The Artificial Intelligence Machine Learning Standing Committee is tasked with exploring AI/ML use cases that will apply to Wi-Fi systems and devices
Artificial intelligence (AI) and machine learning (ML) are being applied everywhere and to everything; and according to Purva Rajkotia, who serves as the lead of IEEE SA’s Connectivity and Telecom Practice, Wi-Fi is no different.
In fact, there is a new committee within the IEEE 802.11 working group called the Artificial Intelligence Machine Learning (AIML) Standing Committee (SC) tasked with exploring AI/ML use cases that will apply to IEEE 802.11 systems and devices. It also looks at the technical feasibility of these use cases.
While it remains unclear how exactly and to what extent AI/ML will be used in 802.11 systems, Rajkotia detailed for RCR Wireless News four ways these advanced technologies are already being used in Wi-Fi networks.
Channel access control parameter selection
Today, AI/ML algorithms are used to assist with transmission channel selection. These algorithms can use data from previous transmissions to help decide the right channel parameters based on things like user history, channel conditions, the time, the day and the environment. “It’s all going to play a role in determining the right channel selection and how the link should be adapted,” said Rajkotia.
Multi-user parameter selection
With 802.11, we have moved from single user to a mutli-user functionality with capabilities like multi-user MIMO (MU-MIMO). MU-MIMO was introduced in Wi-Fi 6 and allows more data to be transferred at one time, enabling access points (APs) to handle larger numbers of devices simultaneously. Wi-Fi 7 expands on multi-user capabilities with Multi-link operation (MLO), which makes it possible for a client device to talk to an AP over multiple radios and frequency bands at the same time, therefore, sending data simultaneously over two radios simultaneously.
Rajkotia said that AI/ML mechanisms can be used in multi-user scenarios to better optimize network performance by improving network traffic management, which will more efficiently distribute bandwidth and minimize latency.
Determining the contention window size
The contention window (CW), or a time period within a wireless network in which a device waits randomly before attempting to transmit data, in 802.11 dynamically adjusts based on the network conditions, increasing it during times of high and decreasing it when fewer devices are transmitting on the channel.
AI/ML can help improve this process, as well, said Rajkotia, but he added: “Here, you want to make sure that you use the right AI/ML algorithm and technique to determine the contention window sizes, so that we can increase the overall capacity and provide a better user experience.”
Improving channel usage
By dynamically selecting the least congested channels on both the 2.4 GHz and 5 GHz bands in real-time, AI/ML can minimize network interference and optimize connection speeds. Put simply, AI/ML analyzes surrounding network activity and decides how each channel should be used for optimal performance and user experience.
“These are some of the areas in which we are already using AI/ML technology,” concluded Rajkotia. “That being said, there is also a lot of working being done around AI/ML systems because we need to determine the emerging use cases for which this technology is going to be applicable for the 802.11 systems and we also want to understand the technical ability of certain features to support AI/ML technology.”
Table of Contents
