What role does Artificial Intelligence (AI) play in optimising service providers’ Radio Access Networks (RAN)?
Radio access networks are growing complex. We see this complexity growth, not only with the growing 5G uptake, but also in legacy networks with increased mobile broadband traffic. AI promises that increased network complexity doesn’t mean network optimisation will be complicated. AI provides optimised use of existing RAN assets and above all, improved end-user experience.
We see traction for our AI-based optimisation solutions across the globe. Our customers have experienced this firsthand with our AI-based optimisation solution.
In mid-2019, with the Japanese mobile network operator KDDI, our AI-based Cognitive Optimisation solution, which employs cutting-edge patented machine learning algorithms to classify network problems and recommend actions, was put in test. The result was an impressive improvement in mobile broadband network speed of almost 10%, while drastically cutting the number of engineering hours from thousands of hours to just a handful.
Just recently, we have been selected by the Japanese mobile network operator NTT DOCOMO to deploy our AI-based optimisation solution into their network operation, after extensive testing in 12 cities across Japan and meticulous vendor benchmarking by NTT DOCOMO. Worth to mention that NTT DOCOMO network RAN is not provided by Ericsson; proving the RAN vendor-agnostic strength of Ericsson’s AI-based optimisation solution and the capability of Ericsson Operations Engine.
How can we use AI to Minimise energy consumption in network operations?
AI creates energy efficiencies from two angels of radio network operations, namely: site energy consumption saving and smart site dispatching. Both enabled by Ericsson Operations Engine’s data-driven approach to Managed Services.
Network densification and 5G introduction put more energy demands on the site infrastructure and, theoretically, lead to more power consumption. AI helps break the energy curve here. For operators who have deployed Ericsson Energy Infrastructure Operations solution, built on AI and advanced data analytics, have seen a 15% reduction of energy related OPEX. And this is just the start!
Not only that; AI helped those operators to intelligently manage site visits and reduce unnecessary site dispatches. In network operations, this has translated to an approximate 15% reduction in site visits related to passive infrastructure, and an approximate 30% reduction in energy-related outages.
AI use has proven an energy OPEX reduction value to operators and a sustainability enabler to society with less CO2 emission. Everyone wins!
How does AI-powered network operations drive efficiency from inside out?
Network data patterns intrinsically carry valuable information. Using advanced analytics, the Ericsson Operations Engine unlocks the real data value. It helps operator predict and manage network problems before they become user issues.
For example, the Ericsson Operations Engine uses AI and Natural Language Processing (NLP) to speed up response time, alleviating operator’s workforce from recurring, simple tasks. In live networks where Ericsson deployed such Cognitive Service Desk, the ticket triage was reduced to as little as 40 seconds and service desk costs reduced by up to 30% from day one.
Using intelligent optimisation, the Ericsson Operations Engine maximises the utilisation of network assets from site to spectrum while minimising total cost of ownership (TCO) for operators. One of our customers maintained and improved KPIs despite an average spectrum reduction of 32% in 900 MHz and 46% in 1800 MHz; making efficient use of a scarce resource as precious as spectrum.
The lock on efficiency can certainly removed by Ericsson Operations Engine.
Could you please elaborate on the importance of managed services solution in ensuring smart, seamless security?
More devices and greater network complexity increase the risk for security-related events. The risk exposure surface gets wider to the level that the support of intelligent machines is mandatory. This is where Ericsson Seamless Security Managed Services solution provides a solid shield to operators.
The Ericsson Operations Engine uses AI and automation to detect threats proactively creating a seamless security shield for networks. The Ericsson Operations Engine can automate all Level 1 alerts, identifying up to 78 per cent of false positives and clearing Level 2 alert queues in as little as 24 hours.
Clearly, having such capability is inevitable for 5G network operations.
Can you also offer managed services in remote areas with little to no access to broadband?
The answer is yes. Through Ericsson Networks solution, we enable affordable broadband to remote areas, and we also extend our global vendor-agnostic Managed Services offerings to operators as well. We recently reached an agreement with Newport Utilities to bring high-speed internet access to underserved portions of their rural Tennessee service territory that otherwise is geographically challenging or difficult to reach with fiber optic cable. The project brings high speed broadband to an underserved portion of the state.
Newport Utilities currently provides a fiber optic cable to home broadband customers and will use Ericsson FWA to augment this offering to serve areas difficult to reach with fiber. Fixed wireless access can be a more cost-effective and efficient alternative to providing broadband in areas with limited access to fixed broadband services such as DSL, cable or fiber.
The FWA solution leverages 3.5Ghz Citizens Broadband Radio Service (CBRS) spectrum and Ericsson’s industry leading 64T64R Massive MIMO technology from the Ericsson Radio System portfolio to provide better coverage and capacity. The network will be operated by Ericsson’s world-class Managed Services solution for first year of operations.
This will surely benefit residents in remote areas. What about private networks for governments and industries?
Private networks come in different flavors. A unique combination of private and public networks is called a “dual-slice” solution, and consists of a private mobile communications network for plant operations and staff, combined with the public network transmitting the same signal strength for devices that are not allowed to transmit in the private network.
In addition to the private campus network, the public network will transmit the same signal strength. This ensures a perfect connection even for terminal devices that are not allowed to transmit in the private network. For example, smartphones of customers or suppliers.
With Ericsson as network supplier, Deutsche Telekom has deployed its innovative campus network solution – a combination of private and public networks – in a new campus network site in Leipzig, the fourth one the German service provider has set up in Europe.icsson and Deutsche Telekom are also further developing the functionalities of campus solutions based on 5G Standalone (SA) technology. The 5G SA architecture supports applications that require an ultra-short response time, also known as Ultra-Reliable and Low-Latency Communication (URLLC).
How do your Managed Services solutions redefine customer experience in real time?
The Ericsson Operations Engine uses a data-driven approach to provide coverage, performance, and reliability for customers worldwide. But it’s not just any data. The Ericsson Operations Engine relies on service KPIs rather than network KPIs to help operators see what their subscribers see, so operators can better analyse, understand, and optimise their networks to deliver a superior customer experience.
With its unique AI use cases, Ericsson Operations Engine provides proactive measurements of customer experience, and it can respond before network issues affect the user. This level of proactivity is unprecedented in network operations, and through which, Ericsson Operations Engine is truly re-defining customer experience management.