AI Operations

The AI Operations page displays all the network performance enhancement recommendations for all network zones and clients. A recommendation is a change suggested to the software configuration, and it is applied to the SmartZone controller.

Note: Access to AI Operations is available exclusively to users with the Prime Administrator role and designated access to the Default Resource Group.

RUCKUS AI creates recommendations after constantly monitoring the software configuration and analyzing network data. Changes in the expected behavior of a network or its performance are commonly detected and corrected by making physical changes to the network environment or by making configuration changes to the networking software. The network system configuration must be constantly monitored and regulated to achieve optimum performance. In the AI Operations page, RUCKUS AI provides recommendations by monitoring dynamic factors influencing network performance and also seeks to tune static factors with an objective to improve performance metrics, key performance indicators (KPIs), and the user experience. You can view the recommendations on this page and apply them to see how network performance can be improved. This page also provides access to insights into network performance before and after a recommendation is applied. A recommendation is a change suggested to the software configuration, and it is applied to the SmartZone controller.

Note: Only SmartZone controllers running SmartZone 5.2.2 or SmartZone 6.0 and later can make use of the AI Operations page in RUCKUS AI.

On the navigation bar, click AI Assurance > AI Analytics > AI Operations. The AI Operations page is displayed. The AI Operations page displays all the network performance enhancement recommendations in a table for the selected network node and date/time filters. When the table spans multiple pages, you can select the number of recommendations displayed per page from the drop-down at the bottom of the page.

You can use the search option to search the required recommendation by its summary, zone, priority, category, and status.

The table has the following attributes:

Priority: Displays the severity of the recommendation as Low, Medium, or High.
Date: Displays the date the recommendation was created.
Category: Displays the type of recommendation:
- Infrastructure: This category considers infrastructure changes specifically related to SmartZone firmware. If RUCKUS AI determines that the APs in a zone would benefit from a newer firmware release for performance or security reasons, this type of recommendation is generated.
- Wi-Fi Client Experience: This category includes changes to auto channel selection mode and/or background scan timers for 2.4GHz and 5GHz radios, enabling or disabling Dynamic Frequency Selection (DFS) for 5GHz, enabling or adjusting band balancing, enabling load balancing, and adjusting transmit radio power without using RRM.
Summary: Displays a short description about the suggested recommendation, for example, "Firmware Upgrade", "Enable DFS Channel", and so on.
Zone: Displays the affected zone name.
Status: Displays the status of the recommendation. The status is New when the recommendation is available but not applied. The status changes to Scheduled after you have scheduled to apply the recommendation. It changes to Applied after the scheduled recommendation has applied the change. If applying the recommendation fails, the status changes to Failed. Every state change triggers an email message to the network administrator providing information about the status. For more information, refer to Table 2.

The network node filter and Date and Time filter are displayed in the upper-right corner of the Content panel. These options control the elements displayed within the Content Panel. To modify these options, refer to Content Panel.

Click the Date attribute for a specific recommendation to display the AI Operations Recommendation Details page. This page displays the description of the recommendation, the reason why it is encouraged to apply the recommendation, the tradeoff in the network performance if it is not applied, the KPIs that are impacted due to the current configuration, and the status trail tracking how the recommendation has changed states. For more information, refer to AI Operations Recommendation Details.

Note: Recommendations are available for a period of 90 days, after which they are removed. RUCKUS AI runs a daily check of the recommendations created to see if they are still relevant and applicable. Recommendations that are no longer relevant or valid are automatically removed from the list.

Recommendations that are applied successfully are also reflected on the Config Change page. For more information, refer to Configuration Change.

Following are the key AI Operations Recommendations provided by RUCKUS AI to improve network performance, key performance indicators (KPIs), performance metrics, and the user experience.

Table 1. AI Operations Recommendations
Category	Summary	Why this Recommendation?
Wi-Fi Client Experience	Auto channel selection mode and background scan on 2.4 GHz radio	Auto Channel Selection feature works well only when RUCKUS APs can perform background scan of the available channels in the network. This helps in building the RF neighborhood. APs can then select an optimum channel for their operation. Hence, it is recommended to enable the Background Scan feature.
Wi-Fi Client Experience	Auto channel selection mode and background scan on 5 GHz radio	Auto Channel Selection feature works well only when RUCKUS APs can perform background scan of the available channels in the network. This helps in building the RF neighborhood. APs can then select an optimum channel for their operation. Hence, it is recommended to enable the Background Scan feature.
Wi-Fi Client Experience	Background scan timer on 2.4 GHz radio	An optimized scan timer for background feature enables RUCKUS APs to scan the channels for an appropriate time interval. Time interval that is too long would result in longer time for radio channel selection.
Wi-Fi Client Experience	Background scan timer on 5 GHz radio	An optimized scan timer for background feature enables RUCKUS APs to scan the channels for an appropriate time interval. Time interval that is too long would result in longer time for radio channel selection.
Wi-Fi Client Experience	Enable DFS channels	Enabling the DFS channels give more available channel options for the AP and enable the AP radio to pick the best available channel. This will help to reduce co-channel interference, enhance user experience and improve throughput for 5 GHz Wi-Fi connections.
Wi-Fi Client Experience	Disable DFS channels	If AP is placed in an area where there are genuine and consistent DFS Radar signals, then the AP should not select DFS channel to ensure uninterrupted operations.
Wi-Fi Client Experience	Enable band balancing	Band Balancing feature intelligently distributes WLAN clients on 2.4 GHz and 5 GHz channels to balance the client load. Band Balancing results in better radio utilization and gives better Wi-Fi experience to the user.
Wi-Fi Client Experience	Change band balancing mode	Band Balancing (BB) feature intelligently distributes the WLAN clients on the 2.4G and the 5G channels to balance the client load. The "PROACTIVE" mode has high efficiency in steering clients from one band to another and thus, balance the load on the AP resulting in better Wi-Fi experience to the user.
Wi-Fi Client Experience	Enable load balancing based on client count	Client Load Balancing allows equal distribution of Clients by allowing heavily loaded APs to move clients to less loaded neighboring APs. This ensures better radio utilization and provides better Wi-Fi experience to the user.
Wi-Fi Client Experience	Tx power setting for 2.4 GHz and 5 GHz radio	The AI Analytics encourages client association to 5 GHz and reduces co-channel interference on 2.4 GHz.
Infrastructure	Zone firmware upgrade	The latest AP Firmware version in the zone ensures all the APs in the Zone to have the best available firmware with appropriate security or bug fixes and new features.
AI-Driven Radio Resource Management (RRM)	AI-Driven RRM	Based on the AI Analytics, enabling the AI-Driven RRM will constantly monitor the network, and adjust the channel plan, bandwidth, and AP transmit power when necessary to decrease the number of interfering links to zero. For more information, refer to AI-Driven RRM Overview.

Applying a Recommendation

Complete the following steps to apply a recommendation:

On the navigation bar, click AI Assurance > AI Analytics > AI Operations. The AI Operations page is displayed.
In the AI Operations table, click on the radio button of a new recommendation.
Click Apply, displayed at the top of the AI Operations table.
A calendar prompt is displayed to schedule the action.
Select the date and time in the calendar prompt to schedule the time at which you want to apply the recommendation.
You can select any date up to 30 days in the future and specify a time for the recommendation to be applied. The chosen time will be stored and applied to future changes if RUCKUS AI determines that a more optimal configuration is available.
Click Apply to enable the recommendation to be applied at the scheduled time.
Once scheduled, the action options are updated to include the ability to adjust the schedule or cancel the action.
Note: After you apply the recommendation, it takes an hour to apply the configuration change to the controller.
Alternatively, you can apply a recommendation from its Recommendation Details page (accessible by clicking the Date attribute of the recommendation in the table). In the Recommendation Details page, click the icon in the top right corner, and from the Actions list, click the icon. Select the date and time in the calendar to schedule the time at which you want to apply the recommendation and click Apply to enable the recommendation to be applied at the scheduled time.

Reverting a Recommendation

Complete the following steps to revert a recommendation and restore the previous configuration settings:

On the navigation bar, click AI Assurance > AI Analytics > AI Operations. The AI Operations page is displayed.
In the AI Operations table, click on the radio button of a recommendation that you want to revert.
Click Revert displayed at the top of the AI Operations table.
A calendar prompt is displayed to schedule the action.
Select the date and time in the calendar to schedule the time at which you want to revert the recommendation.
Click Apply to enable the recommendation to be reverted at the scheduled time.

Note: After you apply the recommendation, it takes an hour to apply the configuration change to the controller.

Alternatively, you can revert a recommendation from its Recommendation Details page (accessible by clicking the Date attribute of the recommendation in the table). In the Recommendation Details page, click the icon in the top right corner, and from the Actions list, click the icon. Select the date and time in the calendar to schedule the time at which you want to revert the recommendation and click Apply to enable the recommendation to be reverted at the scheduled time.

Muting a Recommendation

Complete the following steps to mute a recommendation:

On the navigation bar, click AI Assurance > AI Analytics > AI Operations. The AI Operations page is displayed.
In the AI Operations table, click on the radio button to select a recommendation that you want to mute.
Note: You can mute only one recommendation at a time.
Click Mute, displayed at the top of the AI Operations table to mute the recommendation. When an recommendation is muted, it is hidden in the AI Operations table.

Alternatively, you can mute a recommendation from its Recommendation Details page (accessible by clicking the Date attribute of the recommendation in the table). Refer to Mute and Unmute the Recommendation.

Viewing the Muted Recommendation in the AI Operations Table

In the AI Operations table, click the icon and select the and select the Show Muted Recommendation check box. The muted recommendations are displayed in the AI Operations table with a gray background.

Unmuting a Recommendation

Complete the following steps to unmute a recommendation:

On the navigation bar, click AI Assurance > AI Analytics > AI Operations. The AI Operations page is displayed.
In the AI Operations table, click on the radio button to select a muted recommendation.
Note: You can unmute only one recommendation at a time.
Click Unmute, displayed at the top of the AI Operations table to unmute the recommendation. When an recommendation is unmuted, it is visible in the AI Operations table.

Alternatively, you can unmute a recommendation from its Recommendation Details page (accessible by clicking the Date attribute of the recommendation in the table). Refer to Mute and Unmute the Recommendation.