Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Graph Representation Learning for Contention and Interference Management in Wireless Networks

Zhouyou Gu, Branka Vucetic, Kishore Chikkam, Pasquale Aliberti, Wibowo Hardjawana

TL;DR

Simulations show that the proposed methods achieve higher worst-case user throughput than the existing approaches and that the proposed architecture can further improve the worst-case user throughput by $5\%\sim 30$ while ensuring timely updates of grouping decisions.

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.

Graph Representation Learning for Contention and Interference Management in Wireless Networks

TL;DR

Simulations show that the proposed methods achieve higher worst-case user throughput than the existing approaches and that the proposed architecture can further improve the worst-case user throughput by while ensuring timely updates of grouping decisions.

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 higher worst-case user throughput than the existing approaches and that the proposed architecture can further improve the worst-case user throughput by while ensuring timely updates of grouping decisions.
Paper Structure (38 sections, 1 theorem, 48 equations, 13 figures, 1 table, 2 algorithms)

This paper contains 38 sections, 1 theorem, 48 equations, 13 figures, 1 table, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Works
  3. Markov-model-based Approach
  4. Graph-based Approach
  5. Machine Learning Approach
  6. Our Contributions
  7. Notation
  8. System Model and Problem Formulation
  9. Configurations of the Wireless Network
  10. Network States
  11. User Grouping Problem in RAW
  12. Graph Representation Optimization for User Grouping in Wireless Networks
  13. Actor-Critic Graph Representation Learning
  14. Design of the Graph-Constructing Actor
  15. Design of the Graph Cut Procedure
  16. ...and 23 more sections

Key Result

Lemma 1

Define the optimal edge weights as $W^*_{i,j}$, $\forall i\neq j$, that maximize the network performance objective in eq:prob:user_grouping:graph_cut:adaptive_edge_weighting, and define $\mathbf{z}^*$ as the optimal solution of the LLP of eq:prob:user_grouping:graph_cut:adaptive_edge_weighting for t

Figures (13)

  • Figure 1: Illustration of a wireless network.
  • Figure 2: Illustration of user grouping in RAW, where $K=3$, $Z=2$, $z_1=1$ and $z_2=z_3=2$.
  • Figure 3: Overall structure of the AC-GRL algorithm.
  • Figure 4: The structure of the actor.
  • Figure 5: Tree diagram illustrating the recursive graph cut when $Z=4$ and $\beta=0,1,\dots,\log_2(Z)$.
  • ...and 8 more figures

Theorems & Definitions (3)

  • Lemma 1
  • proof
  • proof