A Deep Neural Network-Based Multi-Frequency Path Loss Prediction Model from 0.8 GHz to 70 GHz

Abstract

Large-scale fading models play an important role in estimating radio coverage, optimizing base station deployments and characterizing the radio environment to quantify the performance of wireless networks. In recent times, multi-frequency path loss models are attracting much interest due to their expected support for both sub-6 GHz and higher frequency bands in future wireless networks. Traditionally, linear multi-frequency path loss models like the ABG model have been considered, however such models lack accuracy. The path loss model based on a deep learning approach is an alternative method to traditional linear path loss models to overcome the time-consuming path loss parameters predictions based on the large dataset at new frequencies and new scenarios. In this paper, we proposed a feed-forward deep neural network (DNN) model to predict path loss of 13 different frequencies from 0.8 GHz to 70 GHz simultaneously in an urban and suburban environment in a non-line-of-sight (NLOS) scenario. We investigated a broad range of possible values for hyperparameters to search for the best set of ones to obtain the optimal architecture of the proposed DNN model. The results show that the proposed DNN-based path loss model improved mean square error (MSE) by about 6 dB and achieved higher prediction accuracy R2 compared to the multi-frequency ABG path loss model. The paper applies the XGBoost algorithm to evaluate the importance of the features for the proposed model and the related impact on the path loss prediction. In addition, the effect of hyperparameters, including activation function, number of hidden neurons in each layer, optimization algorithm, regularization factor, batch size, learning rate, and momentum, on the performance of the proposed model in terms of prediction error and prediction accuracy are also investigated.

Comparative Analysis of Major Machine-Learning-Based Path Loss Models for Enclosed Indoor Channels

“…Large-scale fading models play a vital role in optimizing base station deployments, estimating radio coverage, and characterizing the radio environment to quantify the performance of wireless communication systems [6]. Furthermore, efficient and reliable determination of crucial factors, such as the signal field strength, carrier-to-interference (C/I) ratio, and signal-to-noise ratio (SNR), can be achieved if in-depth knowledge of propagation loss is provided [7].…”

“…These multipath effects result in signal power fluctuation and increase the uncertainty of received signal power [11]. This work mainly focuses on developing large-scale path loss models that are crucial for estimating radio coverage, allocating frequencies properly, optimizing base stations, and identifying the most suitable antennas [6,12].…”

“…These models are easier to apply; however, they are time-consuming since they require massive measurement campaigns in specific environments and communication scenarios. Further, in terms of prediction accuracy, these models provide less accuracy than deterministic models [6,12,21]. Many other well-known empirical path loss prediction models are derived based on the criteria of the minimum mean square error (MMSE) between the models and the measurement data to provide the best fit of these models as a function of the separation distance between the transmitting and receiving antennas in the logarithmic scale.…”

“…Stochastic path loss models have probability distributions and assumptions to be considered in the models' equations. These models suffer from limited precision because of some mathematical expressions since the communications environments are considered random variables [6].…”

“…Multi-frequency path loss prediction models are receiving more attention from some researchers in recent times to develop accurate and stable path loss models for future wireless systems over a wide range of frequency regimes [6,31]. However, these models face the same problems mentioned above as the single-frequency models.…”