Service agility is the key strategic benefit of network virtualisation. It results in an increase in revenues from new services, accelerates the time to market and provides capex and opex savings. In addition, cost optimisation is an implicit benefit of increased service agility.

Operation support systems (OSS) are emerging as key enablers for achieving service agility, operational flexibility and optimisation of costs. To increase service agility and service fulfillment, operators are focusing on OSS functions such as order management, inventory management, activation and provisioning, and planning and optimisation. Software-defined networking (SDN) trials and use cases generally target traffic control and management, with increased emphasis on service-assurance OSS functions. Currently, network functions virtualisation (NFV) and SDN trials are being conducted by service providers to deliver existing services through traditional physical networks. The existing OSS can support network virtualisation by abstracting control of the virtual infrastructure through virtual network function (VNF) managers, NFV orchestrators (NFVOs) and SDN controllers, while preserving the current OSS processes and operations.

Service providers agree that OSS abstraction provides a default migration path for them. Initially, OSS abstraction will deliver service fulfillment and assurance control for existing services through new software control layers of virtualised network function managers, NFVOs and SDN controllers. This will be achieved in a streamlined way, with minimal disruption.

However, for maximum benefit, service providers expect vendors to develop new, mature virtualised next-generation networks (vNGN) OSS, which will orchestrate and manage physical and virtual network resources for both existing and new services, continually reduce the complexity, development and maintenance costs of service providers’ OSS; lower the time and cost of integration through open interfaces, and hardware and software interoperability standards; provide near-real-time view and control of operations with policy-controlled automation and analytics; encompass delivery and lifecycle management of services where resource management is implicit; and potentially modernise operations to converge network and IT planning, build, and operations and maintenance.

Key milestones of OSS maturity with network virtualization

Discussions regarding OSS among service providers and vendors are still at early stages. The exploration of multiple OpenStack projects is ongoing, but OpenStack’s management of virtualised infrastructure is seen as the most applicable at present. This exemplifies the immaturity of network virtualisation technologies.

Over the next three years, service providers and vendors will continue to identify network functions that can and should be virtualised in order to realise business benefits. Service providers have recognised that greater cost reductions could be achieved in the access network than in the core network. In the following three years, service providers anticipate that vendor OSS will become advanced enough to allow the coexistence of physical and virtualised networks through OSS abstraction. Over the subsequent five years, service providers expect to continue the gradual progress of the development, implementation and rationalisation of their OSS, in preparation for transforming into a consolidated, slimmer vNGN-OSS architecture that addresses OSS challenges and gaps to orchestrate the management of vNGNs. As a result, towards the end of the next decade, service providers are likely to complete their migration to vNGN-OSS, which will manage the vNGNs and technologies that have emerged during the decade. However, over a long-term period, service providers would consider other possibilities, including investing based on customer engagements, so that customers part-fund the vNGN-OSS development for particular services, progressing when sufficient VNFs are available for end-to-end delivery of one or more services,  overlapping self-organising networks automation with vNGN-OSS requirements, developing a holistic service agility vNGN-OSS framework and requirements for moving towards automation. They would also include these requirements in all OSS procurement documents issued henceforth.

Network virtualisation facilitates OSS automation, which in turn will drive service agility

Service providers are far more convinced of the benefits that network virtualisation can bring in terms of service agility and operational flexibility, than any cost savings that may be achieved. OSS automation is a cornerstone for service providers to increase service agility. It requires new OSS functions and features that can do auto-additions of new network elements, functions and technologies, and can deliver services using a combination of network and service policies. These OSS functions and features would effectively optimise end-to-end plan-to-provision OSS processes of service providers, consequently increasing their service agility with an OSS that has automated services readiness.

OSS automation does exist today, but it is not as plug-and-play as in IT environments because of the lack of standardisation. However, service providers view OSS automation as crucial to increasing their service agility and allowing them to differentiate themselves from other service providers as well as IT-centric competitors such as over-the-top and data-centre providers. For true OSS automation to be realised, the following vNGN-OSS are needed, in addition to open interfaces and more standardisation:

•             A minimum number of parameters provided by VNFs to the OSS or orchestrator for seamless auto-discovery and configuration

•             A master policy database in the OSS layer that can configure the VNFs based on pre-determined and approved network and service designs and configurations

•             New simulators and emulators of OSS features to pre-test the results of implementing the changes

•             Improved security administration, hierarchy and auditing functions

•             Off-the-shelf, plug-and-play infrastructure for new vNGN service delivery and management

•             A single end-to-end view and operation of service delivery and lifecycle management.

OSS important to realise benefits from network virtualisation

Service providers have expressed uncertainty about the benefits of network virtualisation in terms of total cost of ownership and, in particular, whether the expected 33 per cent capex savings would be lost in the longer term due to the greater opex of virtualisation software. Service providers believe that, potentially, the cumulative cost of a traditional physical network hardware and software with lower operations and maintenance (O&M) costs could be less than the total cost of ownership of a virtualised network, which offers an initial capex saving, but incurs higher O&M costs. So far, very little in-depth analysis has been conducted by the industry regarding the costs and benefits of network virtualisation. Analysis shows that retaining the existing physical network and systems while also investing in “silo” network virtualisation could lead to overspending. In contrast, using a holistic service agility approach to migrating to a vNGN can help realise the maximum benefits of network virtualisation.

Service agility is the linchpin for maximising the benefits of network virtualisation

There are three phases to become an agile service provider with a vNGN-OSS:

• Deployment of vNGN: vNGN investments should be used primarily to augment or replace the network infrastructure that is delivering existing services. The vNGN should not be implemented in a silo environment, which would lead to dual spending on two networks and operations.
• Co-existence with vNGN: In this phase, the benefits gradually begin to match the costs. These will largely come from deferred and reduced hardware costs, and the application of existing OSS to provision, manage and assure physical and virtual network resources for existing and new services using OSS abstraction.
• Transformation to vNGN and vNGN-OSS: Service providers clearly identify legacy systems and infrastructure, and either replace or retire them in favour of newer, lower-cost virtualised alternatives which are more readily integrated into the new vNGN and vNGN-OSS architecture. The faster this transformation is completed, the sharper the inflection point can be in the service agility trend.

Historical trends indicate that if service providers continue their “as-is” operations, costs will increase gradually but continually over the next 10 years. As a result, a holistic service agility framework for migrating to vNGNs with vNGN-OSS could curb the climbing costs for service providers, while increasing competitiveness.

Conclusion

Going forward, a holistic service agility framework is needed to move towards vNGNs, a framework that increases the agility of service delivery and lifecycle management, and uses increased OSS automation, which can provide near-real-time views and control of operations with policy-controlled automation and analytics.

New vNGN-OSS must be cheaper and more agile – matching the flexibility and elasticity of virtualised networks, while still capable of managing traditional networks. vNGN-OSS will also need to orchestrate and manage physical and virtual network resources for both existing and new services.

Based on a white paper “Next-generation OSS is critical to delivering service agility in new virtualised networks” by Analysys Mason