Ahead of his presentation at TM Forum Live!, in Nice, France, next month, Mounir Ladki, President and CTO of MYCOM OSI looks at how analytics can help service providers get more out of the network
Telecommunication networks are undergoing major technology shifts, and many of these are happening at the same time: Virtualization, IoT, customer-centricity and automation, bringing a new set of challenges with them.
As happens with every technology shift, the coverage, capacity and quality of the network are impacted. As an example, it took years to establish the 2G quality, which was ruptured when 3G arrived. This process was repeated when 4G was introduced, and will happen again very soon with IoT. While 5G is still a few years away, it is a good idea to prepare the networks for associated challenges.
Let us look at the most important reasons for the introduction of new technologies – speed and capacity.
While virtualization promises speed of service rollout, 5G promises both greater speed and capacity. However, until 5G becomes a reality, operators have to make use of current networks to deliver the expected capacity, driven by customer behaviour and availability of smartphones.
Rapid increases in data consumption, especially video-centric applications and OTT, have been choking the capacity of many mobile networks, forcing continuous capacity upgrades. In this context, IoT will add another layer of traffic to existing networks. The additional complexity posed by new infrastructure from IoT and 5G will only intensify the network management problem across the many co-existing network layers in the coming years.
And even with introduction of new technologies, operators will continue to maintain legacy technologies because the coverage and capacity for high-speed data, even in the most mature networks, still toggles between EDGE, 3G and LTE.
Customer-centric analytics that highlight the customer problems and prioritize troubleshooting based on customer impact must become the norm to relieve this situation. Through analytics, the operator can be guided to how elements of legacy NOCs (Network Operation Center) can be retained and pulled into a customer-centric SOC (Security Operation Center). A combination of technical and commercial metrics, behaviors, trends and predictions can be visualized by the operator’s C-level, and a new approach to problem resolution that integrates different functions, including analytics and automation, will be adopted by network engineers.
Carrier-grade business analytics with automation capabilities can be developed to solve critical capacity bottlenecks. Using such analytics (real time and non-real time), operators can trigger automatic processes for proactive or prescriptive actions to solve capacity and associated-quality issues.
As a response to the immediate need for optimal infrastructure growth, operators need to utilize analytics that add business value to decisions for planning new sites and capacity upgrades. This is based on the identification of revenue-generating locations, capabilities of handsets, customer behavior, uplink and downlink video traffic, consumption of video/conversational services, etc. This allows network investments and direct marketing campaigns based on business intelligence, for the maximum impact. However, these analytics have to be generated intelligently in real time or as long-term patterns so that revenue-impacting decisions can be accurately taken.
Operators need to introduce more dynamic ways of forecasting as opposed to a static marketing mix, which allows the planning teams to phase out the investment in line with user behavior and consumption patterns. Predictive analytics combined with dynamic forecasting techniques can provide the required insights.
Such analytics, not only help in capacity management but also guide the operator to design their resources and their performance parameters to specific user service requirements (e.g. low latency, high bandwidth, etc.).
In addition to this, analytics can also help with network monetization, i.e. operators can proactively identify congestion-free network locations (where customer traffic is low and predicted to remain so for a while), rapidly fill spare capacity with revenue-generating traffic from new service offers created in near real-time by analytics-driven, multi-team design and rollout, such as video streaming, mobile TV or smartphone apps. Such offers can be bound by time and location, and personalized for specific customer segments.
Finally, a few words on automation. Automation is no longer a ‘good to have’. In order to allow the multiple layers of modern telecom networks to run efficiently with minimal human effort, it is necessary to automate as many processes and workflows as possible. The choice of automating workflows shall be based on telecom expertise of engineers involved in capacity planning and performance management, who understand the delays and inefficiencies that creep in because of human involvement.
Through network analytics, the operator can identify processes/workflows that need a high level of automation, and those that need manual control, and ways to help the operator distinguish between the two. As a quick example, automation of operators’ radio optimization processes, some of them open-loop and others closed-loop, can dynamically optimize network capacity and release skilled teams for detailed/complex analysis.
Join me at the TM Forum Live 2016! for a session on network analytics where I will share some real-life use cases on how analytics have been used to generate value-based capacity management, and can also create opportunities for network monetization.