Abstract:Resource allocation and transceivers in wireless networks are usually designed by solving optimization problems subject to specific constraints, which can be formulated as variable or functional optimization. If the objective and constraint functions of a variable optimization problem can be derived, standard numerical algorithms can be applied for finding the optimal solution, which however incurs high computational cost when the dimension of the variable is high. To reduce the on-line computational complexity, learning the optimal solution as a function of the environment's status by deep neural networks (DNNs) is an effective approach. DNNs can be trained under the supervision of optimal solutions, which however, is not applicable to the scenarios without models or for functional optimization where the optimal solutions are hard to obtain. If the objective and constraint functions are unavailable, reinforcement learning can be applied to find the solution of a functional optimization problem, which is however not tailored to optimization problems in wireless networks. In this article, we introduce unsupervised and reinforced-unsupervised learning frameworks for solving both variable and functional optimization problems without the supervision of the optimal solutions. When the mathematical model of the environment is completely known and the distribution of the environment's status is known or unknown, we can invoke an unsupervised learning algorithm. When the mathematical model of the environment is incomplete, we introduce reinforced- unsupervised learning algorithms that learn the model by interacting with the environment. Our simulation results confirm the applicability of these learning frameworks by taking a user association problem as an example.

Unsupervised Learning for Asynchronous Resource Allocation in Ad-hoc Wireless Networks

Aggregation Design and Resource Allocation for Decentralized Learning over Wireless Networks.

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

Unsupervised Resource Allocation with Graph Neural Networks

Asynchronous Algorithms for Decentralized Resource Allocation over Directed Networks

Unsupervised Learning Neural-Network Method for Resource Allocation in Multi-cell Cellular Networks

Efficient Collaborative Learning over Unreliable D2D Network: Adaptive Cluster Head Selection and Resource Allocation

Knowledge-Driven Resource Allocation for D2D Networks: A WMMSE Unrolled Graph Neural Network Approach

Improving Message-Passing GNNs by Asynchronous Aggregation

Message Passing Assisted Scalable Distributed Link Management for Ubiquitous Network

Optimizing Ultra-Reliable and Low-Latency Communication Systems with Unsupervised Learning.

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

Interference-Limited Ultra-Reliable and Low-Latency Communications: Graph Neural Networks or Stochastic Geometry?

Unsupervised Graph-based Learning Method for Sub-band Allocation in 6G Subnetworks

Sparse Graph Neural Networks for Two-Timescale Wireless Resource Allocation.

Straggler-Resilient Decentralized Learning via Adaptive Asynchronous Updates

Transfer Learning for Mixed-Integer Resource Allocation Problems in Wireless Networks

Optimizing Wireless Systems Using Unsupervised and Reinforced-Unsupervised Deep Learning

Unsupervised Deep Learning For Ultra-Reliable And Low-Latency Communications

Unsupervised Power Allocation Based on Combination of Edge Aggregated Graph Attention Network with Deep Unfolded WMMSE

Learning Power Allocation for Multi-Cell-Multi-User Systems with Heterogeneous Graph Neural Networks.