Path Loss Prediction Based on Machine Learning Methods for Aircraft Cabin Environments

Abstract

Wireless communications in aircraft cabin environments have drawn widespread attention with the increase of application requirements. To ensure reliable and stable in-cabin communications, the investigation of channel parameters such as path loss is necessary. In this paper, four machine learning methods, including back propagation neural network (BPNN), support vector regression (SVR), random forest, and AdaBoost, are used to build path loss prediction models for an MD-82 aircraft cabin. Firstly, machine-learning-based models are designed to predict the path loss values at different locations at a fixed frequency. It is shown that these models fit the measured data well, e.g., at 2.4 GHz central frequency the root mean square errors (RMSEs) of BPNN, SVR, random forest, and AdaBoost predictors are 1.90 dB, 2.20 dB, 1.76 dB, and 2.12 dB. Subsequent research is engaged to forecast path loss at a new frequency based on available information at known frequencies. Additionally, to solve the data limitation problem at the new frequency, we propose a path loss prediction scheme combining empirical models and machinelearning-based models. This scheme uses estimated values generated by the empirical model according to prior information to expand the training set. To verify the performance of this scheme, measured samples at 2.4 GHz and 3.52 GHz, as well as samples generated by the empirical model are employed as the training set for the path loss prediction at 5.8 GHz. The RMSEs of BPNN, SVR, random forest, and AdaBoost models are 2.49 dB, 2.78 dB, 2.54 dB, and 3.76 dB. In contrast, without samples generated by the empirical model, the RMSEs of those models are 3.84 dB, 4.94 dB, 6.57 dB, and 6.77 dB. Results show that the proposed data expansion scheme improves prediction performance when there are few measurement samples at the new frequency. INDEX TERMS Aircraft cabin, data expansion, machine learning, path loss prediction, propagation characteristics.

Space-Frequency-Interpolated Radio Map

“…They showed that this FFNN can predict path loss over multibands in the ultra-high frequency (UHF). Additionally, similar results have been reported in other situations, such as aircraft cabin environments [30] and very high frequency (VHF) bands [34]. Designs of hyper-parameters are discussed by Sotiroudis et al in [33] and Popoola et al in [35].…”

Statistical tuning of cost 231 Hata model in deployed 1800mhz GSM networks for a rural environment

“…Generally, propagation models can be categorized into two categories: deterministic and empirical propagation models [5]. Deterministic radio propagation models are path loss models that uses the laws governing the propagation of electromagnetic wave for determination of the power of a received signal at a given location [5]; while empirical propagation models are mathematical formulations based on observation and measurement obtained from the propagation environment. These type of propagation models are derived empirically from statistical analysis of large number of field measurement [6].…”

Estimation of the Path-Loss Exponent by Bayesian Filtering Method

“…Many more unique methods for creating predictive models have emerged. These approaches are based on machine learning methods, such as back propagation neural networks, support vector regression, or random forest [ 26 ].…”