Mobile networks are growing more expansive and complex. With infinite 5G opportunities on the horizon, communication service providers (CSPs) can gain an edge through AI-powered, data-driven network optimization technologies - maximizing both the network's performance and potential, while intuitively delivering on strategic objectives amidst growing scale and complexity.
What is network optimization?
Network optimization is achieved by applying industry-leading, AI-powered technologies throughout the network lifecycle to align network performance with strategic objectives and maximize return on investment.
Leveraging live and predictive network data, network optimization technologies push the network to its maximum potential, proactively resolving performance issues ahead of subscriber impact. Through active monitoring and predictive forecasting, network planning technologies evaluate future network demands and identify where and when to expand capacity for maximum returns, months in advance.
This results in an always-on, high-performing network – tailored to strategic business objectives and ready to meet increasingly critical performance demands of future 5G use cases.
Four steps of network optimization
To capture the value of the 5G era, CSPs must be ready to adapt to the constantly changing demands and expectations of emerging 5G use cases, for consumers & enterprises alike.
Network benchmark criteria are growing more expansive and dynamic. CSPs must identify which key metrics to focus on and how to evaluate them proactively and effectively.
AI-powered technologies identify gaps where key metrics fall short, drill-down to the root cause and proactively resolve the issue across the full end-to-end network.
With traffic forecasting and performance prediction, network optimization becomes integral to network investment planning – identifying critical congestion areas that cannot be solved with optimization.
Benefits of network optimization
Network optimization is effective in boosting network investment returns, while ensuring superior network experience and easier network operations, without the need for capacity expansion. As validated in our engagements with frontrunners like Swisscom, TDC and other leading CSPs, network optimization is proven to produce many direct and indirect benefits on both network and business performance.
- 31% reduction of cells with poor spectral efficiency
- 15% increase in downlink user throughput during busy hours
- 10% reduction in unplanned carrier expansions
Improve user experience
15% user experience improvements
20% increase in deployment capacity
70% reduction in customer complaints
90% external benchmark wins
- 4x larger network managed with the same team
- 50% faster site acceptance
- 3.4% decrease in site power consumption
5G and network optimization
5G is rewriting the playbook for mobile networks – bringing more applications, devices, spectrum, energy and expenses into network operations. This is generating thousands of new consumer, enterprise, and mission-critical use cases, all with complex requirements and placing increased pressure on network performance.
In this new reality of constant change and complex service level agreements (SLAs), a properly designed network which intuitively aligns with strategic objectives while maintaining a low total cost of ownership becomes more important than ever, serving as the main foundation not only to meet increasingly challenging use case demands and deliver on SLAs, but also ensure long-term business success.
AI, data and network optimization
Just as AI and data are key to winning with 5G, multi-dimensional data sources are key to winning with network optimization. If you can’t measure it, you can’t improve it – which is why network optimization technologies rely on a varied range of data sources including data from the network itself, but also third-party data such as benchmark campaigns or crowdsourced data. Together, this ensures a full end-to-end perspective of the network’s performance against the assigned strategic objectives, which gets harder every day as the better the network performance, the more difficult it can be to achieve performance gains.
How to measure network optimization
Network optimization metrics are divided into two main areas: network-centric resource KPIs (R-KPI) and use-case-centric service KPIs (S-KPI).
Network-centric resource KPIs
R-KPIs are the primary metric of most traditional forms of network optimization. With this approach, KPIs are typically specified per each network domain (such as radio networks, transport networks, core networks, and IP multimedia subsystems (IMS)). They are measured based on the performance of individual network elements such as network accessibility (whether a call can be established), retainability (whether calls are dropped) and integrity/quality (download speed).
Use-case-centric service KPIs
S-KPIs frame end-to-end network performance in ways suitable to deliver on both consumer expectations such as video resolution, video stalls, web access time, VR gaming latency, and over-the-top (OTT) voice mean opinion score, as well as enterprise requirements such as latency, jitter, and availability. A critical requirement for this approach is end-user specific data such as net promoter score, complaints, crowdsourced data, call trace records. It is also critical that this data is proactively monitored and acted upon, for example data from customer call centers can be correlated with network data to derive customer perception of network performance. However, access is increasingly difficult and expensive, as data is spread across various systems, and is sometimes limited by encryption or privacy regulations (e.g., GDPR, iOS).
Network optimization solutions
Network optimization is achieved through a combination of cognitive software technologies which are uniquely developed and applied by software engineers and data scientists across the full end-to-end network and the various lifecycle stages of planning, design, tuning and continuous optimization.
Ericsson currently deploys its leading network optimization solutions across nine million live network cells globally where we continue to propel the world’s best networks, including many frontrunner commercial 5G networks, to their optimal potential.
Ericsson’s cognitive software solutions are fully automated, proactive and precision engineered for high-performance network operations based on live- and historical network data. In addition to generating superior network performance, this is also proven to lower our customers operational costs and network footprint, as well as offering the highest possible return on network investments.
Our AI-powered solutions result in smarter capital expenditure (capex) through improved prediction accuracy, greater return on investment and superior user experience.
We enable our customers to build their networks with greater precision, making them ready to deliver new use cases for both consumers and enterprises, at the right cost.
We deliver a faster time to market for our customers with AI-powered network tuning that ensures faster validation of quality for a successful commercial launch.
We empower our customers to continually maximize performance, efficiency and profit gains across their full end-to-end network chain – delivering benchmark network experiences at the lowest possible TCO.