WiFi Access Point Interference Avoidance via Predictive ML
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
WiFi access points face interference from LTE-U and LTE-LAA systems in unlicensed bands, leading to performance degradation, reduced channel availability, and decreased network capacity due to inefficient interference avoidance methods that result in wasted spectrum and reduced airtime efficiency.
Innovation Solution
A predictive machine-learning process is implemented by WiFi access points to anticipate LTE interference, allowing proactive avoidance through techniques such as preamble puncturing or channel switching, thereby minimizing service interruptions and maintaining optimal network performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If WiFi access points completely avoid channels with LTE-U and/or LTE-LAA interference, then interference from second wireless technology is eliminated, but WiFi channel availability is reduced and spectrum overlap increases
Solution Approach 1:
The system performs preliminary detection of LTE-U/LTE-LAA signals and predicts their upcoming activity patterns using machine learning. This allows WiFi access points to proactively switch to alternative channels or adjust transmission parameters before interference occurs, rather than reactively avoiding all channels with historical LTE activity. The preliminary action enables selective avoidance only when and where interference is predicted, maintaining channel availability when LTE is not active.
Solution Approach 2:
The interference avoidance mechanism dynamically adapts channel selection and transmission parameters based on real-time LTE detection and prediction results. Instead of static channel avoidance, the system continuously adjusts WiFi channel usage, bandwidth, and power levels in response to predicted LTE activity patterns. This dynamic approach optimizes the balance between avoiding interference and maximizing channel utilization efficiency.
2Productivity
If WiFi access points use wider bandwidths such as 80MHz or 160MHz, then network capacity is improved, but the number of alternative channels available to select from is limited
Solution Approach 1:
The machine learning system predicts LTE-U/LTE-LAA activity patterns in advance, enabling WiFi access points to proactively select and switch to alternative wide bandwidth channels before interference occurs. By predicting which channels will be interfered with and when, the system can pre-position WiFi transmissions on alternative 80MHz or 160MHz channels, ensuring high network capacity is maintained without being constrained by the limited number of wide bandwidth options.
Solution Approach 2:
The system dynamically changes transmission parameters including channel bandwidth, center frequency, and power levels based on predicted LTE activity. When LTE interference is predicted on a current wide bandwidth channel, the system adjusts to alternative bandwidth configurations or frequency positions, optimizing the balance between maintaining high network capacity through wide bandwidths and adapting to available spectrum conditions.
3Productivity
If LTE systems use unlicensed spectrum in an on-demand basis without Listen before Talk, then LTE channel access frequency is improved, but WiFi air time efficiency is reduced
Solution Approach 1:
The system performs preliminary detection and prediction of LTE-U/LTE-LAA transmission patterns using machine learning analysis of energy detection data and signal characteristics. By predicting when LTE will transmit in advance, the system can proactively schedule WiFi transmissions during predicted LTE idle periods or switch to alternative channels, thereby reducing WiFi air time loss caused by LTE's on-demand access without LBT.
Solution Approach 2:
The system continuously monitors energy levels and LTE signal characteristics on WiFi channels, using this feedback to refine machine learning predictions of LTE activity patterns. This feedback loop enables increasingly accurate prediction of LTE transmission timing, allowing WiFi access points to optimize their transmission schedules and channel selection to minimize air time efficiency losses while accommodating LTE's aggressive channel access behavior.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems, methods, and computer-readable media are provided for predicting presence of interfering signals on given wireless channel(s) on which an access point is operating and proactively implementing an interference avoidance process. In one aspect of the present disclosure, a method includes determining, by an access point operating according to a first wireless technology and at a first time, that interference from signals of a second technology will occur, at a second time that is later than the first time, on a channel on which the access point is currently operating; selecting, by the access point, an interference avoidance process based on a plurality of factors; and implementing, by the access point, the interference avoidance process such that at the second time the access point is not operating on one or more sub-channels spanning the channel on which the access point is currently operating.