Employing AIOps to Establish Self-healing Networks: The Mandate for Hyper-connected Businesses in a 5G World

By: Ashok Reddy, General Manager for the Enterprise Software Division

Business success is increasingly contingent upon optimally performing, highly reliable network communications. However, for the teams responsible for tracking and managing network performance, the job only seems to get tougher. In this post, I outline why current trends are making the move to AI-enabled networks an urgent requirement.

In addition, I discuss why Broadcom is uniquely positioned to help organizations navigate this transition successfully. Today, 99.9% of internet traffic goes through Broadcom technology and 100% of public clouds use our products. This market presence puts us in an unrivaled position to help businesses make the move to the AI-enabled networks of the future.

Challenges of Modern Network Environments

Today, it is through digital channels that businesses serve customers, manage operations, track supply chains, and more. Now, it is through these channels that businesses compete. Today’s businesses are now hyper-connected, constantly reliant upon complex, multi-cloud environments connected through a mesh of networks.

Ultimately, business fortunes are resting on these networks. While ensuring network availability and performance has never been more critical, it has also never been more difficult. Network operations continue to grow in size and complexity, with traditional networks running alongside emerging mobile and fixed networks, including software-defined infrastructures, 5G, and more. Further, the volume of traffic, devices, and device types continues to expand, with innovations in IoT and virtual reality poised to introduce ever more explosive growth.

These growing, modern networks create complexity and overwhelming scale. Teams have to contend with massive volumes of operational data, leaving operators bracing for a Tsunami of metrics that have to be tracked, aggregated, analyzed, and acted upon. Ultimately, these volumes surpass any human’s ability to keep pace. Therefore, while network performance grows ever more critical, networks continue to get more susceptible to outages and performance issues than ever before.

Requirement: A New Approach to Network Operations

As they look to contend with these emerging requirements and challenges, network operations teams need to adopt a new approach. Toward that end, it will be critical to establish tight integration between network infrastructure and network monitoring software in order to gain the right context and intelligence. Teams will also need to establish unified visibility of the entire ecosystem, including new and traditional network implementations.

Perhaps most importantly, to keep up with growing volumes and more intelligently track, manage, and optimize networks, teams are realizing that they need to move away from a reliance solely on manual efforts, and start harnessing AI and machine learning.

In short, as Wall Street Journal reporter Sara Castellanosa recently wrote, companies are “finding that AI can identify and fix problems more quickly than humans.” Castellanosa’s article cites the example of Adobe. The software vendor instituted an AI-based program that automated around 30 tasks, including correcting data batching failures. The program has reduced the average time it takes to correct a failure from 30 minutes to three minutes.

With these kinds of results becoming possible, it’s no surprise that the move to AI-enabled networks is gaining momentum. The article quoted an IDC analyst as saying that AI-enabled networks will be mainstream at big companies in the next three to five years. The most critical objective? An IDC survey of 301 IT professionals found that 50% of respondents said the most important thing an AI-enabled network would do is improve application availability and performance.

How Broadcom Can Help

As network operations teams seek to establish AI-enabled networks, working with the right vendors and technologies will be crucial. This is one key reason why so many leading enterprises are turning to Broadcom. Broadcom is the top vendor of networking, storage, and computing connectivity technologies. As mentioned earlier, 100% of public clouds use Broadcom chipsets today, and 99.9% of internet traffic goes through our technology. Further, all the leading smartphone vendors, including Apple, Google, and Samsung, rely on Broadcom technology.

Broadcom delivers the market’s leading AIOps solution, which offers open and scalable data analytics. AIOps from Broadcom provides the key capabilities teams need to optimize modern environments:

  • Automation and predictive analytics, enabling the triggering of autonomous remediation so issues can be addressed before they have an impact on users or the business.
  • Full-stack observability, providing monitoring coverage of applications, infrastructure, and networks.
  • Service modeling, offering views that connect business services with all their underlying components.
  • Alarm noise reduction, eliminating the noise from the signal and helping IT speed mean time to resolution.
  • Application-centric network topology, mapping issues to associated business services, enabling more intelligent prioritization.

Combined, our chipsets and AIOps solution can deliver breakthrough capabilities. With these combined offerings, teams can do granular, packet-level network monitoring, leveraging in-band network telemetry to do real-time flow monitoring. AIOps from Broadcom can ingest rich packet-level network data from Trident chipsets. Now, Broadcom’s new Trident 4 switch series combined with our AIOps solution delivers AI-driven network triage and remediation. With these offerings, Broadcom is uniquely qualified to help organizations harness real-time analysis and correlation of network data, so over time teams can institute reliable, self-healing networks.

Conclusion

For today’s hyper-connected businesses, the establishment of AI-enabled, self-healing networks represents an increasingly vital mandate. For further insights, watch the below video.