Juniper Networks has made repeated attempts on the wireless LAN market, with limited success. Its 2010 purchase of enterprise WiFi pioneer Trapeze Networks failed to deliver the hoped-for market share boost against Cisco (it generated just under $50m a year from 2011-2013 and then fell away); HPE acquired Juniper’s next strategic WLAN partner, Aruba; and its latest (and most promising) alliance, with Ruckus Wireless, ended when that firm was snapped up by Brocade in 2016 (and subsequently sold on to Arris).
Undeterred, Juniper is making its fourth attempt with plans to buy WLAN start-up Mist Systems for $405m, a deal which should close in the second quarter. Juniper CEO Rami Rahim wrote in a blog post that the acquisition would extend the firm’s enterprise wireless portfolio and allow it to stake a claim “to AI-driven operations in the era of multi-cloud”.
In an analyst call to discuss the purchase, Rahim was more bullish, saying it was an “offensive move” against Cisco and HPE/Aruba. He said that Mist would fill a gap in Juniper’s software-defined enterprise portfolio and help it increase its share of the wireless networking market, and of the campus switching space. Juniper could also become more self-sufficient in WLAN, which it currently sources mainly from HPE/Aruba and Aerohive.
Based in Cupertino in Silicon Valley, Mist was founded in 2014 by three former Cisco executives – Bob Friday, now CTO; Brett Galloway, now chairman; and Sujai Hajela, now president and CEO. To date, it has raised $88.4m in three funding rounds.
It first worked with its soon-to-be parent last fall when Juniper integrated Mist’s Learning WLAN with its Contrail SD-WAN technology, providing enterprises with visibility into their wired and wireless LANs and WANs.
Mist’s differentiation from other WLAN platforms lies, it says, in its use of artificial intelligence to manage and predict the network’s performance and make the WiFi experience more predictable, reliable and measurable. Its cloud-based system features an AI-driven virtual assistant called Marvis, which uses patented dynamic packet capture and machine learning technology to identify and fix network issues automatically.
It also uses virtual Bluetooth LE technology to enable indoor location services like wayfinding, proximity messaging and asset visibility.
Its new parent says it will integrate Mist’s cloud management and AI engine into its enterprise platform and will combine the Mist solution with its own (W)LAN, SD-WAN and security offerings. That combination will deliver three benefits, wrote Rahim in his post:
- Mist will deepen Juniper’s offering in the wireless networking market, enabling it to bring an end-to-end software-defined stack to the enterprise. “Operations is inherently an end-to-end proposition. User experience is not siloed. It requires all of the elements, from wireless access to the wired LAN across the WAN, to servers in the data center or cloud to work together,” he wrote. “Having the breadth of product to service this entire space is important.”
- Mist’s AI engine will extend AI-driven operations to the full IT stack including networks.
- The start-up’s cloud-first microservices architecture will accelerate Juniper’s push to build AI-driven operations software.
Verizon Enterprise Solutions launched a software-defined WLAN managed service based on Mist’s technology last year; Mist has also worked with VMware to build a joint product with interoperability between Mist’s Learning WLAN and VMware’s VeloCloud-based NSX SD-WAN.
“WiFi is increasingly becoming the primary access mechanism for most companies, yet it is harder than ever to operate and manage WLANs given the vast array of devices, operating systems and applications,” said Mist CEO Hajela. “The top WLAN vendors are dealing with platforms that are over a decade old, making it difficult – if not downright impossible – to leverage the latest cloud, AI and wireless technologies to meet modern mobile user needs.”