In discussion on network performance
It was this realisation that led the founders of Mist Systems to take a different approach when founding their company in 2014, by adopting newer generation technologies such as microservices, artificial intelligence (AI), and cloud computing as core components of their architecture.
According to product manager Yedu Siddalingappa, the result is a wireless network platform that is both highly adaptable and able to deliver superior user experience.
Critical to these outcomes is the ability of the network to collect and analyse vast amounts of data about its own performance.
"We have built a capability for access points to collect a large amount of data and process it using AI algorithms to give service level experience (SLE) metrics in real time," Siddalingappa said. "So we are able to show if clients are getting proper throughput and proper coverage, and if they are not, we can identify the root cause at the most granular level."
The AI capability within Mist equipment can identify where problems are occurring, provide advice to administrators on the steps they should take, or even remedy them automatically in some cases.
This AI-based capability is presented to administrators in the form of Marvis, an AI-powered engine which enables administrators to post queries using natural language.
"Marvis gives a natural language interface for network operators to manage the system," Siddalingappa said. "It can accept chat commands like 'troubleshoot my client' or 'who are the unhappy users on my site', or 'get a list of all the users on this floor during this time'.
"IT systems are becoming very complex nowadays and IT teams are shrinking, so people have to do more with less resources. Marvis proactively hunts problems in the network and presents the steps needed to be taken to rectify them."
Because Mist uses a cloud-based microservices architecture, upgrades and patches can be applied automatically without interruption to network performance. This also boosts security by ensuring all equipment is automatically patched with the latest fixes.
"Since the firmware updates happen very fast, that gives better security for the end users," Siddalingappa said. "And for the network operators, they feel more empowered, because they have more tools at their disposal now to measure the user experience, understand the problems and rectify them."
All of this comes together to create an SLE-based operating model where the management dashboard is focused on reporting user experience rather than device performance.
And because Mist was developed to address the challenges of the next decade, it was also designed with modern concepts such as the Internet of Things and Bluetooth Low Energy networking technology in mind.
Siddalingappa said this suits the needs of business users who are looking to their networks for more than just connectivity.
"In a healthcare facility the Wi-Fi network is not just used for connectivity, it is used for guest access, asset tracking, and to improve the visitor experience using things like wayfinding," he said. "So there is a lot of intelligence present within the network, and with all this intelligence we can enable many other use cases which are applicable to many other industries."
In 2019 Mist Systems was acquired by Juniper Networks. Accordingly to Siddalingappa, this represents the perfect marriage due to the two companies' respective focuses on wireless and wired technology, and their shared vision for an AI-driven future.
"Going forward our vision is to bring the AI capability which is present on the Wi-Fi networks today to all the other elements of the network," Siddalingappa said. "We have already integrated the wired stack of Juniper on our AI platform and going forward you will see SD-WAN, security and many more things, towards the vision of AI-driven networking."
Get results with AI-driven enterprise.