Radio Resource Management (RRM) is all about delivering an automated way to tune the RF in your Cisco network.
However - for a variety of reasons, it doesn't always get it right. Misconfigured RRM can cause a lot of issues for clients - particularly with roaming.
There is now a better option - Cisco Artificial Intelligence Enhanced RRM.
What is AI-Enhanced RRM?
To answer this question, we will first of all need to describe what RRM is. This stands for Radio Resource Management and is a collection of tools on the WLAN controller which allows us to tune the RF.
The Artificial Intelligence (AI) piece is the new add-on. This essentially describes the process of taking a larger data set than the instant data set normally used by RRM, to make better, longer term decisions about how to tune the network.
Here's a quick review of the AI-Enhanced RRM Control Center:
Cisco AI-Enhanced RRM Overview
*Image copyright Cisco Systems
OK – let’s move on to look at why you would want to deploy AI-Enhanced RRM – and what some of these RRM tools we’ve mentioned here actually are.
ASSOCIATED BLOGS:
- How does Cisco AI-Enhanced RRM Work?
- How to Configure Cisco AI-Enhanced RRM
- Wi-Fi 6
- Wi-Fi Design Technical Requirements
- Wi-Fi Design Business Requirements
What are the Benefits of AI-Enhanced RRM?
To implement any changes in your network, you need to understand why you're making those changes.
So - in a nutshell - what are the benefits of AI-Enhanced RRM?
Here's a quick bullet list:
- Simplicity: First up - the profiles are simple to configure and use familiar concepts in their configuration
- Configurability: While the profiles are simple - you still have access to configure the deeper-dive options
- Granularity: You can typically apply per building - so nice and easy to tune different buildings, with different RF characteristics
- Busy Hours: The ability to not make changes during busy hours is handy to ensure stability
- AI-Driven Actionable Insights: This is the major advantage - getting advanced insights into your network performance allows you to tune and operate your network with much greater confidence of success
ASSOCIATED BLOGS:
Why do you need AI-Enhanced RRM
There’s a bit of a gap currently between the current RRM (Radio Resource Management) algorithm and what we actually need in order to tune a Wi-Fi network - better control over the network, cell size and so on. To tune via RRM, we use a set of pre-configured WLAN Controller features.
.png?width=212&name=Wi-Fi%20Surveys%20(Instagram).png)
What are those features? Here’s a quick recap:
-
-
-
-
-
-
- TPC: Transmit Power Control
-
-
-
-
-
-
-
-
-
-
-
- DCA: Dynamic Channel Allocation
-
-
-
-
-
Within the bounds of the above controls, we can also configure static (or semi-static) configurations – sometimes you need to overrule the standard RRM operation with these static settings.
For example – edge APs- tend not so tune their power properly with TPC, so we might need to use power caps. Of note, and I think often overlooked, is the TX Threshold – which is an incredibly important tool in ensuring a refined RF / RRM environment.
ASSOCIATED BLOGS:
AI-Enhanced RRM: Expectation vs Reality
Things don't always work as you expect: RRM is all about providing automated tuning for your Wi-Fi network, but it doesn't always get it right.
Your expectation sits on one site of the equation (you expect it to be plug and play and deliver great outcomes), but the tuning doesn't always produce the optimal output.
The image below describes this equation: AI-Enhanced RRM is here to close the gap in expectation vs reality:
The RRM Challenge Expectation vs Reality
*Image copyright Cisco Systems
ASSOCIATED BLOGS:
The Solution: Cisco AI-Enhanced RRM
Proactive, Artificial Intelligence driven RRM is the new tool in Cisco DNA Center to allow you to set these functions and allow Cisco DNA Center to apply AI to your RRM setup.
The image below provides an overview of the reporting screen (left hand side) and the configuration screen (right hand side), used to deliver AI-Enhanced RRM from Cisco DNA Center:
Display - and Configure for AI-Enhanced RRM in Cisco DNA Center
*Image copyright Cisco Systems
This is the first point: AI-Enhanced RRM is delivered via Cisco DNA Center. This is required as the RRM data is processed in the cloud (made possible via Cisco DNA Center).
On ticking the box to enable the AI-Enhanced RRM feature, your anonymised RF data will be sent to the Cisco AI cloud, where it is processed to spot patterns and from these patterns, information is sent back to allow your particular network to be optimised.
ASSOCIATED BLOGS:
Longer term RRM View
Normally RRM is making periodic determinations on short term data (depending how you’ve configured your network, this could be from a 10 minute window up to a 24 hour window), but it is only making a decision for that point in time – with AI-Enhanced RRM, there are up to two weeks worth of data used to make the best decision on how to optimise your network.
This means that patterns can be used to optimise, and prevent the see-saw effect that can happen with standard RRM, where it is optimising, making a change, then 10 minutes later, another change and so on – and never quite achieves the optimum equilibrium for your site.
Here's how Cisco DNA Center displays the status changes from a high level:
RRM Changes - and Performance in Operation
*Image copyright Cisco Systems
This longer term view means that periodic events in your network – such as an all hands meeting on a Wednesday morning – are taken into account when analysing and optimising.
ASSOCIATED BLOGS:
Introduction to Cisco AI-Enhanced RRM Summary
In this first blog, we have sought to just define what AI-Enhanced RRM is - and what problem is is actually solving.
.gif?width=200&name=Wi-Fi%206%20Key%20Benefits%20(Instagram).gif)
The answer to the second question is pretty simple: let's apply a better set of logic to the problem of tuning RRM and we'll get better outcomes.
To do that, we're using Artificial Intelligence to take a much wider data set into account for the purpose of tuning RRM - this allows periodic things that happen in your Wi-Fi network to be taken into account.
No algorithm will ever get RF tuning right all the time - but this is a major leap forward.
The outcome for you and your network - a better tuned network, delivering a better set of outcomes for your end client devices - and your end users.
ASSOCIATED BLOGS:
