You are on page 1of 5

White Paper

ExtremeAI for Smart OmniEdge™


Transform your user experience and enterprise network
operations with Artificial Intelligence/Machine Learning

We are living in a time where the amount of data traversing over networks
is so astronomical that it is impossible for humans to process, analyze
and act on it. That makes it essential that we utilize the power of artificial
intelligence and machine learning to harness the data and process it faster,
smarter and more efficiently.

ExtremeAI for Smart Few areas in a network can benefit more from Artificial Intelligence (AI) /
Machine Learning (ML) than Wireless LANs RF management. As business-
OmniEdge automatically
critical systems, Wireless LANs demand flawless operation. They are
monitors critical RF and critical to enterprises. Not only do they connect people, they also enable
network parameters, to everything from point-of-sale systems to enterprise applications, real-time
optimize the network communication, video conferencing, and more.
and make proactive However, given the dynamic nature of a wireless network, it can be
extremely challenging to maintain and operate. For IT departments to
adequately support Wireless LAN applications and users, a proactive
rather than reactive approach is required. ExtremeAI for Smart OmniEdge
leverages artificial intelligence and machine learning to deliver a better
networking experience for both IT and the end-user. By detecting and
correcting problems before they manifest themselves to the end-user,
ExtremeAI can greatly enhance the end-user experience and reduce the
maintenance and troubleshooting burden on IT, freeing up time for IT to
focus on value-add initiatives that supports their business’ objectives.

WWW.EXTREMENETWORKS.COM 1
ExtremeAI for Smart The Benefits of Artificial
OmniEdge can help reduce Intelligence/Machine Learning
reactive troubleshooting and
Machine Learning is a type of Artificial Intelligence that enables a
support calls by 50%.
computer to teach itself by analyzing new data, without being
programmed, with the goal of uncovering hidden insights. Unlike humans
Majority of Network Issues who can only react to problems once they have occurred, machines can
are "Reactive" automate processes and proactively help find failures before they become
major issues.
By gathering and analyzing data in real time, machines can correlate
...of downtime is information, identify patterns, “learn” to predict what may happen
37% due to human error.
next, and act on that information. Actions may include optimizing the
system, preventing failures, sending alarms, automatically healing
breaches, and generating actionable information for human experts to
base decisions upon.

...of network issues The Enterprise Wireless LAN Challenge


75% are identified by
end users.
With wireless networks now mission critical and over 50% of global traffic
running over Wi-Fi, downtime is no longer affordable or acceptable.
Enterprise Wireless LANs are the most challenging environment for IT
teams to manage given their dynamic nature - varying user densities, user
locations, distance and coverage, etc. This combined with the exponential
growth of new mobile users and devices puts a heavy burden on IT teams.

90% ...of time spent


is identifying the
ExtremeAI for Smart OmniEdge revolutionizes the IT operations experience
problem.
by providing proactive, automated, self-healing capabilities that enhance
the end-user experience and reduce IT outages and help desk calls. By
using machine learning technology and artificial intelligence to automate all
aspects of WLAN RF management, including configuration, dynamic tuning
Majority of Support Issues and optimization, load balancing, troubleshooting, diagnosis and correction,
are Wi-Fi related the network continuously learns and adapts the Wi-Fi environment,
proactively ensuring a best in class experience for mobile users.
64%
...of reported network issues
Proactive Wi-Fi Operations
are Wi-Fi issues. At the core of ExtremeAI is a machine learning-based engine offering
predictive analytics technology. The engine learns the typical characteristics
47% and behaviors of devices, environment and network that impact a Wireless
...of Wi-Fi issues impacted LAN’s performance. To access the same level of information solely from
at least 10% of employees. human intelligence would be virtually impossible.
Statistical data or real-time information is communicated to the cloud, while
40% the customer’s data and content remain untouched. Once in the cloud, this
...of reported Wi-Fi issues information is processed in the ML Engine, in some cases in real-time, as
are connectivity issues. network performance indicators.
The information is organized, processed, and translated into solutions that
23% trigger specific actions or actionable alerts, or automatically optimize and
...of reported Wi-Fi issues heal the network.
are performance issues. By correlating the wide variety of data that can be gathered, the system
quickly learns to identify issues with device compatibility, interoperability,
Figure 1: ExtremeAI for Smart OmniEdge access, use or misuse, and network security – as well as gaps in coverage or
reducing troubleshooting. capacity in the network.

WWW.EXTREMENETWORKS.COM 2
With this power at hand, even small IT organizations can give all users
an exceptional wireless experience across multiple locations, with fewer
headaches for IT personnel. This reduces or eliminates the need for
organizations to rely on deep RF knowledge or to undergo specialized
training to run the Wireless LAN.

“No Worry” Wi-Fi


ExtremeAI for Smart OmniEdge provides 24x7 machine-based real-
time diagnosis and optimization, proactively monitoring all network
Benefits of ExtremeAI services associated with a Wireless LAN. The service takes responsibility
for Smart OmniEdge for the performance of the wireless network day and night across all
• Automatically detects and corrects
the enterprise’s locations. When anomalies arise, it provides human-like
problems before they become decision making to raise actionable alerts, automatic correction of problems,
user-observable or advise the customer how to address a problem even before it arises.
• Human-like diagnosis using ML and
corrections at machine speed How ExtremeAI Works
• Significantly reduces support and ExtremeAI is a subscription-based service available on ExtremeWireless,
IT cost ExtremeWireless WiNG, ExtremeCloud and ExtremeCloud Appliance.
• Instant resolution for most common After subscribing to the service, the IT department specifies the services
Wireless LAN problems and interfaces that are to be enabled, and the security policies to be
• When external problems are detected, implemented. From that point, ExtremeAI takes over Wireless LAN
provides actionable alerts management, proactively monitoring and managing the health of all Access
• Real-time cross-layer analysis Points and network operations.
for root causing problems
IT managers can easily access their network data through the user interface.
(e.g., RF, DNS, DHCP)
It gives insight into what’s happening in the network environment, and the
• When manual support is required, all
mix of iOS, Android and other devices accessing an organization’s system.
the detailed information regarding the
problem is available with a single click
• Self-learning algorithms for
dynamic environments
Near-Zero
• Learns and adapts to changing
Configuration
local RF environment, devices, and
neighbor networks
• Learns and adapts to global events Preventive and
Last Resort –
such as known firmware/OS upgrades Call Support
Periodic RF
and related changes in behavior Management

• Zero configuration and tuning of


complex network parameters
• Hundreds of threshold and
parameters for various algorithms
controlled by a learning engine Automatic
Auto-tech Problem
Support Identification
and Resolution

Proactive
Notification
to Admin

Figure 2: Benefits of ExtremeAI for Smart OmniEdge

WWW.EXTREMENETWORKS.COM 3
Customer Data is Untouched
All customer’s proprietary data remains untouched by the ExtremeAI
service. All data traffic for the users remains within the local network. Only
key statistical parameters are forwarded on to the cloud for processing by
the Machine Learning Engine.
The service isolates each customer’s individual network data—and can also
isolate data from multiple networks for a single customer if desired.

Common Use Cases


ExtremeAI for Smart OmniEdge can improve Enterprises in many ways
by simplifying and optimizing the RF management. Below are common
examples that IT operators frequently face:

1. Wi-Fi networks designed for coverage, don’t guarantee capacity


• To optimize for capacity, network operators need to master concepts
such as co-channel interference, band steering, spectral capacity,
collocated APs and load balancing, which is not an easy feat. ExtremeAI
for Smart OmniEdge makes it easy through its learning and automatic
algorithms. The system learns from changing use patterns, and adapts
as needed to find the right balance between coverage and capacity.
Learning-Based WLAN Services
• Radio Resource Management 2. Identifying client device capabilities
• Channel and Cell Size Management • Users frequently bring their own devices to work and expect them to
• Intelligent client steering work seamlessly with the Wi-Fi infrastructure, creating major workloads
for IT operators. ExtremeAI for Smart OmniEdge automatically
• Link quality management
monitors critical RF and network parameters, and either adjusts those
• Fast Roaming
parameters or make proactive recommendations to IT.
• Voice and Video QoS
3. Configuring a “target” application throughput per device type,
or channel capacity
• Too much time is spent by network operators making manual
adjustments to RF in an attempt to optimize the Wi-Fi performance.
Too often this is an exercise in futility, as often what is planned for
differs greatly from what is actually experienced. After the initial
deployment, ExtremeAI for Smart OmniEdge monitors the usage
and types of applications and makes the necessary adjustments and
recommendations. IT no longer has to worry about which devices are
running what applications, and what their preset configurations
allowed them.

4. Optimizing Signal-to-Noise (SNR) ratio or measuring co-channel


interference
• Radio is a scarce resource, and in the Wi-Fi heavy world we live
in, devices will aggressively compete for that resource putting your
network (and you) under a lot of stress. The ExtremeAI for Smart
OmniEdge built-in radio resource management features, together with
the self-learning system and in-house RF experts, can analyze and
proactively implement performance enhancing configuration changes—
without having you spend valuable time on such optimization activities.

WWW.EXTREMENETWORKS.COM 4
High Quality Experience—for the End-User
and the IT Department
ExtremeAI provides, above all things, an exceptional Wi-Fi user experience,
no matter how large or multi-faceted the network is. With Machine
Intelligence-based operational management of the network, the enterprise
can deliver the following, without burden on the IT department:
• Consistent network throughput, speed, and performance when and
where it’s needed, 24x7
• Fast user authentication and network access, no matter what device
is connecting
• Seamless application experiences for VoIP and video conferencing,
among others
• Rapid, accurate troubleshooting and fast network recovery in the case
of failures
The result is a superior experience for both end-users and for network
IT personnel.

Figure 3: ExtremeAI for Smart OmniEdge – Analytics Dashboard Example – Summary of automated Issue
detection, Root Cause Determination, Machine Learned Mitigations and Performance results

http://www.extremenetworks.com/contact Phone +1-408-579-2800

©2018 Extreme Networks, Inc. All rights reserved. Extreme Networks and the Extreme Networks logo are trademarks or registered trademarks of Extreme Networks, Inc. in
the United States and/or other countries. All other names are the property of their respective owners. For additional information on Extreme Networks Trademarks please
see http://www.extremenetworks.com/company/legal/trademarks. Specifications and product availability are subject to change without notice. 17022-0618-18

WWW.EXTREMENETWORKS.COM 5