Perhaps this pattern is predictable, but what happens when the vehicle connectivity comes through an enterprise contract, not through the subscriber account? How can an operator now attach behavioral profiling to the subscriber with no visibility into the enterprise connection to that subscriber? Hence, operators now need to manage profiles on both subscriber and enterprise devices, and try (through Big Data analytics) to find a connection between both, in real time, to generate a truly representative profile for planning purposes.

Access and Transparency. A final piece to the real-time puzzle facing operators in their quest for agile networks and accurate planning is the move into deeper omni-channel experiences for customers. Customers expect similar, consistent experiences when accessing content through their devices. Whether they are sitting in front of a PC or a 17-inch Tesla vertical screen, they want the same ability to view, select, interact, or purchase a given product or service. How can operators maintain the same service, and quality of service, as that customer transitions from a fixed broadband experience to a mobile LTE experience, ultimately ending on a Wi-Fi network that is offloading users in a large urban area, saving cellular spectrum? The customer does not care how it happens. He or she simply wants access and transparency. The service must be maintained, in good quality, uninterrupted, and in real-time.  

Addressing the Challenges —Where Rubber Meets the Road!

Wave Pools and Capacity Forecasting. Without over-simplifying network capacity predicting, we can illustrate capacity movement across a network by using the example of the Oceanographic Wave Pool. Network consumption has followed our lives from our homes in the mornings, through traffic and transit arteries, to our places of work, then back along similar paths during the afternoon and evening. In a wave pool, we can simulate this by initiating a linear “wave” (in this case, capacity consumption) along that path, terminating at our place of work, only to reverse that wave in the opposite direction later that day. Identification of residential clusters, transit paths, business clusters has, therefore, made the science of network capacity planning at least somewhat achievable.

Unfortunately, capacity planning has dramatically changed. To account for the changing profiles of customers, devices, and usage patterns in a complex digital world, we’ll now need to widen that wave pool (to allow for multiple wave directions), and launch waves from any point in the pool, at any time. To further complicate, let’s change the wave height (equivalent to capacity consumed) at any point along the path of any given wave as it progresses across the pool. Digital services, the devices they operate from when active, are now a constantly changing (and growing) variable. Now one begins to understand how a digitized world can take the complex science of Capacity forecasting, and make it far more difficult to manage. 

Tackling today’s challenge:  Network Agility. To serve the needs of a real-time customer, network agility must evolve and become far more efficient. Through the rapid rise in data-consuming services, and the amount of data consumed, operator revenues no longer match the increased capacity needs of the network when using previous strategies of “building the network for worst-case capacity scenarios." Instead, service and capacity support needs to be better at deploying when needed, where needed, and then redeploying elsewhere once demand has subsided or relocated. This type of operational model is virtually impossible to meet by rolling hundreds, if not thousands, of field personnel daily to perform these network changes, especially when one considers that these changes can happen in moments, not hours. To achieve agility, networks need to be “smart” while also mechanized. Essentially, the network needs to manage itself, which is realistically achievable once the network becomes software-defined, and network functions become virtualized. Once this transformation is completed, the role of network operations can fully evolve to now ensure the smart network has the tools in place to effectively conduct itself as (and when) needed.

Software-Defined, Virtualized Networks and Functions (SDN/NFV): a differentiator.  The advent of SDN/NFV has severed a long-standing link between function and platform in telecom networking. The industry has finally abstracted specialized network functions into a detached, movable software layer, while at the same time allowed that layer to run on a wide variety of commoditized hardware environments.