Game changers was the theme at Juniper Networks NXTWORK 2019 event in London this week, with the highlight being an audience with a true game changer, John McEnroe. Appledore identified three network “game changing” ideas at the event.
Edge and not a use case in sight
Juniper’s CEO Rami Rahim, in his slides on edge, did something rare in the industry; he didn’t start his discussion with edge use cases. Instead, he talked about the structural change that edge brings to networks, and its position as a natural continuation of network/service disaggregation. In simple terms, Juniper positioned edge as the enabler of data and services disaggregation from the network data plane.
Appledore believe the focus on specific compelling and killer edge applications is a problem, because it distracts from the open ended and unknown scope for edge innovation. It also hides the fact that edge is disruptive technology that does not necessarily drive vertically integrated solutions.
By highlighting that edge is fundamentally the disaggregation of the data and services plane, Juniper neatly encapsulated the challenge of edge cloud for CSPs. If you can disaggregate the data plane at the edge, then anybody can innovate network based services at the edge. There is no barrier to others innovating and competing at the edge and CSPs are potentially left with the commoditised dumb connectivity pipes.
Appledore will be releasing a market outlook “Edge Investment Model and Impact – How to invest in Edge and make money” in the new year which will look at the opportunities in edge cloud including disaggregation.
Closing the loop with AI
Juniper’s recent acquisition of Mist was centre stage and was exciting for two reasons:
- Mist is closing the loop in terms of enabling self managing and self optimising networks
- Mist is achieving this with AI and machine learning.
Mist comes from the WiFi network space, but Juniper has big ambition to roll out this capability to other network types. With five years experience in building AI managed networks, Mist brings methodology and experience of AI driven networks. This will be key to enabling the building of trust in AI driven networks by operational staff.
Is better WiFi good enough for most private networks?
Mist is being used to support factory automation using WiFi, providing higher QoS and better connection setup, including for latency critical applications.
LTE is a better standard for delivering performance guarantees, licensed spectrum, mobility and coverage for a private network. However, Juniper believe that WiFi is significantly cheaper to implement than LTE for private networks. With Mist potentially solving key WiFi challenges, like poor configuration and management, some latency sensitive applications could be delivered on a cheaper/existing private WiFi network. Suppliers of LTE based private networks, need to ensure that their private network ROI is based on real technological difference and customer benefits, not simply the fact that WiFi networks are often poorly managed.