NFV and OSS: Virtualization meets reality

Changing telcos’ systems from a legacy to a virtualised model is a bit like building an autonomous car from a moving steam locomotive. In this report, we look at the relationship between NFV (Network Functions Virtualization) and OSS (Operations Support Systems), and the difficulties that operators and the developer community are facing in migrating from legacy OSS to NFV-based methods for delivering and managing communications services.

Introduction: New virtual network, same old OSS

The relationship between NFV and OSS

This report discusses the relationship between NFV (Network Functions Virtualization) and OSS (Operations Support Systems), and the difficulties that operators and the developer community are facing in migrating from legacy OSS to NFV-based methods for delivering and managing services.

OSS are essentially the software systems and applications that are used to deliver services and manage network resources and elements in legacy telecom networks – such as, to name but a few:

• Service provisioning: designing and planning a new service, and deploying it to the network elements required to deliver it
• Service fulfillment: in its broader definition, this corresponds to the ‘order-to-activation’ (O2A) process, i.e. the sequence of actions enabling a service order to be logged, resourced on the network, configured to the relevant network elements, and activated
• Service assurance: group of processes involved in monitoring network performance and service quality, and in proactively preventing or retrospectively repairing defective performance or network faults
• Inventory and network resource management: managing the physical and logical network assets and service resources; keeping track of their utilization, condition and availability to be allocated to new services or customers; and therefore, closely related to service fulfillment and assurance.

As these examples illustrate, OSS perform highly specific management functions tied to physical network equipment and components, or Physical Network Functions (PNFs). As part of the migration to NFV, many of these PNFs are now being replaced by Virtualized Network Functions (VNFs) and microservices. NFV is developing its own methods for managing VNFs, and for configuring, sequencing and resourcing them to create, deliver and manage services: so-called Management and Orchestration (MANO) frameworks.The MANO plays a critical role in delivering the expected benefits of NFV, in that it is designed to enable network functions, resources and services to be much more easily programmed, combined, modified and scaled than is possible with PNFs and with OSS that perform isolated functions or are assigned only to individual pieces of kit.The problem that operators are now confronting is that many existing OSS will need to be retained while networks are transitioning to NFV and MANO systems. This is for a number of reasons. 

• Executive Summary
• Next Steps
• Introduction: New virtual network, same old OSS
• The relationship between NFV and OSS
• Potential solutions and key ongoing problem areas
• Conclusion: OSS may ultimately be going away – but not anytime soon
• OSS-NFV interoperability: three approaches
• OSS-NFV integration method Number 1: use the existing BSS / OSS to manage both legacy and virtualized services
• OSS-NFV integration method number 2: Use a flexible combination of existing OSS for legacy infrastructure and services, and MANO systems for NFV
• OSS-NFV integration method number 3: Replace the existing OSS altogether using a new MANO system
• Three critical problem areas: service assurance, information models, and skills
• 1. Closed-loop service fulfillment and assurance
• 2. A Common Information Model (CIM)
• 3. Skills, organization and processes

• Figure 1: Classic TMN BSS / OSS framework
• Figure 2: Telcos’ BSS / OSS strategy for NFV
• Figure 3: Transition from BSS / OSS-driven to NFV-driven service management as proposed by Amdocs
• Figure 4: NFV / SDN functions as modules within the Comarch OSS architecture
• Figure 5: Closed-loop network capacity augmentation using Netscout virtual IP probes and a common data model
• Figure 6: Service-driven OSS-MANO integration according to Amdocs
• Figure 7: HPE’s model for OSS-MANO integration
• Figure 8: BSS and OSS still out of scope in OSM 1.0
• Figure 9: Subordination of OSS to the MANO system in Open-O
• Figure 10: Vodafone Ocean platform architecture
• Figure 11: Vodafone’s VPN+ PoC
• Figure 12: Operators’ main concerns regarding NFV
• Figure 13: Closed-loop service fulfillment and assurance
• Figure 14: Relationship between Information Model and Data Models