Artificial neural network optimal modelling of received signal strength in mobile communications using UAV measurements

Artificial neural network optimal modelling of received signal strength in mobile communications using UAV measurements. Goudos, S., K., Tsoulos, G., & Athanasiadou, G. In IET Conference Publications, volume 2018, 2018.
doi abstract bibtex

In this paper, we apply an alternative procedure for the prediction of received signal strength in mobile communications based on Artificial Neural Networks (ANN). We use experimental data measurements taken from an unmanned aerial vehicle (UAV) for ANN training. We apply several evolutionary algorithms (EAs) in conjunction with the Levenberg-Marquardt (LM) backpropagation algorithm in order to train different ANNs. We design two new hybrid training methods by combing LM with self-adaptive Differential Evolution (DE) strategies. These new training methods achieve better convergence of neural network weight optimization than the original LM method. The received results are compared to the real values using representative ANN performance indices and exhibit satisfactory accuracy.

@inproceedings{
 title = {Artificial neural network optimal modelling of received signal strength in mobile communications using UAV measurements},
 type = {inproceedings},
 year = {2018},
 keywords = {ANN,DE,LM,Mobile communications,Propagation prediction,UAV},
 volume = {2018},
 issue = {CP741},
 id = {a365915c-1912-3029-9777-6e4a4da3262f},
 created = {2020-02-29T16:57:42.220Z},
 file_attached = {false},
 profile_id = {c69aa657-d754-373c-91b7-64154b7d5d91},
 last_modified = {2023-02-11T18:54:03.449Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {true},
 hidden = {false},
 citation_key = {Goudos2018e},
 private_publication = {false},
 abstract = {In this paper, we apply an alternative procedure for the prediction of received signal strength in mobile communications based on Artificial Neural Networks (ANN). We use experimental data measurements taken from an unmanned aerial vehicle (UAV) for ANN training. We apply several evolutionary algorithms (EAs) in conjunction with the Levenberg-Marquardt (LM) backpropagation algorithm in order to train different ANNs. We design two new hybrid training methods by combing LM with self-adaptive Differential Evolution (DE) strategies. These new training methods achieve better convergence of neural network weight optimization than the original LM method. The received results are compared to the real values using representative ANN performance indices and exhibit satisfactory accuracy.},
 bibtype = {inproceedings},
 author = {Goudos, Sotirios K. and Tsoulos, George and Athanasiadou, Georgia},
 doi = {10.1049/cp.2018.1079},
 booktitle = {IET Conference Publications}
}

Downloads: 0