Dimensionality Reduction Applied to Direct Regression Localization Systems in Regions with Different Levels of Urbanization

• Gabriel W. A. Silva UFPE
• Daniel C. Cunha UFPE

DOI: https://doi.org/10.5753/wperformance.2022.222508

Abstract

This work analyzes the application of the direct regression localization (DRL) method in two regions with different levels of urbanization. In addition, the effect of dimensionality reduction, through feature extraction algorithms (FEAs), is addressed on the accuracy and execution times of the radiolocalization method. Experimental results evidenced that the average prediction error of the DRL method decreased as a function of the increase in the training set in the region with the highest level of urbanization. Furthermore, the FEA kernel principal component analysis using sigmoid function provided an approximate seven-fold decrease in training time and approximately a four-fold decrease in the prediction time of the DRL method without impairing its accuracy.

Keywords: Radiolocalization, regression, feature extraction, execution time

References

Anagnostopoulos, G. G. and Kalousis, A. (2019). A reproducible analysis of RSSI fingerprinting for outdoor localization using Sigfox: Preprocessing and hyperparameter tuning. In 2019 Int. Conf. on Indoor Positioning and Indoor Navigation (IPIN), pages 1–8.

Anatel (2021). Technical report, ANATEL – Agência Nacional de Telecomunicações. Acessado em 14 de maio de 2021.

Anowar, F., Sadaoui, S., and Selim, B. (2021). Conceptual and empirical comparison of dimensionality reduction algorithms (PCA, KPCA, LDA, MDS, SVD, LLE, ISOMAP, LE, ICA, t-SNE). Comput Sci Rev, 40:100378.

FCC (2010). Second Report and Order on Wireless E911 Location Accuracy Requirements. Technical report, Federal Communications Commission, Washington-US. Rep. 10-176.

Feurer, M. and Hutter, F. (2019). Hyperparameter Optimization. In Automated Machine Learning, pages 3–33. Springer, Cham.

Hackeling, G. (2014). Mastering Machine Learning With Scikit-Learn. Packt Publishing.

Mahdavinejad, M. S., Rezvan, M., Barekatain, M., Adibi, P., Barnaghi, P., and Sheth, A. P. (2018). Machine learning for Internet of Things data analysis: a survey. Digit Commun Netw, 4(3):161–175.

Neto, A. C. and Levada, A. L. M. (2020). Isomap-KL: a parametric approach for unsupervised metric learning. In 2020 33rd SIBGRAPI Conf. on Graphics, Patterns and Images (SIBGRAPI), pages 287–294.

Oliveira, L. L., Jr., L. A. O., Silva, G. W. A., Timoteo, R. D. A., and Cunha, D. C. (2019). An RSS-based regression model for user equipment location in cellular networks using machine learning. Wireless Netw, 25:4839–4848.

Qi, G., Jin, Y., and Yan, J. (2018). RSSI-based floor localization using principal component analysis and ensemble extreme learning machine technique. In 2018 IEEE 23rd Int. Conf. on Digital Signal Processing (DSP), pages 1–5.

Refaeilzadeh, P., Tang, L., and Liu, H. (2009). Cross-Validation. Springer US, Boston, MA.

Shafique, K., Khawaja, B. A., Sabir, F., Qazi, S., and Mustaqim, M. (2020). Internet of Things (IoT) for next-generation smart systems: A review of current challenges, future trends and prospects for emerging 5G-IoT scenarios. IEEE Access, 8:23022–23040.

Yousaf, M., Khan, M. S. S., Rehman, T. U., Ullah, S., and Li, J. (2021). NRIC: A noise removal approach for nonlinear isomap method. Neural Process. Lett., 53(3):2277–2304.