Deep Reinforcement Learning for Continuous Power Allocation in Flexible High Throughput Satellites

Luis, Juan; Guerster, Markus; Portillo, Iñigo del; Crawley, Edward F.; Cameron, Bruce

doi:10.1109/ccaaw.2019.8904901

Cited by 40 publications

(45 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In a multibeam system, the power allocation is usually a continuous variable [25]; however, in this article, the power allocation is considered to be performed by selecting one value among several in a set of possible power levels per beam, due to existing technologies that allow power to be modified in 0.5-dB steps [7].…”

Section: Numerical Results and Analysismentioning

confidence: 99%

“…7). The DRM will attempt to minimize KPI 1 in (24) while maximizing KPI 2 in (25). In this sense, the performance of the algorithm used for the DRM is evaluated by exploiting a joint KPI, defined as…”

Section: B Cnn Performance For Drm and Comparison With Benchmark Algmentioning

confidence: 99%

“…However, in their proposed study, DRL architectures are based on quantizing the resources before they are allocated, whereas satellite resources, such as power, are inherently continuous. In that sense, Luis et al [25] explore a DRL architecture for power allocation that uses continuous and state action spaces, avoiding the need for discretization. Nonetheless, the policy is not optimal, as some demand is still being lost.…”

Section: Introductionmentioning

confidence: 99%

“…Some of the recent literature considers that DRM is able to take into account only two possible flexible resources [10], [11], [23], i.e., power and bandwidth, whereas other studies consider that only one flexible resources can be handled [14], [15], [25]. Differently from previous approaches, the main objective of this article is to solve the DRM problem using an ML algorithm based on CNNs.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Convolutional Neural Networks for Flexible Payload Management in VHTS Systems

Ortíz

Tarchi

Martínez

et al. 2021

IEEE Systems Journal

View full text Add to dashboard Cite

Very high throughput satellite (VHTS) systems are expected to have a large increase in traffic demand in the near future. However, this increase will not be uniform throughout the service area due to the nonuniform user distribution, and the changing traffic demand during the day. This problem is addressed using flexible payload architectures, enabling the allocation of the payload resources in a flexible manner to meet traffic demand of each beam, leading to dynamic resource management (DRM) approaches. However, DRM adds significant complexity to the VHTS systems, which is why in this article, we are analyzing the use of convolutional neural networks (CNNs) to manage the resources available in flexible payload architectures for DRM. The VHTS system model is first outlined, for introducing the DRM problem statement and the CNN-based solution. A comparison between different payload architectures is performed in terms of DRM response, and the CNN algorithm performance is compared by three other algorithms, previously suggested in the literature to demonstrate the effectiveness of the suggested approach and to examine all the challenges involved.

show abstract

Section: Numerical Results and Analysismentioning

confidence: 99%

Section: B Cnn Performance For Drm and Comparison With Benchmark Algmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Convolutional Neural Networks for Flexible Payload Management in VHTS Systems

Ortíz

Tarchi

Martínez

et al. 2021

IEEE Systems Journal

View full text Add to dashboard Cite

show abstract

“…From the four resources introduced above, recent research has focused on power allocation, [4][5][6][7][8] while the other satellite resources such as the frequency assignment and beam placement have been less studied. The latter two problems have been shown to be NP hard, 9,10 which is why brute force algorithms are unfeasible for medium-to large-scale scenarios (ie, >850 beams, as we show later).…”

Section: Literature Reviewmentioning

confidence: 99%

Static beam placement and frequency plan algorithms for LEO constellations

Osa

Guerster

Barrios

et al. 2020

Satell Commun Network

Self Cite

View full text Add to dashboard Cite

LEO satellites are growing rapidly in number, but also in configurability of the payload, which presents new challenges of how to allocate the satellite's resources. We present an algorithm to solve two subproblems of the resource allocation problem for multi-beam satellites in large LEO constellations. Our approach produces a static beam placement and a static frequency plan, taking into account the user demands and interference conditions, respectively. We provide an easy to implement and fast approximation for the beam placement problem, which is converted into an Edge Clique Cover problem using graph theory. The frequency assignment problem is translated into a constraint satisfaction problem, which is later solved with a first-fit heuristic. We apply our algorithm to a scenario based on SpaceX's Starlink constellation, and when compared to two different baselines, we show that our method requires 40% fewer beams while being able to provide service to 2% more users.

show abstract

Supervised machine learning for power and bandwidth management in very high throughput satellite systems

Ortíz

Tarchi

Martínez

et al. 2021

Satell Commun Network

View full text Add to dashboard Cite

In the near future, Very High Throughput Satellite (VHTS) systems are expected to have a high increase in traffic demand. However, this increase will not be uniform over the service area and will be also dynamic. A solution to this problem is given by flexible payload architectures; however, they require that resource management is performed autonomously and with low latency. In this paper we propose the use of Supervised Machine Learning, in particular a Classification algorithm using a Neural Network, to manage the resources available in flexible payload architectures. Use cases are presented to demonstrate the effectiveness of the proposed approach and a discussion is made on all the challenges that are presented.

show abstract

Deep Reinforcement Learning for Continuous Power Allocation in Flexible High Throughput Satellites

Cited by 40 publications

References 7 publications

Convolutional Neural Networks for Flexible Payload Management in VHTS Systems

Convolutional Neural Networks for Flexible Payload Management in VHTS Systems

Static beam placement and frequency plan algorithms for LEO constellations

Supervised machine learning for power and bandwidth management in very high throughput satellite systems

Contact Info

Product

Resources

About