The Evolution of Bandwidth on Demand
Jamie Jefferies, Vice President and General Manager of EMEA, Ciena
Thursday, August 29, 2019
As a result of an increase in the use of multiple cloud providers and over-the-top services (particularly video), increased mobility of end users, flexible placement…

The Evolution of Bandwidth on Demand
Enterprises, content providers, data centre operators, mobile operators, and retail network service providers all need to have the appropriate capacity between designated locations at the right time—consuming network resources with the same ease and cost-effectiveness as cloud-based X-as-a-Service offerings.
But as a result of an increase in the use of multiple cloud providers and over-the-top services (particularly video), increased mobility of end users, flexible placement of visualised workloads, and transient relationships among an open ecosystem of digital businesses, anticipating workload peaks accurately is a near impossible challenge. Traffic patterns are becoming increasingly more unpredictable across networks and traffic spikes are becoming more pronounced. If the spikes exceed available bandwidth, applications will falter or fail.
Manual processes for connectivity
To order a connectivity service today, requests must be entered into an order management system which initiates a manual workflow. This fulfilment process can take weeks to complete even for already on-net locations. The requirement for guaranteed bandwidth forces customers to over-order.
They either purchase fixed capacity services sized for the peak demand forecasted between their locations, or inefficiently route some traffic through intermediate sites to minimise the number of transport services due to the high cost of an underutilised full mesh. They make purchase decisions largely based on price, commoditising connectivity services.
Making a new service available, or adjusting the bandwidth on an existing service, involves manual handoffs between departments and monotonous, potentially error-prone lifecycle operations. Network engineers must manually design the path using offline inventory systems or spreadsheets. Plus, provisioning requires administrators to manually enter data into multiple Network Management Systems (NMSs) or domain controllers to configure network elements within respective domains.
These operational silos and manual processes can result in delayed revenue and higher operational costs. To reduce costs, the product catalogue is often limited to fixed duration contract offerings, limiting the addressable market and constraining revenue opportunities. A service provider faces either lower margins or non-competitive prices which hampers competitiveness.
Automate using Bandwidth on Demand
Help is at hand though, Bandwidth on Demand solutions are now available for automating end-to-end path computation and provisioning connections across multi-vendor, multi-domain infrastructure. Using an order management system (OMS) self-service portal, a customer can select the desired connectivity product, bandwidth, activation and termination day and time, the A and Z locations, and quality of service (QoS) parameters.
Customers select different bandwidth options and the OMS calculates then displays the price of the service. If the price works for the customer, an on-demand order is placed, and the OMS uses a REST API to trigger the BoD solution to turn up the service or schedule it for future activation. For guaranteed scheduled services, the solution requests the domain controllers to calculate a path and reserve resources. For an immediate activation or when the activation date/time arrives for a scheduled service, the domain controllers are triggered to provision their respective network segments (or activate the pre-reserved path).
The solution delivers self-serve options to dynamically adjust bandwidth on demand. For the service provider, it simplifies the many steps involved in the multi-domain connectivity fulfilment process, reduces manual effort and total OPEX, and decreases the time interval from quote to cash.
By using a BoD solution, customers can consume connectivity resources as quickly and simply as cloud-based XaaS offerings. Service providers can intelligently automate operations and end-to-end service fulfilment across multi-vendor domains, increasing service agility and delivering services in minutes, instead of weeks or months. This enables them to offer new, competitive services, increase monetisation of their current assets and gain increased revenue streams. While unified control and management of the network helps ensure error-free service activation, eliminating unnecessary overhead and costs associated with mistakes, while increasing end-customer satisfaction.
