Modular Meta-Learning for Power Control via Random Edge Graph Neural Networks

Nikoloska, Ivana; Simeone, Osvaldo

doi:10.48550/arxiv.2108.13178

Cited by 4 publications

(5 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With the success of machine learning, and particularly deep learning, over the past few years, learning-based algorithms have emerged to solve challenging problems in wireless communications, including for resource management [13]. As a prominent example, for the class of power allocation problems, several approaches have been proposed that leverage techniques based on supervised, unsupervised, self-supervised, and reinforcement learning, as well as meta-learning and graph representation learning [14]- [29].…”

Section: Introductionmentioning

confidence: 99%

Learning Resilient Radio Resource Management Policies with Graph Neural Networks

Naderializadeh¹,

Eisen²,

Ribeiro³

2022

Preprint

View full text Add to dashboard Cite

We consider the problems of downlink user selection and power control in wireless networks, comprising multiple transmitters and receivers communicating with each other over a shared wireless medium. To achieve a high aggregate rate, while ensuring fairness across all the receivers, we formulate a resilient radio resource management (RRM) policy optimization problem with per-user minimum-capacity constraints that adapt to the underlying network conditions via learnable slack variables. We reformulate the problem in the Lagrangian dual domain, and show that we can parameterize the user selection and power control policies using a finite set of parameters, which can be trained alongside the slack and dual variables via an unsupervised primal-dual approach thanks to a provably small duality gap. We use a scalable and permutation-equivariant graph neural network (GNN) architecture to parameterize the RRM policies based on a graph topology derived from the instantaneous channel conditions. Through experimental results, we verify that the minimum-capacity constraints adapt to the underlying network configurations and channel conditions. We further demonstrate that, thanks to such adaptation, our proposed method achieves a superior tradeoff between the average rate and the 5 th percentile rate-a metric that quantifies the level of fairness in the resource allocation decisions-as compared to baseline algorithms.

show abstract

Section: Introductionmentioning

confidence: 99%

Learning Resilient Radio Resource Management Policies with Graph Neural Networks

Naderializadeh¹,

Eisen²,

Ribeiro³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…This is, however, not the case for conventional learning in slow-varying environments, for which the features tend to be too correlated, resulting in overfitting. As a final note, although absolute NMSE values close to 1 may be insufficient for use in applications such as precoding, they can provide useful information for other applications such as proactive resource allocation [ 40 , 64 ].…”

Section: Methodsmentioning

confidence: 99%

“…Previous applications of transfer learning to communication systems include beamforming for multi-user, multiple-input, single-output (MISO) downlink [ 25 ] and for intelligent reflecting surfaces (IRS)-assisted MISO downlink [ 26 ], and downlink channel prediction [ 27 , 28 ] (see also [ 25 , 27 ]). Meta-learning has been applied to communication systems, including demodulation [ 29 , 30 , 31 , 32 ], decoding [ 33 ], end-to-end design of encoding and decoding with and without a channel model [ 34 , 35 ]; MIMO detection [ 36 ], beamforming for multiuser MISO downlink systems via [ 37 ], layered division multiplexing for ultra-reliable communications [ 38 ], UAV trajectory design [ 39 ], and resource allocation [ 40 ].…”

Section: Introductionmentioning

confidence: 99%

Speeding up Training of Linear Predictors for Multi-Antenna Frequency-Selective Channels via Meta-Learning

Park

Simeone

2022

Entropy

View full text Add to dashboard Cite

An efficient data-driven prediction strategy for multi-antenna frequency-selective channels must operate based on a small number of pilot symbols. This paper proposes novel channel-prediction algorithms that address this goal by integrating transfer and meta-learning with a reduced-rank parametrization of the channel. The proposed methods optimize linear predictors by utilizing data from previous frames, which are generally characterized by distinct propagation characteristics, in order to enable fast training on the time slots of the current frame. The proposed predictors rely on a novel long short-term decomposition (LSTD) of the linear prediction model that leverages the disaggregation of the channel into long-term space-time signatures and fading amplitudes. We first develop predictors for single-antenna frequency-flat channels based on transfer/meta-learned quadratic regularization. Then, we introduce transfer and meta-learning algorithms for LSTD-based prediction models that build on equilibrium propagation (EP) and alternating least squares (ALS). Numerical results under the 3GPP 5G standard channel model demonstrate the impact of transfer and meta-learning on reducing the number of pilots for channel prediction, as well as the merits of the proposed LSTD parametrization.

show abstract

“…It is clear that as the network size grows, generating high-quality labeled samples becomes exponentially more costly. This computational complexity has led to alternatives to supervised learning for training deep learning models in RRM problems, including unsupervised learning, reinforcement learning, self-supervised learning, and meta-learning, which do not necessarily rely on (extensive) labeling of the data for training the underlying neural networks [3][4][5][9][10][11][12][13][14][15][16]. However, except for a few recent studies, such as [3,8,11], little effort has been made to thoroughly compare the performance of models trained using supervised and unsupervised learning procedures.…”

Section: Introductionmentioning

confidence: 99%

Wireless Link Scheduling via Graph Representation Learning: A Comparative Study of Different Supervision Levels

Naderializadeh¹

2021

Preprint

View full text Add to dashboard Cite

We consider the problem of binary power control, or link scheduling, in wireless interference networks, where the power control policy is trained using graph representation learning. We leverage the interference graph of the wireless network as an underlying topology for a graph neural network (GNN) backbone, which converts the channel matrix to a set of node embeddings for all transmitter-receiver pairs. We show how the node embeddings can be trained in several ways, including via supervised, unsupervised, and selfsupervised learning, and we compare the impact of different supervision levels on the performance of these methods in terms of the system-level throughput, convergence behavior, sample efficiency, and generalization capability. 1

show abstract

Modular Meta-Learning for Power Control via Random Edge Graph Neural Networks

Cited by 4 publications

References 33 publications

Learning Resilient Radio Resource Management Policies with Graph Neural Networks

Learning Resilient Radio Resource Management Policies with Graph Neural Networks

Speeding up Training of Linear Predictors for Multi-Antenna Frequency-Selective Channels via Meta-Learning

Wireless Link Scheduling via Graph Representation Learning: A Comparative Study of Different Supervision Levels

Contact Info

Product

Resources

About