Graph Neural Networks for Decentralized Controllers

Gama, Fernando; Tolstaya, Ekaterina V.; Ribeiro, Alejandro

doi:10.48550/arxiv.2003.10280

Cited by 5 publications

(11 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In these works, GNNs are generally used in the centralized manner. Recently, some works have used GNNs for decentralized control [31], [32], [33]. However, they generally assume perfect information exchanges between nodes.…”

Section: Related Work In Wireless Communication Fieldmentioning

confidence: 99%

“…v , node v requests each neighbor u with vu > U v to retransmit their node features. By using the MRC technique, node v coherently combines the received copies for each retransmitted signal, computes their corresponding effective SNR as (33), and updates the estimated sliced feature matrix Xv . Then, node v re-predicts its label ĉv and re-computes the robustness probability p (r) v .…”

Section: Retransmission In Uncoded Communication Systemsmentioning

confidence: 99%

“…Obviously, not all received signals but only those in outage need to be retransmitted. In other words, if log(1 + γ vu ) < R u , node u needs to retransmit x u to node v. By using the MRC technique, all received copies of x u are coherently combined and the effective SNR is given by (33), which is used to check whether the effective received signal is in outage. Specifically, after T transmissions, if log(1 + γ vu (T )) ≥ R u , it indicates that xvu (T ) = x v and q vu (T ) = 0.…”

Section: Retransmission In Coded Communication Systemsmentioning

confidence: 99%

See 2 more Smart Citations

Decentralized Inference with Graph Neural Networks in Wireless Communication Systems

Lee¹,

Yu²,

Dai³

2021

Preprint

View full text Add to dashboard Cite

Graph neural network (GNN) is an efficient neural network model for graph data and is widely used in different fields, including wireless communications. Different from other neural network models, GNN can be implemented in a decentralized manner with information exchanges among neighbors, making it a potentially powerful tool for decentralized control in wireless communication systems. The main bottleneck, however, is wireless channel impairments that deteriorate the prediction robustness of GNN. To overcome this obstacle, we analyze and enhance the robustness of the decentralized GNN in different wireless communication systems in this paper. Specifically, using a GNN binary classifier as an example, we first develop a methodology to verify whether the predictions are robust. Then, we analyze the performance of the decentralized GNN binary classifier in both uncoded and coded wireless communication systems. To remedy imperfect wireless transmission and enhance the prediction robustness, we further propose novel retransmission mechanisms for the above two communication systems, respectively. Through simulations on the synthetic graph data, we validate our analysis, verify the effectiveness of the proposed retransmission mechanisms, and provide some insights for practical implementation.

show abstract

Section: Related Work In Wireless Communication Fieldmentioning

confidence: 99%

Section: Retransmission In Uncoded Communication Systemsmentioning

confidence: 99%

Section: Retransmission In Coded Communication Systemsmentioning

confidence: 99%

See 1 more Smart Citation

Decentralized Inference with Graph Neural Networks in Wireless Communication Systems

Lee¹,

Yu²,

Dai³

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…As the CNN is acting on signals synthesized from locally available information, such an architecture naturally encodes the information sharing specified by the graph shift operator of the graph convolution layer. In the interest of space, we omit a detailed description of AGNNs, and instead refer the reader to Gama et al (2020a) for a more detailed overview of their use in the context of distributed control.…”

Section: Graph Neural Networkmentioning

confidence: 99%

“…For the problem of communication topology co-design, we propose parameterizing distributed controllers using GRNNs -GRNNs have been successfully used for decentralized control in the context of imitation learning in Gama et al (2020a). GRNNs extend the aforementioned GNN-based controller parameterizations by introducing a local hidden state z i (t) ∈ R h at each sub-controller, resulting in a dynamic distributed controller.…”

Section: Graph Recurrent Neural Networkmentioning

confidence: 99%

Communication Topology Co-Design in Graph Recurrent Neural Network Based Distributed Control

Yang¹,

Matni²

2021

Preprint

View full text Add to dashboard Cite

When designing large-scale distributed controllers, the information sharing constraints between sub-controllers, as defined by a communication topology interconnecting them, are as important as the controller itself. Controllers implemented using dense topologies typically outperform those implemented using sparse topologies, but it is also desirable to minimize the cost of controller deployment. Motivated by the above, we introduce a compact but expressive graph recurrent neural network (GRNN) parameterization of distributed controllers that is well suited for distributed controller and communication topology co-design. Our proposed parameterization enjoys a local and distributed architecture, similar to previous Graph Neural Network (GNN)-based parameterizations, while further naturally allowing for joint optimization of the distributed controller and communication topology needed to implement it. We show that the distributed controller/communication topology co-design task can be posed as an 1 -regularized empirical risk minimization problem that can be efficiently solved using stochastic gradient methods. We run extensive simulations to study the performance of GRNN-based distributed controllers and show that (a) they achieve performance comparable to GNN-based controllers while having fewer free parameters, and (b) our method allows for performance/communication density tradeoff curves to be efficiently approximated.

show abstract

Multi-robot Coordination and Planning in Uncertain and Adversarial Environments

Zhou

Tokekar

2021

Curr Robot Rep

View full text Add to dashboard Cite

Purpose of Review Deploying a team of robots that can carefully coordinate their actions can make the entire system robust to individual failures. In this report, we review recent algorithmic development in making multi-robot systems robust to environmental uncertainties, failures, and adversarial attacks.Recent Findings We find the following three trends in the recent research in the area of multi-robot coordination: (1) resilient coordination to either withstand failures and/or attack or recover from failures/attacks; (2) risk-aware coordination to manage the trade-off risk and reward, where the risk stems due to environmental uncertainty; (3) Graph neural networks based coordination to learn decentralized multi-robot coordination policies. These algorithms have been applied to tasks such as formation control, task assignment and scheduling, search and planning, and informative data collection. SummaryIn order for multi-robot systems to become practical, we need coordination algorithms that can scale to large teams of robots dealing with dynamically changing, failure-prone, contested, and uncertain environments. There has been significant recent research on multi-robot coordination that has contributed resilient and risk-aware algorithms to deal with these issues and reduce the gap between theory and practice. Learning-based approaches have been seen to be promising, especially since they can learn who, when, and how to communicate for effective coordination. However, these algorithms have also been shown to be vulnerable to adversarial attacks, and as such developing learning-based coordination strategies that are resilient to such attacks and robust to uncertainties is an important open area of research.

show abstract

Graph Neural Networks for Decentralized Controllers

Cited by 5 publications

References 19 publications

Decentralized Inference with Graph Neural Networks in Wireless Communication Systems

Decentralized Inference with Graph Neural Networks in Wireless Communication Systems

Communication Topology Co-Design in Graph Recurrent Neural Network Based Distributed Control

Multi-robot Coordination and Planning in Uncertain and Adversarial Environments

Contact Info

Product

Resources

About