It’s no exaggeration to say that the RAN is one of the most complex network domains that service providers manage and operate. The intricate traffic patterns caused by user equipment mobility or unpredictability of the radio frequency can be problematic for RAN operations. These complexities are one of the biggest obstacles for our customers to enhance their quality of service.
When it comes to enhancing quality of experience, there are two fundamental issues associated with the RAN:
The first issue is the inability to use real-time subscriber-level trace data, which can be used to enhance customer experience for every subscriber. The challenge is that the subscriber-level traces generate massive volumes of data, as opposed to the traditional cell-level counters. Translating the data associated with each subscriber into actionable insights is unattainable by human operators in real time.
This is where AI-powered software applications can be hugely beneficial – being able to analyze the massive amount of subscriber-level data, potentially in the order of millions of events per second. With AI, our customers can gain unprecedented levels of insights into each subscriber, leading to enhanced quality of experience. However, this leads to the second issue, which is the implementation of such software applications, especially if they are introduced by newer application developers on closed and proprietary RAN.
To that end, our customers are engaged in a RAN transformation – from traditional, proprietary RAN to virtualized RAN and then to open RAN. By opening the RAN, our customers can introduce innovative applications, including AI-powered applications.
To facilitate this RAN transformation, we have recently introduced our RAN Intelligence Controller (RIC), enabling our customers to host various innovations in the form of applications – rApps and xApps. When paired with Subscriber QoS Optimization rApps, VMware RIC allows our customers to use subscriber-level trace information to detect subscriber-level anomalies, analyze the root causes of anomalies, predict the subscriber-level impact, identify the signal interference, and then optimize the use of underpinning virtualized RAN functions with closed-loop load-balancing – all in an automated manner. With our AI and automation rApps running on VMware RIC, our customers can drastically improve their customer experience for every subscriber.
Our AI-powered Subscriber QoS Optimization rApps utilize subscriber-level trace information to automate the RAN operations. The key features are:
· Anomaly detection: Monitors subscriber-level trace information, as opposed to monitoring only cell-level information, for every subscriber and utilizes AI to detect subscribers with anomalous QoS and generates alerts
· Root cause analysis: Automates the process of finding the root cause – the analysis is performed in real time and is highly accurate, compared to the analysis done manually, due to utilizing automation and AI models sifting through large amounts data
· Impact prediction: Uses a “what-if engine” with AI models to predict the subscriber-level impact based on the relationship between configuration parameters of the RAN and subscriber QoS from large amounts of historical data in diverse deployment scenarios
· Uplink/external interference localization: Uses AI models to identify the external interferer of the network and displays the likely location of the issues, eliminating manual, cumbersome, and error-prone processes
· Closed-loop load balancing: Uses AI models and automation to determine how the load balancing thresholds should be changed to resolve the load imbalance issue and thereby ensuring that subscribers receive better customer experience
Our Subscriber QoS Optimization rApps, with VMware RIC and its APIs, enable our customers to gain unprecedented real-time visibility of the quality of service that each subscriber is receiving through the subscriber-level traces. This insight empowers our customers to proactively configure and optimize their RAN to enhance the quality of service for the most-impacted subscribers. With AI models and automation, combined with the ability to predict the impact of failures, our customers gain the confidence to frequently explore better configuration settings of their RAN. Ultimately, our rApps running on VMware RIC improve customer experience for each subscriber.
It is important to note that one of the key benefits of VMware RIC, specifically of VMware Centralized RIC (a non-real-time RIC in the O-RAN Alliance reference architecture), is its ability to support traditional RAN. This capability enables our customers to utilize our Subscriber QoS Optimization rApps, running on VMware Centralized RIC, in their existing RAN deployments for both 4G and 5G networks now.
As part of VMware RIC, we also offer VMware Distributed RIC, which is a near-real-time RIC as defined by the O-RAN Alliance reference architecture.
VMware Centralized RIC and VMware Distributed RIC enable our customers to accelerate innovation in the RAN and build new apps for RAN automation, optimization, and monetization.
For more information, check out our VMware RIC eBook.
Authored by Yusuke Kanamori and Rakesh Misra