Remember that time you lost internet in the middle of a digital meeting you were leading? Many of us have stories like this to lament at this stage in the COVID-19 pandemic as our lives function now predominantly online, and our networks are strained – albeit holding up respectively considering the conditions.
One truth laid bare by the pandemic is how vital technology is for much of our existence – from first responders saving lives to everyday people staying connected to family, friends, and coworkers. As we all increasingly rely on these capabilities (and expect perfect performance), it’s no wonder that the next generation of communications technology is a hot topic.
Interest in 5G technology understandably spiked as the global pandemic set in, with Google search data indicating peak popularity in early April. While there are a range of opinions circulating about the pandemic’s impact on 5G adoption, our need for faster, more agile network infrastructure is unquestionable. In our current state of affairs, we need 5G more than ever to support this massive influx in demand.
Beyond the significant jump in download speeds, 5G’s ability to manipulate the network to address real-time demands makes it uniquely suited for today’s challenges. Slicing dedicates specific portions of the network by frequency range for specific use cases depending on their bandwidth demands and overall importance. Use cases that are mission-critical (such as communications for autonomous vehicles or medical devices) are prioritized in the more robust slices of the network, but they’re adaptable, so bandwidth can be made available to other use cases as demand ebbs and flows.
With much ground still left to cover in terms of widescale 5G deployment, it’s important that we accelerate the process. Intelligent automation is a vital tool for advancing 5G rollouts in a number of ways. It’s not simply a tool though – automation is fundamental to the very existence of 5G.
The 5G-Automation Link
5G and automation are inextricably tied together. For starters, automation is the only way to manage the network slices and appropriate them accordingly, ensuring the all-important reliability of throughput on designated frequency ranges. But this is just the beginning; there’s a much wider range of use cases that benefit from automation than you might initially imagine.
Apart from dynamic provisioning, automation is also necessary to connect all the dots in the network. The antennas, transmitters, receivers, and all the other devices need to be communicating and working together properly; automated heath checks can quickly surface problems in the making, so 5G providers can get ahead of issues.
With the transition to 5G, there’s a massive leap in infrastructure that needs to be managed, monitored, and maintained. There’s a lot more hardware to start with as the 5G antenna size is much smaller – the size of a fingernail in some cases – which means far more antennas are required to create the network. The new antennas also consume significantly more power. As an example, estimates indicate that upgrading to 5G would increase Vodafone’s base station power draw by 60%. Automation is the only practical way to handle management as equipment proliferates exponentially.
The underlying infrastructure expansion that accompanies 5G makes managing and monitoring the network a more complex job while fixing issues ASAP is even more important, given the types of traffic the network is carrying. All of the new network devices translate to far more points of failure, and new underlying technologies like Software Defined Networking (SDN) and increasing virtualization can make it difficult to identify the root cause of issues. Without technologies like automation and AIOps to aid network operations teams, incident response quickly becomes a game of Whack-a-Mole – except far less entertaining. Automation can collect diagnostics in nanoseconds and even pinpoint the root cause of many issues, massively accelerating incident resolution.
Provisioning and deploying new equipment on the 5G network also sit squarely in the automation wheelhouse, making it significantly faster to turn up and turn down new infrastructure as needed. Automation can also provision the networks dynamically to adjust for peaks and valleys in traffic consumption, many of which are predictable (especially with machine learning) based on days of the week, time of day, and geographies. Automation has a similar application in managing network throttling used to regulate traffic and minimize congestion.
Predicting and Preventing are Poised to Go Mainstream
Artificial intelligence and machine learning can further facilitate 5G management when coupled with automation. While in earlier stages of adoption, 85% of IT organizations indicated that AI for IT operations (AIOps) is either a major initiative or well underway, according to new data from EMA Research. AIOps solutions ingest, aggregate, and analyze large volumes of data, relying on machine learning to produce valuable insights, perform event correlation and clustering, and reduce alarm noise. Over time, these tools can also predict future problems and automatically remediate them before they have an impact, moving us closer to self-healing IT.
When we look at how network consumption has grown during the course of the last three months, the numbers are staggering. Working from home has driven VoIP and video conferencing application usage up 210-285% with VPN traffic growing 30-40% on Comcast networks, the largest residential provider in the U.S. Microsoft Teams hit a daily record of 2.7 billion meeting minutes in one day on March 31, a 200% increase in just two weeks. Zoom daily sessions were up well over 6 million in late March.
There’s no shortage of similar statistics, and there’s no sign of abatement even as shelter in place orders are being relaxed. It’s clear that we need to up the ante on our network capacity with 5G, and automation is fundamental to getting there.