Configure Coverage Hole Detection

In this post we will learn about CHD @RRM

Coverage holes are areas where clients can’t receive a signal from the wireless network. If clients on an AP are detected at low received signal strength indicator levels, Cisco lightweight APs send a coverage hole alarm to the cisco WCS/NCS or PI.

The RRM coverage hole detection algorithm can detect areas of radio coverage in a wireless LAN that are below the level needed for robust radio performance. This feature can alert us to the need for an additional (or relocation) lightweight access point.

If clients on a lightweight access point are detected at threshold levels lower than those specified in the RRM configuration, the access point sends a “coverage hole” alert to the controller. The alert indicates the existence of an area where clients are continually experiencing poor signal coverage, without having a viable access point to which to roam.

The controller uses the quality of client signal levels reported by the APs to determine if the power level of that AP needs to be increased. Coverage hole detection is controller independent, so the RF group leader is not involved in those calculations. The controller knows how many clients are associated with a particular AP and what the signal-to-noise ratio (SNR) values are for each client.

If a client SNR drops below the configured threshold value on the controller, the AP increases its power level to try to compensate for the client. The SNR threshold is based on the transmit power of the AP and the coverage profile settings on the controller.

The controller uses the following equation for detecting a coverage hole:

Client SNR Cutoff Value (ldB|) = [AP Transmit Power (dBm) – Constant (17 dBm) -Coverage Profile (dB)]

Depending on the number of clients that are at or below this value for longer than 60 seconds, coverage hole correction might be triggered, and the AP could increase its power level to try to remove the SNR violation.

If the AP is already at power level 1, it cannot increase the power any further, and clients at the edge of the cell coverage suffer a performance hit or disassociate altogether if the signal gets weak enough.

Aside from a real coverage hole, a client with a poor roaming logic might not roam to another AP as expected and be “sticky.” A sticky client can remain associated with an AP until the SNR is very low and triggers a false coverage hole detection.

The coverage hole algorithm also allows the network to heal itself if an AP fails. When a neighbor AP is lost, it increases the power of nearby APs as needed to compensate. Again, the increase in power for an AP is a gradual process, increasing the power one level at a time.

Configure Coverage Hole Detection

Enable Coverage Hole Detection check box to enable coverage hole detection, or unselect it to disable this feature.

Data/Voice RSSI text box, enter the minimum receive signal strength indication (RSSI) (It must be between -60 to -90 dBm and can be different for voice and data) value for data/voice packets received by the access point. The value that we enter is used to identify coverage holes within our network.

Min Failed Client Count per AP text box, the minimum number of clients on an access point with an RSSI value at or below the data or voice RSSI threshold. The range can be from 1 to 75, and default value is 3.

Coverage Exception Level per AP text box, the percentage of clients on an access point that are experiencing a low signal level but cannot roam to another access point. The range is 0 to 100%, and default value is 25%.

Note: Coverage hole detection is no longer a global setting and can be enabled or disabled on a per-WLAN basis: Coverage hole detection is enabled by default on the WLAN. One of the reasons we might want to disable this is because if we know a device is going to roam, it is advised that we enable the wireless on the device so that it can assist in finding coverage holes. Conversely, if several devices are stationary and have wireless as a backup, it would be advisable to disable this because we know the devices are not going to move and will not be able to provide intelligent information to help the coverage hole detection algorithm with its calculations.

Enable/Disable Coverage Hole Detection per WLAN basis: WLC software release 5.2 or later, we can disable coverage hole detection on a per-WLAN basis

Coverage hole detection is enabled by default on the WLAN.

Configure Transmit Power Control

This is one of the features of RRM on WLC and in this post we will see and learn the option under TPC.

This algorithm is responsible for reducing the power level on the APs to reduce excessive cell overlap and co-channel interference. TPC uses the RSSI calculations for the neighbor APs, and it determines effective changes only if there are more than three neighbor APs.

The TPC algorithm runs every 10 minutes (600Secs). The RF group leader runs TPC on a per-radio, per-AP basis. Therefore, a power adjustment on 802.11b/g has no bearing on the 802.11a power level settings for the same AP.

The minimum requirement for TPC is that a single AP needs to be heard by at least three other APs at -70 dBm or greater. Therefore, we must have at least four APs total. The logic behind the lowering of the power levels is that the third loudest neighbor is heard at -70 dBm or lower after the change.

The final purpose of the algorithm is to make sure that the third-loudest neighbor AP is heard at a signal level lower than the configured threshold (by default its –70 dBm).

***Note: The TPC algorithm is only responsible for turning power levels down.

TCP goes through these stages which decide if a transmit power change is necessary:

Find out if there is a third neighbor, and if that third neighbor is above the transmit power control threshold (-70dBm).
Determine the transmit power using this equation:

Tx_Max for given AP + (Tx power control thresh – RSSI of 3rd highest neighbor above the threshold).

Compare the calculation from step two with the current Tx power level and verify if it exceeds the TPC hysteresis.

If Tx power needs to be turned down: TPC hysteresis of at least 6dBm must be met. OR
If Tx power needs to be increased: TPC hysteresis of 3dBm must be met.

***Note: When a brand new APs boot up for the first time, it transmit at their maximum power level (its 1). When AP is power cycled, it uses their previous power settings.

***Note: It is important to remember that decreases in AP radio power levels are gradual, whereas increases can take place immediately. Therefore, if we change the RRM configuration settings, do not expect to start seeing the APs changing channels and adjusting their power as soon as we click Apply.

Now we will see the configuration steps@TPC

Via GUI:

Go to Wireless -> 802.11a/n or 802.11b/g/n -> RRM ->TPC

On this screen we have these options:

Power Level Assignment Method: There are 3 ways to configure TPC algorithm:

Automatic: This is the default configuration and the TPC algorithm runs every ten minutes (600 seconds).
On Demand: The algorithm can be manually triggered if we click the Invoke Channel Update Now
Fixed

Min/Max Power: Maximum and minimum power level assignment and we can choose between -10 to 30dBm.

Power Threshold: Default value for this parameter is –70 dBm but can be changed when access points are transmitting at higher (or lower) than desired power levels.

Power Neighbor Count: The minimum number of neighbors an AP must have for the TPC algorithm to run.

Power Assignment Leader: This field displays the IP address of the WLC that is currently the RF Group Leader. Because RF Grouping is performed per-AP, per-radio, this value can be different for the 802.11a & 802.11b/g networks.

Last Power Level Assignment: The TPC algorithm runs every 600 seconds (10 minutes). This field only indicates the time (in seconds) since the algorithm last.

Verification:

(WLAN1) >show advanced 802.11a txpower
 Automatic Transmit Power Assignment
 Transmit Power Assignment Mode................. OFF
 Transmit Power Update Interval................. 600 seconds
 Transmit Power Threshold....................... -70 dBm
 Transmit Power Neighbor Count.................. 3 APs
 Min Transmit Power............................. -10 dBm
 Max Transmit Power............................. 30 dBm
 Transmit Power Update Contribution............. SNI..
 Transmit Power Assignment Leader............... WLAN1 (10.35.80.1)
 Last Run....................................... 98 seconds ago

Configure Dynamic Channel Assignment

In this post we will learn about DCA and it’s a really cool feature of RRM.

DCA is managed by RF Group Leader (How to define RF leader, we saw in one of my last post)

DCA used to determine the optimal AP channel based on these parameters.

Load: Percentage of time spent transmitting 802.11 frames

Noise: Measurement of non-802.11 signals on every serviced channel

Interference: Percentage of radio time used by neighbor 802.11 transmissions

Signal strength: Received signal strength indication (RSSI) measurement of the received neighbor messages

These values are then used by the Group Leader to determine if another channel schema will result in at least a bettering of the worst performing AP by 5dB (SNR) or in other words: Based on these metrics, if the worst performing AP will benefit by at least 5 dB or more, a channel change will take place. The decision to change the channel of an AP is also weighted to prevent a mass change within the RF group. We would not want to have a single AP change channel and have that change result in 20 other APs having to change their channel. The controller also takes into account how heavily an AP is used. A less utilized AP is more likely to have a channel change instead of a heavily used neighbor (isn’t it an interesting feature?). This helps mitigate client disassociations during a DCA event because a radio channel change disconnects all associated clients.

***Note: When an AP first boots up out the box, it transmits on channel 1 on the 802.11b/g radio and channel 36 for the 802.11a radio. The channels change according to any DCA adjustments if necessary. If a reboot occurs, the APs remain on the same channel they were using before the reboot until a DCA event occurs. If an AP is on channel 152 and reboots, it will continue to use channel 152 when it comes back up.

***Note: Radios using 40-MHz channels in the 2.4-GHz band or or 80MHz channels are not supported by DCA.

The RRM startup mode is invoked in the following conditions:

In a single-controller environment, the RRM startup mode is invoked after the controller is rebooted.
In a multiple-controller environment, the RRM startup mode is invoked after an RF Group leader is elected.

Configure DCA:

***We must disable 802.11a and b radio before changing the config. for DCA and then enable it again. Simplest way to enable/disable the radio is via CLI:

(WLAN1) >config 802.11a disable network
(WLAN1) >config 802.11a enable network

Go to Wireless > 802.11a/n or 802.11b/g/n > RRM > DCA

There is three type of Channel Assignment Method.

Channel Assignment Mode:

Automatic: This mode will cause the controller to periodically evaluate and, if necessary, update the channel assignment for all joined access points.
Freeze: It will Causes the controller to evaluate and update the channel assignment for all joined access points, but only when we click Invoke Channel Update Once.
OFF: Turns off DCA and sets all access point radios to the first channel of the band.

Avoid Foreign AP Interference: It detect foreign AP and take into consideration while changing the channel.

Avoid Cisco AP Load: When its enabled then the AP load is taken into account before result in which AP will change the channel (least loaded AP will change the channel first.

Avoid Non-802.11a (802.11b) Noise: It cause the controller’s RRM algorithms to consider noise (non-802.11 traffic) in the channel when assigning channels to lightweight access points.

Avoid Persistent Non-Wi-Fi Interference: Its enable the controller to ignore or avoid from persistent non-Wi-Fi interference.

Channel Assignment Leader: The IP address of the RF group leader, which is fully responsible for channel assignment.

Last Auto Channel Assignment: The last time RRM evaluated the current channel assignments.

DCA Channel Sensitivity: We have 3 levels (Low, Medium and High)

Channel Width: depends on the 802.11a or b radios: 5GHz select 40MHz. In 2.4 GHz it will be 20MHz.

Avoid check for non-DFS channel: Enabled then the controller avoid checks for non-DFS channels. (Apply only for outdoor APs)

DCA Channel List: This option shows the selected channel on this radio.

Dont forget to enable both radios after changing the parameter in this section by using these commands 🙂

RRM (Radio Resource Management) Overview

The RRM feature is also known as Auto-RF or act as a built –in RF engineer in controller, uses the RF information gathered by the APs to make decisions on whether channel assignment or power levels need to be adjusted.

In other words: It uses the RF information gathered by the APs to make decisions on whether channel assignment or power levels need to be adjusted. Just because the RF environment has changed does not necessarily mean that the controller will change.

Before covering the intricacies of the RRM algorithm and RF grouping, following is a high-level overview of the basic workflow involved:

Step 1: The controllers and their APs use the configured RF group name to determine if other APs they hear are part of their RF group.

Step 2: The APs use neighbor messages (sent every 60 seconds) that are authenticated by other APs that hear them. The neighbor messages include information about the AP, the controller, and the configured RF group name.

Step 3: The APs that hear the neighbor message of another AP authenticate that message using the RF group name and pass it to their respective controller.

Step 4: The controllers use this information to determine what other controllers should be in their RF group, and then form logical groups to share the RF information from their respective APs, and elect an RF group leader.

Step 5: The RF group leader runs the RRM algorithm against the RF information from all the APs in the RF group. Depending on the outcome, a power level or channel change for an AP or group of APs might take place.

To know more details about RRM, check this previous post:

https://rscciew.wordpress.com/2013/12/04/radio-resource-management/

Also don’t forget to see these YouTube video by Jerome Henry:

http://www.youtube.com/watch?v=gwCxVwmHnRw – describes RRM principles
http://www.youtube.com/watch?v=XhmnXeeLQBc – goes deeper into RRM and provides useful information if you are to take a Cisco exam on Wireless related topics! 🙂
http://www.youtube.com/watch?v=3EnvhxjzEWU – details how RRM controls the AP channel assignment with DCA (Dynamic Channel Assignment).
http://www.youtube.com/watch?v=32YWzuXTg5M – explains how RRM dynamically reduces AP power with TPC (Transmit Power Control)
http://www.youtube.com/watch?v=yot63RsKOCg – explains how the Radio Coverage Detection Algorithm works.

RRM feature enables controllers to continually monitor their associated LAP for the following information:

Traffic load: The total bandwidth used for transmitting and receiving traffic. It enables wireless LAN managers to track and plan network growth ahead of client demand.
Interference: The amount of traffic coming from other 802.11 sources.
Noise: The amount of non-802.11 traffic that is interfering with the currently assigned channel.
Coverage: The received signal strength (RSSI) and signal-to-noise ratio (SNR) for all connected clients.
Other: The number of nearby access points.

RRM performs these functions:

Radio resource monitoring
Transmit power control
Dynamic channel assignment
Coverage hole detection and correction

In this post we will see the configuration guide of RRM on WLC.

Configure an RF Group Name

Via GUI:

First step to configure RRM is to ensure WLC has the RF Group Name configured. This can be done through the controller web interface. Go to Controller > General and then type a RF Group Name value.

Via CLI:

Create an RF group by entering the config network rf-network-name name command:

(WLAN1) >config network rf-network-name mywlc

Configuring the RF Group Mode:

Via GUI

Go to Wireless > 802.11a/n or 802.11b/g/n > RRM > RF Grouping

Via CLI:

config advanced {802.11a | 802.11b} group-mode {auto | leader| off | restart}

(WLAN1) >config advanced 802.11a group-mode ?
 auto           Sets the 802.11a RF group selection to automatic update mode.
 leader         Sets the 802.11a RF group selection to static mode, and sets this controller as the group leader.
 off            Sets the 802.11a RF group selection off.
 restart        Restarts the 802.11a RF group selection.
(WLAN1) >config advanced 802.11a group-mode auto

On this screen we can see the details of RF group

Group Mode: Auto (It can be static or we can disable it)

Group Role: Auto Leader or Static Leader

Group Update Interval: The group update interval value indicates how often the RF Grouping algorithm is run and it cannot be modified.

Group Leader: This field displays the IP Address of the WLC that is currently the RF Group Leader.

Last Group Update: The RF Grouping algorithm runs every 600 seconds (10 minutes). This field indicates the time (in seconds) since the algorithm last ran.

*** A configured static leader cannot become a member of another controller until its mode is set to “auto”.

No we will change the Group mode on Controller”WLAN1” as leader.

Add a controller as member:

Via CLI:

Add a controller as a static member of the RF group (if the mode is set to “leader”) by entering this command:

config advanced {802.11a | 802.11b} group-mode {auto | leader| off | restart}

(WLAN1) >config advanced 802.11agroup-mode leader

config advanced 802.11a | 802. group-member add controller_name controller_ip_address

(WLAN1) >config advanced 802.11a | 802. group-member add WLAN2 10.35.80.3

To see RF grouping status

(WLAN1) >show advanced 802.11a group
 Radio RF Grouping
 802.11a Group Mode............................. STATIC
 802.11a Group Update Interval.................. 600 seconds
 802.11a Group Leader........................... WLAN1 (10.35.80.1)
 802.11a Group Member......................... WLAN1 (10.35.80.1)
 802.11a Group Member......................... WLAN2 (10.35.80.3)
 802.11a Last Run............................... 17 seconds ago
 * indicates member has not joined the group.
 (WLAN1) >

*** Same procedure for 802.11b network

***Info:

There are few things we must take care before forcing a WLC to be a RF leader:

All WLC members must have the same mobility and RF group name.
All WLCs AP must be in the range of each other.

In next post we will learn TPC, DCA and CHD.

Radio Resource Management

This feature is very importent to understand; Resource Management (RRM) to continuously monitor the RF environment. The controller uses the information from the access points (AP) and makes any changes to AP channels and power levels to try to mitigate such things as non-802.11 signal (noise), interference from other 802.11 devices, coverage gaps, and co-channel interference caused by the network.

Here is links to understand the RRM process:

Radio Resource Management Part 1
Radio Resource Management Part 2

The RRM feature, also known as Auto-RF, uses the RF information gathered by the APs to make decisions on whether channel assignment or power levels need to be adjusted.

RRM enables controller to monitor their associated AP for the following information.

RRM Analyze	RRM Performs
Traffic Load Interference Noise Coverage Others AP	Radio resource monitoring Transmit power level Dynamic channel assignment Coverage hole detection and correction

RRM Key Figures

RRM Neighbor messages is sent

At lowest mandatory speed, max power(standard according to country regulations)
Every 60 seconds by default
On all serviced channels

Important points when dealing with RRM

The controllers elect the RF group leader.
The RF group leader is responsible for dynamic channel assignment (DCA) and transmits power control (TPC).
An individual controller handles coverage hole detection and correction.
RF groups and mobility groups are independent functions.
RF grouping is per radio. The RF group leader for the 802.11b/g network might not be the same RF group leader for the 802.11a network.
With code Release 4.2.99 or higher, RRM supports up to 20 controllers and 1000 APs in a single RF group.
RF fluctuations can cause the RF group leader to change.
By default, the RF group leader polls the other controllers in the RF group for AP statistics and neighbor messages.
The transmit power threshold setting should be the same between all controllers in the same RF group, because you do not want an entire network to start fluctuating because of a group leader change.
Each AP maintain a list of up to 34 RRM neighbors per radio
Controller keeps the best 24 (per radio)
Controller forwards the list of 24 to the RF-Leader
Neighbor AP must be -80dBm or better to be on the list
An AP on the list gets dropped if its signal falls below -85dBm
Group leader tries to maintain -70dBm threshold between APs

Basic workflow:

The controllers and their APs use the configured RF group name to determine if other APs they hear are part of their RF group.
The APs use neighbor messages (sent every 60 seconds) that are authenticated by other APs that hear them. The neighbor messages include information about the AP, the controller, and the configured RF group name.
The APs that hear the neighbor message of another AP authenticate that message using the RF group name and pass it to their respective controller.
The controllers use this information to determine what other controllers should be in their RF group, and then form logical groups to share the RF information from their respective APs, and elect an RF group leader.
The RF group leader runs the RRM algorithm against the RF information from all the APs in the RF group. Depending on the outcome, a power level or channel change for an AP or group of APs might take place.

RRM Inter-controller communication

Controller must be part of the same mobility group
RF group is a subset of mobility group
Max limits 20 controllers or 1000 AP per RF-Groups
Group Leader elected automatically based on controller id value and highest MAC address
Updates group member every 600 seconds for channels and power values
For 802.11b/g ports 12124 and 12134 are used
For 802.11a ports 12125 and 12135 are used.

RRM message contains:

Radio ID: If the AP had multiple radios, this field identifies the radio used to transmit the message.
Group ID: The 16-bit value and controller MAC address.
Management IP address of the controller (if OATP enabled): RF group leader’s management IP address.
Channel Count (unused)
Antenna Pattern: The antenna pattern currently in use.
Measurement Interval
Key
Channel: The native channel that the AP uses to service clients.
Power (Unused)

When controllers learn of another controller from the AP neighbor messages, they communicate directly with one another to form a system-wide RF group. After the system-wide RF group is formed, the controllers elect an RF group leader.

The group leader is the controller with the highest group leader priority. The group leader priority is based on the group identifier (group ID) information element (IE) contained in the AP neighbor messages. Every controller maintains a 16-bit counter that starts at 0 and increments following events like adding or leaving an RF group or the controller being rebooted. This counter value and the MAC address of the controller make up the group ID IE. Every controller in the RF group selects one controller, or itself, that has the highest group ID value and compares this to the selected controller from the other controllers in the group. The single controller with the highest group ID is elected the RF group leader.

If the RF group leader goes offline, the entire RF group breaks up and the RF grouping process and election of an RF group leader starts over.

For an RF group to form, it takes only one AP on one controller to hear another AP on a different controller.