Infovista has announced Precision Drive Testing to bring an ML/AI data-driven approach and automation to network testing, significantly reducing the cost and time of 5G network testing. The patent-pending Precision Drive Testing leverages 5G network, service and customer data, and ML/AI techniques to increase the speed and accuracy of 5G testing process.
“There’s a misconception that 5G makes drive testing redundant, but the reality is that although enhanced from LTE, the ‘minimisation of drive testing’ feature in 5G still only works if users are in the geographic area that needs testing. Drive testing will still be needed to complement the gaps of MDT in testing, at least until 5G is fully autonomous,” said Dr Irina Cotanis, Technology Director, Network Testing at Infovista. “But such is the complexity of 5G networks and the proliferation of device types that traditional drive testing processes are not fit-for-purpose. Operators can’t afford for their highly qualified RF engineers to be driving around manually testing; it’s time to automate and make the cloud do the heavy lifting. Precision Drive Testing transforms the drive testing process from being engineering-driven to AI/ML data-driven, from manual to autonomous and from something very few can do to something that can be done by anyone.”
Infovista’s cloud-based Precision Drive Testing solution automates and guides testing triggered by use cases, based on information and/or analytics results from network planning and performance, fault and configuration management, services assurance, and customers’ (or Crowdsource) data, which an operator will use to inform and improve the network lifecycle processes. Results are instantly available as actionable insights to be used by the requesting systems/solutions, creating a closed-loop of automated and guided testing. This reduces testing time and effort and improves the accuracy of service-specific testing KPIs.
When triggered, the Precision Drive Testing use cases automatically calculate the best test route, generating test scripts to run along with their Definition of Done criteria, while the correspondent KPIs log masks and context-sensitive criteria. The automatic calculation of sweet spots and drive routes, coupled with directions to the tester, including error handling and Edge Analytics, improve the speed and accuracy of the testing processes. Finally, real-time reporting and data feed of the expected result ensures that testing data is quickly acted upon, and in turn, used to inform the network/service/customer data source.
Use cases that can trigger Precision Drive Testing across the 5G network lifecycle include, but are not limited to:
Automated Single Site Verification (SSV) – Powered by Precision Drive Testing, the Infovista Automated SSV solution directs lay people using commercially available smartphones to conduct tests in the field, leaving valuable engineering expertise free to focus on high-value network optimisation tasks. Site testing routines and criteria are defined at the backend and key measurements are autonomously evaluated against pre-defined thresholds rather than manually by a field engineer. This reduces errors in data collection and accelerates the time to Definition of Done status in network-wide testing processes
Fine-tuning 5G planning propagation models – The solution identifies when the 5G propagation model is likely to be less accurate due to varying external factors, for example, in areas of known building construction, fast-growing trees, or in extreme and challenging locations such as quarries and container ports where the landscape changes frequently. Precision Drive Testing enables the user to get actual live data that can then improve the model and accelerate the planning of public and private 5G networks
Network Acceptance Automation – ‘Site acceptance’ routines in multiple conditions have a challenge using real traffic with enough samples, particularly in greenfield networks. This means if a problem occurs when real traffic starts running, it can be a time-consuming and costly process to identify the underlying problem. In triggering Precision Drive Test, the algorithms identify the sweet spot for testing the live network, define the testing routine, direct the tester to the site and automatically conduct the tests. This reduces the time and cost of 5G network acceptance through automated and highly specific testing processes
Geolocated troubleshooting – Deeper and more advanced assurance can detect problems and prompt Root Cause Analysis for troubleshooting. But if the problem still exists, it can then trigger a Precision Drive Test to get precise fine-grained geolocated data to help solve issues such as ‘Inter-vendor Ping Pong Handover’ that have been detected by the assurance system
Proactive network health monitoring – Monitoring a network to ensure delivery of the end-user experience, for example for a network’s top 10 VIPs, can be a labour intensive and unforgiving process. The software can identify and define drive testing routes, such as a VIP’s commute, and trigger Precision Drive Testing to test degradations/problems on the routes and proactively automate regular quarterly drive tests.
Assurance of Critical Networks – The growth in private networks, and in the future 5G network slices, will see more widespread deployment of uRLLC-based services such as remote control of mining equipment or drone surveillance of utilities’ critical national infrastructure. These critical networks require predictive SLA management. Assurance software can trigger a Precision Drive Test to proactively test critical networks, with assurance processes in turn improved and optimised through the input of live network test data.