company practically inventing Cloud Services, I recently heard about ambitions of communication service providers to become application service providers.
Is this one more cycle for this myth? Or is there a definite possibility this time?
Let me set down some definitions lest we should end up on two different galaxies…
Communication Infra Provider (CIP) is a provider of the bearer / carrier infrastructure for communication and application services (viz. connectivity, voice, video, messaging, search engines, social media etc.). They own the infrastructure and are accountable for its upkeep although they may hire 3rd party CIOs to do this. A Public Utility provider (not unlike electricity, water, railways or gas companies), CIPs need to make the infrastructure available to a competitor too. As a result, they are highly regulated.
Communication Infra Operator (CIO) provides services to operate and maintain the bearer / carrier of a CIP. In cases where third-party agencies are not involved in the O&M of the bearer, the CIP themselves play the role of CIO.
Communication Service Provider (CSP) is one who has the customer contract to provide bearer services and, typically, a few application services as well. This is best exemplified in the case of the Mobile Virtual Network Operators (MVNOs).
Application Service Provider (ASP) is the provider of applications delivered on top of a bearer. This can include applications that can provide complete or component functionality of Communication Services too. In fact, one can say that this singular concept is probably what is a fundamental underpinning for the concept of Network Function Virtualisation (NFV).
A World Without CIPs?
As one can see from the image given above, the entire value chain has two ways to reach the customer and his wallet. These two paths have been at loggerheads for most part of the last decade and a half – CSPs being the incumbent and ASPs being the challengers. The incumbency of the CSPs also casts them in additional roles as CIP and CIO too.
When CSPs state that they would like to become ASPs, essentially they seem to imply that they would like to relinquish the role of CIPs (& CIOs).
Although they don’t seem to be contemplating it yet, I wonder if the CSPs can ever relinquish their roles as CIPs completely? And, if yes, how can a world without CIPs exist?
One possibility is what we observe in the case of In-building Solutions for public buildings or buildings that house customers of multiple CSPs. Instead of letting all the CSPs run riot by putting up their own infrastructure inside those buildings for their respective customers, it is typical to allow a neutral CIP to operate in the building. (This would also be required in the brave new world of BYOD, if it will ever happen 🙂 ).
What about interworking?
With the degree of standardisation of basic communication interfaces over the last three decades, I don’t foresee issues at that level at all. However, for an Application-layer Service to customers to succeed, the end-to-end bearer demands seamless management. It is here that Software Defined Networks (SDN and clones) come in.
The standardisation of management capability across networks without introducing excessive complexity is what makes SDN quite capable of delivering applications seamlessly across a federation of bearers provided by multiple players – big and small.
So, I think it is possible for CSP’s to spin off a federation of CIPs who provide only bearer services and become pure-play ASPs like never before – thanks to NFV and SDN.