The BSS/OSS Landscape
Existing OSS/BSS systems are based around functional roles with the systems acting externally on the network. This has become codified with FCAPS (fault, configuration, accountancy, performance, security) and FAB (Fulfilment, Assurance, Billing) architectures such as eTOM.
This works well where there is clarity and stability about what the network is and the services it provides. However, with increasingly complex layered networks this becomes harder. Continuing with a functional split demands more integration, more movement and replication of data between systems and more layers of management software. The more this system integration grows, the more rigid these systems become and the less they can be changed or adapted for new services.
This problem is compounded by virtualization and software defined networks. In this world we are forced to ask: what exactly is the network, what is a network function? What is an IT function and what is a management function?
If we look at existing network architectures, like IMS or 4G, as virtualized software we can see that they are simply arrangements of functions with connectivity. They are no different to any other software applications with some particular latency and distribution requirements.
Similarly, OSS/BSS systems such as charging, performance management or fulfilment system are simply arrangements of functions with connectivity. Much of the focus of NFV architectures has been on the concept of service chaining, but again, this simply represents a particular configuration of functions and connections.
The Threat to Telco
If everything is functions and connections, then distinctions between network solutions and enterprise IT solutions are small. We should expect increasing competition from enterprise IT solutions in the network space and a continued attempt to bypass telecommunication applications and simply use connectivity for those solutions.
New areas like Internet of Things, which are seen as the great hope for telecommunications providers, could in reality only require basic network connectivity.
The Opportunity for Telco
On the other hand, Enterprise IT doesn't have all the answers.
When it comes to function distribution within a geographically distributed network or managing connectivity quality requirements, such as latency, today’s enterprise IT systems often avoid the issue by constraining themselves to best efforts or relying on the approximation of infinite virtual connectivity in a datacentre. Similarly issues of identity and security are often limited.
Yet many of these issues have already been addressed comprehensively by telecom software systems. The problem is that network and management functions are currently bound up in existing systems tied to very fixed business models and products that others seek to avoid. Telecom's untapped opportunity is to exploit these capabilities, by decoupling them from the unnecessary overhead.
A Proven Approach
The principles of abstraction and legacy reuse actually flourish in one particular part of telecommunications: Interconnection and Wholesale.
Whilst the exact nature of these APIs is more aligned with B2B interaction than web APIs the principle is the same: the management of a service, including its fulfilment, experience and charging, is abstracted through an API. A user of these interfaces does not see inside the supplier system but rather sees a set of APIs that provide access to those functions. Virtualized World's approach is to achieve this at a much finer scale and enable these functions to be readily assembled into products by and with partners. Moving from a course grain approach to integration to potentially a much finer grain integration with multiple parties.
Amazon’s success with AWS could be argued to be based on Amazon exposing its ‘OSS’ and creating new value.
Realizing a comparable value-creating opportunity requires adhering to a set of clear virtualization principles.