The harsh reality of today’s communication services market is forcing service providers to re-evaluate the way new services are provisioned across their networks. The potent combination of converging networks and technologies, a more competitive landscape, and a steep rise in the number and variety of services in demand has made conventional approaches to provisioning unsustainable. Notoriously inefficient from the outset, the high cost and slow deployment speeds of traditional provisioning practices has left many service providers scrambling for revenues and market share in the face of prevailing market conditions.
Service providers today need to dramatically improve the efficiency of new service development and deployment, but must do so while simultaneously reducing overall network operating and maintenance costs. Compounding the dilemma are demands from customers for more accessible self-care systems, requiring service providers to incorporate higher levels of expensive process automation technologies into their networks.
Since de-regulation service providers have implemented a number of new COTS products to try and solve the provisioning problem, but many of these are standalone and have very little automation to ensure data accuracy or completeness. This has left many providers with data accuracy problems which has lead to provisioning processes that can be best described as “Best Effort”. In order for service providers to be successful, they must take the guesswork out of the service provisioning process; they need a more predictable approach to the way services are provisioned.
The Conventional Approach
The telecommunications industry has long struggled with the accurate and predictable provisioning of new services. The convention has been to adopt a top-down approach; that is, start with a static document that represents a view of what the network is perceived to look like and use that view as a reference point for the planning, design, and activation of every new service.
In this scenario, companies typically begin by deploying inventory management software to render a virtual view of what the network looks like, often using existing data from a number of sources, such as Excel-based records. This electronic view then becomes the basis for all network resource management planning and network and service design with the expectation that implemented processes will ensure that the network view is kept up to date. In theory, this approach works well. However, in practice, due to a number of factors, the inventory often diverges considerably from the actual network. As a result, it is not until the service-activation stage, as a service design is implemented, that the real impact of any data inaccuracies is felt, and that any discrepancies in the original data or network view manifest themselves.
For each data inconsistency that presents itself, there is delay and added cost to the provisioning process as time-intensive manual intervention is required to resolve the discrepancy. As more and more, sometimes interdependent, discrepancies come to light, this can spark a seemingly never-ending battle to keep the original network view in agreement with the network so that subsequent fall-outs are minimized.
The bottom line is that this top down approach can only ever be a best effort situation and will never be capable of delivering zero-fall-out provisioning because there is no guarantee that the data upon which everything hinges is, indeed, accurate. This antiquated approach may have worked in the older, simpler days of the telecom industry, but for today’s service providers, the problems are amplified as networks become more complex, involving multiple layers, many different vendors, and disparate fast changing technologies. In this landscape, the successful provisioning and bandwidth management of last-gen, this-gen, and next-gen services becomes even more of a challenge.
An Alternative Approach
Today’s service providers struggle with their provisioning systems because they lack a true understanding of the configuration and capabilities of their networks. Efforts to automate critical processes, such as service activation, are consistently thwarted by data-related inconsistencies.
It is true that many attempts have been made by service providers to improve the quality of their data. This is typically done via manual audits or by extracting network configuration data from network elements or EMSs, comparing this to the inventory data, and then manually resolving the discrepancies. While some service providers have been successful with this approach, it can be very labor intensive and, in many cases, hard to sustain or scale when costs need to be cut or new network elements are added due to business acquisition or changes to a provider’s target market.
A more effective, organic approach to this issue is to treat network discovery and reconciliation as the absolute foundation for service activation. This gives service providers a far better understanding of actual network capabilities and enables the realization of predictable provisioning with near zero fall-out. This approach assumes that the network itself provides the truth - rather than the documented views of the network. Therefore, creating a mirror image of the network and basing all planning, design, and subsequent implementation on that mirror image, can achieve truly “predictable provisioning”.
It can be quite challenging to discover and reconcile the network configuration, and the services that it carries, with the inventory system. However, a detailed and disciplined approach to managing the network information that is required for service provisioning can yield the true, up-to-date view of the network required for automated provisioning. Information that is brought from the network itself must displace any inaccurate data and any impacts of these changes of data clearly identified for attention. As this approach is applied, the data used for provisioning becomes accurate, allowing new services to be efficiently provisioned while any equipment or service configurations that don’t match the actual network are investigated and corrected to align with the actual network implementation. The approach eliminates the waste associated with the cumbersome manual activities required to determine the configuration of specific portions of the network to allow the implementation of a service, thereby reducing the overall cost of operating and maintaining the network. Over time the discrepancies are reduced drastically as real data replaces the incorrect data.
If service providers are to overcome the tough demands that today’s telco environment is placing on their networks and business, they must rethink the way they manage and maintain network resources. Only by improving efficiency and eliminating wasteful processes will they be able to streamline their organization and meet the more demanding requirements of both services and customers.
NETPortal Provisioning
The Multi-Domain Provisioning platform by TierOne was developed with discovery and reconciliation being fundamental to the core of the product. This combined with a tight integration between service activation and the ability to provision services across network layers in a multi-vendor, multi-technology environment, makes the TierOne solution unique in the marketplace today.
The overall design approach is to make the solution “network aware”. Many solutions on the market place today require a large implementation effort to define, model and implement the network configuration. The TierOne product is delivered with knowledge that allows it to self populate the current view of the network configuration. Each network interface (adapter) understands the network element current implemented capacity. The network configuration is then synchronized with the “Network Image”. In addition to the synchronization of physical components such as cards and ports, logical information is also synchronized. The product then stitches together circuits and services based on the logical information to form a complete picture of the network resources being used.
Services can then be provisioned directly against the “Network Image”. Because the network image is a direct representation of the current network, services will be provisioned with extreme accuracy. For those services requiring a “service design” the product can support both a manual routed design or an automatic design using the Auto-Routing capability.
Unique to the TierOne product is the ability to develop “Simple Adapters” through a graphical user interface. This allows new services or features to be turned up very quickly on existing or new network devices.
The products “Service Components” abstract multiple network interactions into a service level context. As an example, a Service Component could be for VOIP while another might be for a value add service such as Callerid. Service Components are reusable and may be invoked through the user interface, the APIs, or may be grouped together with “Service Templates”.
Service Templates take the complexity of delivering complex provisioning tasks away from either the user or an upstream system by managing the sequence in which Service Components are called. This is key as changes to the network can be abstracted so that upstream processes can remain the same.
Figure 1 - TierOne™ Multi-Domain Provisioning
Key Benefits:
- Rapid product implementation and turn-up
- Accurate provisioning through a synchronized image of the network
- Auto-Routing for network path creation
- Customizable adapters for quick service turn-up
- Cross domain and multi-vendor provisioning in a single solution
Support for services include:
- MPLS/IP (L3-VPN, VPLS, VPWS)
- DWDM, ROADM
- SONET/SDH (Next-gen support for VCAT, LCAS)
- Fibre Channel / FICON support
