Skip to main content

AI for network operations

Augment human intelligence to improve decision making & increase efficiency

One of the most important concerns for many Chief Technology Officers (CTOs) and Chief Financial Officers (CFOs) is finding the most effective ways to build network capacity in line with projected growth in demand.

Network complexity will explode in the next five years as the roll out of 5G accelerates and many CSPs refresh their LTE networks. Networks are expected to grow by 73% in the next five years, more than five times the rate seen in the last five years. Yet budgets remain flat and CTOs need to do more with less. AI can maximize existing assets and help plan CAPEX in the right places.  

A future-proof network evolution plan must be innovative and reliable enough to support new services, business models and use cases as they arise. However, before new CAPEX investments are made, it’s important to maximize the assets on the ground and understand exactly the best areas to make investments that add the most capability to launch new 5G services and win the highest returns. 

On this page

How can telcos increase network efficiency

 
Radio Frequency (RF) optimization is the most important way to guide CSP investments in coverage and quality to deliver the highest quality of experience (QoE) for customers. To guide their investment decisions, CTOs need a more detailed view of network performance than is possible with conventional drive testing and geolocation tools.

Spectrum is one of a CSP’s most valuable resources, as well as one of its most expensive. Making use of finite spectrum is paramount and network investments that focus on the efficient use of spectrum are the most likely to achieve the biggest returns. Conventionally, CSPs have relied on analyzing cells with low average capacity performance indicators at peak traffic times. A more effective approach is the Nokia Spectral Performance Analysis. Instead of simply analyzing individual cells, the service divides cells into multiple zones with similar characteristics, measures the spectral efficiency for each zone and automatically recommends ways to improve data capacity. AI makes the analysis more granular, so CAPEX investments can be more precise and bring a higher ROI. 

case study

Spectral Performance Analysis for better customer experience

How to improve Quality of Service and CAPEX

 
Nokia collects, stores and analyzes data from multiple sources, including Minimization of Drive Test (MDT) data. MDT allows performance data to be collected from Nokia and other vendor networks, tapping into billions of anonymized measurement reports sent by ordinary mobile phones.

Optimize RF coverage by identifying coverage gaps and interference, enabling CSPs to prioritize improvement spending.

80%

lower cost than manual drive testing

100%

CAPEX savings against legacy geolocation tools

80%

faster network optimization
 

Boost spectral efficiency by using big data from devices and the network to provide detailed spectral efficiency improvement recommendations. Hutchinson 3 Indonesia wanted to better understand how its increasingly complex network was performing and make improvements to create a better experience for its subscribers. With its AI-driven solution Nokia delivered the following benefits:

17%

higher spectral efficiency
 

60%

faster optimization through automation

How COOs can reduce OPEX by boosting productivity?

 
The explosion in network complexity means that advanced data and AI solutions are essential for reducing Operational Expenditure (OPEX), a key performance aim of many CSP Chief Operations Officer (COO) executives.

Network and service operations must be managed in real time to ensure reliable network performance and customer experiences. Yet most network performance issues and alerts that degrade services are identified only after customers have been affected. Operational efficiency can be improved applying data-led and AI-based solutions that quickly identify and even predict network issues before they arise.
 

What is the best way to reduce operating costs and improve efficiency with automation?

 
With thousands of alarms being generated every day, network operations engineers are under intense pressure to handle a growing number of trouble tickets. Filtering the alarms using root cause analysis relies on specialists’ experience, but this is slow and error prone.  

Tickets are dispatched repeatedly, adding to the burden, and low priority alarms and the accuracy of tickets are difficult to establish, making it hard to locate and fix faults across the network’s layers and different vendors’ equipment.

AI is a powerful tool that augments expert human operators and replaces rules-based approaches. This helps operations teams to optimize field resource allocation to prioritize high-value sites and plan schedules more efficiently. Root cause analysis and the automatic recommendation of actions give the teams further insight into how best to maximize network availability and boost service quality for subscribers.

Nokia AVA solutions to help COOs reduce OPEX

 
Nokia AVA AI solutions run diagnostic and predictive analysis on a range of data to predict future performance for critical network KPIs. This enables CSPs to predict network degradations in advance, dealing with issues before they occur to improve service quality and reduce customer complaints.

Cell Site degradation prediction enables CSPs to detect network incidents in advance and perform corrective actions to deliver uninterrupted network services.

The Nokia AVA Cell site degradation prediction use case helps engineers predict the 20 sites with KPIs most likely degrade within next seven days, allowing them to be rectified to avoid network degradation, improving service quality and reducing customer complaints.

The AI prioritizes incidents on high value sites, enabling operation teams to focus where it matters most to the customer experience.

KPI degradation detection quickly identifies issues that degrade network quality yet often do not generate an alarm, for example sleeping cells. Applying machine learning and automation accelerates detection and resolution of issues.

95%

near real-time detection of KPI degradation

100%

success in detecting silent eNodeBs

82%

reduction in mean time to resolution (MTTR)

Tier 1 Japanese operator automatically detects network issues with 100% accuracy and in near real time

Challenge

  • Silent eNB (or Sleeping Cell) is a well known problem that affects network equipment from all vendors.
  • The base station is unable to carry traffic resulting in a service degradation but appears to be functioning normally and no alarm is triggered!
  • Identifying the issue and detecting the root cause was a very manual, time consuming process for the customer: taking up to 24 hours after the event and only yielding an 80% success rate

Ready to talk?

Please complete the form below.