Making a Molehill out of a Mountain

Service providers are under increasing pressure to meet demands for high-quality voice services as the popularity of Voice over Internet Protocol (VoIP) continues to grow. One industry report, for example, estimates that the global market for VoIP services will reach $102.5 billion by 2026, with a compound annual growth rate (CAGR) of 3.8 percent for the period from 2021 through 2026.

For engineers and operations professionals at service-provider (SP) organizations, the growing complexity of networks and exponentially expanding volume of visibility data being gathered presents tremendous service assurance challenges. Monitoring multiple domains, sources, protocols, dimensions, nodes, devices, and so forth has resulted in a mountain of data. Because of this avalanche of data, SPs practically need an army of skilled experts reviewing reports, monitoring dashboards, and responding to alerts.  

The sheer volume of data being collected, compounded by the complexity of the network, makes it nearly impossible for human eyes to spot the problems. Even a single-layer analytics approach doesn’t overcome this challenge. The problematic interactions may be so small within a vast amount of data that an issue can easily be overlooked. This is the proverbial quandary of needing to find a needle in a haystack. 

To address this challenge, SPs need an automated solution that can sift through the monitoring data and look at multidimensional, multiprotocol, multilayered correlations to identify patterns in order to reveal the cause of disruptions to voice services. A task this complex and extensive can’t be accomplished taking a one-call-at-a-time approach. Artificial intelligence (AI) and machine learning (ML) algorithms with built-in domain expertise are needed to generate actionable business intelligence. 

Omnis Automation: An Actionable Business Intelligence Solution 

NETSCOUT’s Omnis Voice solution, which runs on the Omnis Automation platform, takes the raw, unstructured data and separates the signal from the noise, ensuring all the necessary metrics are available to provide a clean, powerful data set—in short, NETSCOUT Smart Data. This powerful tool then applies AI/ML algorithms with built-in domain expertise that has a deep understanding of the network, including interactions between nodes, protocols, and so forth. Omnis examines vital voice metrics, such as registrations, call setups, call drops, mean opinion score (MOS), and audio gaps. This data is then run through microservice-based analytics chains to detect the most relevant outliers. Omni’s service chaining architecture makes it easy to change the algorithm’s metrics without having to change the entire chain, making this solution highly flexible and adaptable.
Omnis Voice automates mundane monitoring tasks so skilled resources can be used more efficiently. Examples of how this solution can help SPs include determining the cause of audio gaps down to the specific device type, uncovering the source of noise and low voice quality issues for 5G customers related to specific device software versions, and managing the handover from 5G to 4G. Omnis also provides multiprotocol analytics, enabling SPs to monitor the signaling layer, media layer, and access layer for 4G and 5G wireless or wireline networks and allowing engineers and operators to view how these layers interact and deliver voice services. A final example is outlier detection related to customer complaints about lengthy call setup. Omnis applies outlier analytics algorithms that can isolate latency problems to a specific area and then go deeper, applying multidimensional analytics to pinpoint more precisely what is causing the problem.

What makes Omnis Voice different is its built-in domain expertise that delivers true automation. By running specialized algorithms that understand the context and generate smart outcomes that are actionable business intelligence, SPs are better able to assure voice service performance and find that molehill in the mountain. 

Learn more about Omnis Voice.