Abstract:Restricted access window (RAW) in Wi-Fi 802.11ah networks manages contention and interference by grouping users and allocating periodic time slots for each group's transmissions. We will find the optimal user grouping decisions in RAW to maximize the network's worst-case user throughput. We review existing user grouping approaches and highlight their performance limitations in the above problem. We propose formulating user grouping as a graph construction problem where vertices represent users and edge weights indicate the contention and interference. This formulation leverages the graph's max cut to group users and optimizes edge weights to construct the optimal graph whose max cut yields the optimal grouping decisions. To achieve this optimal graph construction, we design an actor-critic graph representation learning (AC-GRL) algorithm. Specifically, the actor neural network (NN) is trained to estimate the optimal graph's edge weights using path losses between users and access points. A graph cut procedure uses semidefinite programming to solve the max cut efficiently and return the grouping decisions for the given weights. The critic NN approximates user throughput achieved by the above-returned decisions and is used to improve the actor. Additionally, we present an architecture that uses the online-measured throughput and path losses to fine-tune the decisions in response to changes in user populations and their locations. Simulations show that our methods achieve $30\%\sim80\%$ higher worst-case user throughput than the existing approaches and that the proposed architecture can further improve the worst-case user throughput by $5\%\sim30\%$ while ensuring timely updates of grouping decisions.

What problem does this paper attempt to address?

### Problems Addressed by the Paper This paper aims to address the issue of managing and reducing competition and interference among users in Wi-Fi 802.11ah networks by optimizing user grouping, thereby maximizing the throughput of the worst-performing user in the network. Specifically, the paper focuses on finding the optimal user grouping decisions under the Restricted Access Window (RAW) mechanism to avoid low throughput for the worst user, which could lead to service interruptions. ### Background and Problem Description 1. **Network Background**: - The Wi-Fi 802.11ah (also known as Wi-Fi HaLow) standard provides low-power, long-distance wireless connectivity. - In such networks, users access the channel using the Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) mechanism. - When users detect that other users are transmitting data, they wait for the channel to be idle and further randomly back off for a period to avoid packet collisions. 2. **Existing Problems**: - **Contention Issue**: When there are many other users around, users experience significant waiting and backoff times, making data transmission difficult. - **Hidden Terminal Problem**: When two users are far apart and cannot detect each other's transmissions, simultaneous transmissions can occur, causing interference at the receiving access point (AP), reducing the Signal-to-Interference-plus-Noise Ratio (SINR), and making it difficult for the AP to decode the data. - These issues lead to low throughput for the worst-performing user, affecting applications that require continuous service, such as telemedicine and environmental disaster monitoring. 3. **Limitations of Existing Methods**: - **Markov Model-Based Methods**: These can model the channel access process of users but do not consider interference caused by hidden users. - **Graph Theory-Based Methods**: These represent user competition and interference through graph construction but use fixed rules, lacking flexibility in optimization. - **Machine Learning-Based Methods**: These can use neural networks (NN) to generate user grouping decisions but lack retention of interference information between individual users. ### Solution 1. **Graph Representation Learning Framework**: - Transform the user grouping problem into a graph construction problem, where the edge weights of the graph represent competition and interference between user pairs. - Use the Max Cut algorithm to group users, maximizing the elimination of competition and interference between different groups. 2. **Actor-Critic Graph Representation Learning Algorithm (AC-GRL)**: - **Actor NN**: Train a neural network to estimate the optimal edge weights of the graph, using path loss between users and the AP as input. - **Critic NN**: Evaluate user throughput under given edge weights and use this to optimize the Actor. - **Online Adjustment Architecture**: Use online measurements of throughput and path loss to fine-tune decisions, adapting to changes in user locations. ### Experimental Results - Simulation results show that the proposed method improves the worst user throughput by 30%~80% compared to existing methods. - The online adjustment architecture can further improve the worst user throughput by 5%~30%. - When users move, adjusting the graph structure online can improve the worst user throughput by about 200%. ### Conclusion By introducing a graph representation learning framework and an Actor-Critic algorithm, the paper successfully addresses the user grouping problem in Wi-Fi 802.11ah networks, significantly improving the throughput of the worst-performing user and ensuring network stability and service quality.

Graph Representation Learning for Contention and Interference Management in Wireless Networks

Graph Representation Learning for Contention and Interference Management in Wireless Networks

Implementing Graph Neural Networks over Wireless Networks Via Over-the-Air Computing: A Joint Communication and Computation Framework

Interference Mitigation In Wireless Mesh Networks Through Radio Co-location Aware Conflict Graphs

Conflict Graph Embedding for Wireless Network Optimization

Learning to Optimize Resource in Dynamic Wireless Environment Via Meta-Gating Graph Neural Network

A Graph Neural Network Approach for Scalable Wireless Power Control

Graph Representation Learning for Wireless Communications

Learning the Interference Graph of a Wireless Network

Resource Allocation for Ultradense Networks with Machine-Learning-Based Interference Graph Construction.

Reinforcement Learning-based Wi-Fi Contention Window Optimization

Cooperative Beamforming and RISs Association for Multi-RISs Aided Multi-Users MmWave MIMO Systems through Graph Neural Networks

Graph Reinforcement Learning for Radio Resource Allocation

Decentralized Routing and Radio Resource Allocation in Wireless Ad Hoc Networks via Graph Reinforcement Learning

Deep Reinforcement Learning Based Contention Window Optimization for IEEE 802.11 Bn

Deep Reinforcement Learning for Optimizing Restricted Access Window in IEEE 802.11ah MAC Layer

Adaptive Contention Window Design using Deep Q-learning

AI Empowered Resource Management for Future Wireless Networks

Graph Neural Networks for Scalable Radio Resource Management: Architecture Design and Theoretical Analysis

Graph Neural Networks for Wireless Networks: Graph Representation, Architecture and Evaluation

Knowledge-Driven Resource Allocation for Wireless Networks: A WMMSE Unrolled Graph Neural Network Approach