Unfolding WMMSE using Graph Neural Networks for Efficient Power Allocation

Chowdhury, Arindam; Verma, Gunjan; Rao, Chirag; Swami, Ananthram; Segarra, Santiago

doi:10.48550/arxiv.2009.10812

Cited by 8 publications

(21 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…5(c) and 5(g) present the influence of the scale parameter γ [cf. (2)]. Combining the results in Fig.…”

Section: Parameter Sensitivitymentioning

confidence: 74%

“…We set b to the 70th percentile of the entries in the FE distance matrix ∆ FE η so that 70% of entries in the similarity matrix S FE η,b,γ will be positive [cf. (2)]. We set γ to make the largest entry in S FE η,b,γ equal to 6.…”

Section: Baseline Methodsmentioning

confidence: 99%

“…The produced low-dimensional embeddings can be used in several network analysis tasks such as node classification, link prediction, community detection, and visualization. They can also be used as the input to downstream graph neural networks with applications in varied fields including wireless communications [2], [3], traffic prediction [4], [5], and neuroscience [6], [7]. In this paper, we focus on the popular subclass of unsupervised node embedding, where only the graph structure is given and no extra information about the nodes is available.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Free Energy Node Embedding via Generalized Skip-gram with Negative Sampling

Zhu¹,

Swami²,

Segarra³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

A widely established set of unsupervised node embedding methods can be interpreted as consisting of two distinctive steps: i) the definition of a similarity matrix based on the graph of interest followed by ii) an explicit or implicit factorization of such matrix. Inspired by this viewpoint, we propose improvements in both steps of the framework. On the one hand, we propose to encode node similarities based on the free energy distance, which interpolates between the shortest path and the commute time distances, thus, providing an additional degree of flexibility. On the other hand, we propose a matrix factorization method based on a loss function that generalizes that of the skip-gram model with negative sampling to arbitrary similarity matrices. Compared with factorizations based on the widely used 2 loss, the proposed method can better preserve node pairs associated with higher similarity scores. Moreover, it can be easily implemented using advanced automatic differentiation toolkits and computed efficiently by leveraging GPU resources. Node clustering, node classification, and link prediction experiments on real-world datasets demonstrate the effectiveness of incorporating free-energy-based similarities as well as the proposed matrix factorization compared with state-of-the-art alternatives.

show abstract

“…5(c) and 5(g) present the influence of the scale parameter γ [cf. (2)]. Combining the results in Fig.…”

Section: Parameter Sensitivitymentioning

confidence: 74%

Section: Baseline Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Free Energy Node Embedding via Generalized Skip-gram with Negative Sampling

Zhu¹,

Swami²,

Segarra³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…A fundamental RRM problem is the optimization of transmission power levels at distributed links that share the same spectral resources in the presence of time-varying channel conditions [8]. This problem was addressed by the data-driven methodology introduced in [9], and later studied in [10]- [12]. In it, the power control policy mapping channel state information (CSI) and power vector is parametrized by a graph neural network (GNN).…”

Section: A Motivationmentioning

confidence: 99%

“…GNNs are enjoying an increasing popularity in the wireless communication community. In addition to power allocation [9]- [12], GNNs have been used to address cellular [25] and satellite [26] traffic prediction, link scheduling [27], channel control [28], and localization [29]. Due to their localized nature, GNNs have also been applied to cooperative [30] and decentralized [31] control problems in networked systems.…”

Section: Prior Workmentioning

confidence: 99%

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

Nikoloska¹,

Simeone²

2021

Preprint

View full text Add to dashboard Cite

In this paper, we consider the problem of power control for a wireless network with an arbitrarily time-varying topology, including the possible addition or removal of nodes. A data-driven design methodology that leverages graph neural networks (GNNs) is adopted in order to efficiently parametrize the power control policy mapping the channel state information (CSI) to transmit powers. The specific GNN architecture, known as random edge GNN (REGNN), defines a non-linear graph convolutional filter whose spatial weights are tied to the channel coefficients. While prior work assumed a joint training approach whereby the REGNN-based policy is shared across all topologies, this paper targets adaptation of the power control policy based on limited CSI data regarding the current topology. To this end, we propose both black-box and modular meta-learning techniques. Black-box meta-learning optimizes a general-purpose adaptation procedure via (stochastic) gradient descent, while modular meta-learning finds a set of reusable modules that can form components of a solution for any new network topology.Numerical results validate the benefits of meta-learning for power control problems over joint training schemes, and demonstrate the advantages of modular meta-learning when data availability is extremely limited.

show abstract

Distributed Scheduling Using Graph Neural Networks

Zhao

Verma²,

Rao³

et al. 2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

A fundamental problem in the design of wireless networks is to efficiently schedule transmission in a distributed manner. The main challenge stems from the fact that optimal link scheduling involves solving a maximum weighted independent set (MWIS) problem, which is NP-hard. For practical link scheduling schemes, distributed greedy approaches are commonly used to approximate the solution of the MWIS problem. However, these greedy schemes mostly ignore important topological information of the wireless networks. To overcome this limitation, we propose a distributed MWIS solver based on graph convolutional networks (GCNs). In a nutshell, a trainable GCN module learns topology-aware node embeddings that are combined with the network weights before calling a greedy solver. In small-to middle-sized wireless networks with tens of links, even a shallow GCN-based MWIS scheduler can leverage the topological information of the graph to reduce in half the suboptimality gap of the distributed greedy solver with good generalizability across graphs and minimal increase in complexity.

show abstract

Unfolding WMMSE using Graph Neural Networks for Efficient Power Allocation

Cited by 8 publications

References 39 publications

Free Energy Node Embedding via Generalized Skip-gram with Negative Sampling

Free Energy Node Embedding via Generalized Skip-gram with Negative Sampling

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

Distributed Scheduling Using Graph Neural Networks

Contact Info

Product

Resources

About