Data-Driven Spectrum Cartography via Deep Completion Autoencoders

Teganya, Yves; Romero, Daniel

doi:10.48550/arxiv.1911.12810

Cited by 2 publications

(7 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To see this, suppose that the number of frequencies N f in F is significant, e.g. 512 or 1024 as would typically occur in practice, and consider a fully connected first layer p (1) w1 with N N neurons. Its total number of parameters becomes (N y N x (N f + N m ) + 1)N N plus possibly additional parameters of the activation functions.…”

Section: Exploiting Structure In the Frequency Domainmentioning

confidence: 99%

“…This work was supported by the Research Council of Norway through the FRIPRO TOPPFORSK grant 250910/F20 and the IKTPLUSS grant 280835 (LUCAT). Parts of this work will be presented at the IEEE International Conference on Communications 2020 [1].…”

Section: Introductionmentioning

confidence: 99%

“…Unfortunately, none of the existing approaches can learn from data, which suggests that their estimation performance can be significantly improved along this direction. 1 To this end, the first contribution of this work is a datadriven paradigm for radio map estimation. The idea is to learn the spatial structure of relevant propagation phenomena such as shadowing, reflection, and diffraction using a data set of past measurements in different environments.…”

Section: Introductionmentioning

confidence: 99%

“…The conference version of this work[1] presents the core ideas here. Relative to[1], the present paper contains improved neural network architectures (including fully convolutional networks), a methodology for PSD estimation with basis expansion models, and extensive empirical validation and comparison with existing algorithms through a realistic data set. Some of the ideas in[33] are similar to those in[1].…”

mentioning

confidence: 99%

“…Relative to[1], the present paper contains improved neural network architectures (including fully convolutional networks), a methodology for PSD estimation with basis expansion models, and extensive empirical validation and comparison with existing algorithms through a realistic data set. Some of the ideas in[33] are similar to those in[1]. However, since the former was submitted after our paper[1], it is regarded as parallel work.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Deep Completion Autoencoders for Radio Map Estimation

Teganya¹,

Romero²

2020

Preprint

Self Cite

View full text Add to dashboard Cite

Radio maps provide metrics such as power spectral density for every location in a geographic area and find numerous applications such as UAV communications, interference control, spectrum management, resource allocation, and network planning to name a few. Radio maps are constructed from measurements collected by spectrum sensors distributed across space. Since radio maps are complicated functions of the spatial coordinates due to the nature of electromagnetic wave propagation, model-free approaches are strongly motivated. Nevertheless, all existing schemes rely on interpolation algorithms unable to learn from data. In contrast, this paper proposes a novel approach in which the spatial structure of propagation phenomena such as shadowing is learned beforehand from a data set with measurements in other environments. Relative to existing schemes, a significantly smaller number of measurements is therefore required to estimate a map with a prescribed accuracy. As an additional novelty, this is also the first work to estimate radio maps using deep neural networks. Specifically, a deep completion autoencoder architecture is developed to effectively exploit the manifold structure of this class of maps.

show abstract

Section: Exploiting Structure In the Frequency Domainmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Deep Completion Autoencoders for Radio Map Estimation

Teganya¹,

Romero²

2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Aerial Spectrum Surveying: Radio Map Estimation with Autonomous UAVs

Romero¹,

Shrestha²,

Teganya³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Radio maps are emerging as a popular means to endow nextgeneration wireless communications with situational awareness. In particular, radio maps are expected to play a central role in unmanned aerial vehicle (UAV) communications since they can be used to determine interference or channel gain at a spatial location where a UAV has not been before. Existing methods for radio map estimation utilize measurements collected by sensors whose locations cannot be controlled. In contrast, this paper proposes a scheme in which a UAV collects measurements along a trajectory. This trajectory is designed to obtain accurate estimates of the target radio map in a short time operation. The route planning algorithm relies on a map uncertainty metric to collect measurements at those locations where they are more informative. An online Bayesian learning algorithm is developed to update the map estimate and uncertainty metric every time a new measurement is collected, which enables real-time operation.

show abstract

Data-Driven Spectrum Cartography via Deep Completion Autoencoders

Cited by 2 publications

References 0 publications

Deep Completion Autoencoders for Radio Map Estimation

Deep Completion Autoencoders for Radio Map Estimation

Aerial Spectrum Surveying: Radio Map Estimation with Autonomous UAVs

Contact Info

Product

Resources

About