The telecommunication industry is preparing for the next technology upgrade cycle which will bring with it changes in lifecycle management and a focus on maximum capital efficiency. 5G will disrupt the RAN, core and edge network enabling 100X increase in bandwidth, ultra reliable low latency, and explosive growth in M2M connectivity. How can CSPs prepare for this disruption in their planning and operations to support industrial automation, virtual reality applications, and enhanced mobile broadband services?

Learning from 4G

It was a decade ago that 4G was launched and the promise of increased returns on capital investments did not materialize against business plans. In fact, in Europe and CALA, revenues declined. Today we are on the edge of another technology upgrade cycle that presents a similar difficult business case. 5G will require massive CAPEX investments in spectrum, radio upgrades, and fiber backhaul. These investments will be needed to support the bandwidth and latency requirements of new services that include autonomous vehicles, virtual reality applications, and gigabyte mobile broadband across a wide coverage area.

Preparing for 5G

Network cost for operators will increase between 50% and 300% in deploying 5G depending on geography and network technology deployed. How do CSPs prepare themselves for capital intensive 5G expenditures in the face of declining smartphone shipments? CSPs must consider all options in making the transition from their existing mobile networks to 5G. First refarming spectrum from 2G/3G to 4G networks is an obvious choice to better utilize available spectrum. Second, network sharing can reduce deployment cost by 30 to 50%. This will be necessary for many operators serving dense urban cities. Third, automating manual labor-intensive task in the planning, optimization, troubleshooting, and resolution of service impacting events can dramatically reduce operational expenses. The closed loop management area is ripe for innovation because up until now very little has been done in the automation and analytics field within network operations.

5G Disruption

5G will bring major changes to the RAN, edge, and core network infrastructure. In the RAN, small cell antenna shipments globally will increase from 500,000 in 2018 to 8.5 Million in 2025. As traffic moves out to the edge, we will see increases in fiberization. Fiber will be deployed to the antenna and to the site. 5G will impact the core network as more network functions are moved from hardware into software. Appledore Research forecast that $100 Billion of the total global network infrastructure will be virtualized in 2025. The ratio of CAPEX to revenues has remained constant over two decades, but at the same time traffic on both fixed and broadband networks will increase threefold over the next five years. The Telco Cloud will replace legacy technologies with cheaper, flexible, cloud native technologies, allowing CSPs to continue to support more and more traffic with the same level of capital investment. Managing a 5G network will introduce challenges that most experts in the NOC have not seen with prior mobile technologies. Network traffic and software as a service packaged offers will continue to increase. Combining smart telemetry collection, analytics, and closed loop automation will enable CSPs to manage the coming 5G technology upgrade cycle and support the new revenue generating services.

Elisa Automation Use Case

As we look to the market adoption rates for automation, we have concluded that very few CSPs have begun this journey. One exception is Finland based Elisa that started with SON and now is far along in its journey to automate as much of its NOC operational activities. Elisa is a leading CSP applying automation in their network operation center (NOC). Elisa has not increased CAPEX more than 12% of their revenue and OPEX continues to decline. Elisa pioneered unlimited data subscription in Finland, but to achieve profitability they required flat or declining CAPEX/OPEX. 90% of network incidents are resolved automatically and as much as 75% are proactively corrected before they become visible to the customers. They have implemented closed loop automation in their network implementing over 4 million network checks and over 4000 change processes. They have achieved a massive cultural change to one where they have no people in their NOC.


CSPs should be preparing now with their key partners and suppliers to assure QoS for data flows and network slicing across the 5G network. In the early deployment cycle these data sessions will transit multiple technology domains so backward interoperability will be a requirement. KPIs should be mapped to specific service and applications which can be rapidly collected and analyzed in real time. End to end sessions will flow over physical and virtual networks in the core and edge network. The ongoing analysis of active and simulated sessions can be modeled in a system we refer to as Rapid Automated Service Assurance (RASA). This system can trigger and provide the control loop to orchestration and policy management systems using a trigger or “notification event” to add capacity, change route paths in the network, and reallocate network resources which are no longer required.
CSPs who innovate in automating network management functions will become disruptors in their market. Automation in the planning and optimization cycle will yield higher network utilization in the radio and core network that will defer CAPEX in network and spectrum deployments. Innovation in the service impact and assurance of key performance indicators will improve customer experience. Appledore Research has published in-depth research, use cases, best practices, and solution profiles that reveal how and where CSPs can benefit from AI and analytics and closed loop automation in 5G, edge computing, IoT, and SD-WAN.

By Patrick Kelly, Founder and Principal Analyst Appledore Research