https://secure.leadforensics.com/js/130070.js //px.spiceworks.com/px.js
Technology

Efficient Client Load Balancing with Samsung WLAN Infrastructure

By March 20, 2018 No Comments

As part of high density Wi-Fi deployments one of the issues that Wi-Fi engineers try to solve via design is the even distribution of stations across multiple access points. The ideal solution is for every AP to be evenly loaded and the nearest client (ideally) connected to the nearest AP. But this is almost never the case as most client load balancing schemes rely on both the WLAN infrastructure and the clients. Nevertheless this should not stop network managers from implementing these schemes because at the least they provide an opportunity to better our design. The purpose of this post is to explain the tools available for engineers who want to implement this feature within a Samsung Wi-Fi infrastructure.

The goals of the station load-balancing feature follow:

  1. Distribute as evenly as possible across multiple APs as the number of stations exceed a threshold value on an AP.
  2. Distribute the stations based on RSSI triggers (low/high).

Station Count Based Load Balancing

There are two variables involved in count-based load balancing: the load-balancing threshold and the max denial count. For those who are more technically inclined, the algorithm below shows how this is implemented.

If ( (current number of stations + a new station attempts to associate < load balancing threshold) or (current denial count + a new station attempts to associate > max denial count) )

{allow the station to be association denial count = 0}

Else

{send association response (Status Code 17) denial count++}

So, ideally, if an AP is above the station association threshold every new association will be sent an association response with status code 17.

 

 

The idea here is that the devices that comply to this response will then start looking for other APs in the candidate table to connect to. As the algorithm shows, no station is denied access.


 

In the diagram above a potential station is deemed sticky. (Its measured signal strength is lower than the RSSI_low value.) Once the AP is overloaded the following process occurs.

  1. The AP sends a load-balancing request message to the controller with a potential load-balancing operation station’s list.
  2. The Access Point Controller (APC) selects a target AP from the criteria which consist of the number of stations, RSSI_high, and RSSI_low factors.  Keep in mind the controller knows the neighboring APs and is also aware of the load these APs are carrying based on RRM.
  3. The APC sends a no_probe command to neighboring APs other than the target AP.  This instructs the APs not to respond to probe requests coming from the stations in the load-balancing operation list.
  4. The APC sends a kick out command to the requesting AP and this AP will send the disassociation frame to candidate STA(s).
  5. The kicked-out STA sends a probe request for scanning.
  6. Neighboring APs other than the target AP do not send a probe response and eventually associate to the target AP because of the target AP’s response.

Station-based adaptive load balancing can be applied to each AP group and specific APs separately.

It supports two kinds of operational modes:

  1) Passive mode: default mode.

   An AP sends the load-balancing request message to an APC with the candidate stations list.

  2) Active mode : optional mode.

   The controller actively sends a load-balancing message to the AP. The AP then responds with the candidate stations list.

Learn more about Samsung’s WLAN Controllers.

VISIT THE PRODUCT PAGEVISIT THE PRODUCT PAGE

stay connected.

Sign up and be the first to see our latest technology innovations.

Email Address

First Name

Last Name

stay connected.

Sign up and be the first to see our latest technology innovations.

Email Address

First Name

Last Name