Academia.eduAcademia.edu

Path Loss

description5,176 papers
group874 followers
lightbulbAbout this topic
Path loss refers to the reduction in power density of an electromagnetic wave as it propagates through space or a medium. It quantifies the attenuation of signal strength due to factors such as distance, obstacles, and environmental conditions, and is a critical parameter in wireless communication system design and analysis.
lightbulbAbout this topic
Path loss refers to the reduction in power density of an electromagnetic wave as it propagates through space or a medium. It quantifies the attenuation of signal strength due to factors such as distance, obstacles, and environmental conditions, and is a critical parameter in wireless communication system design and analysis.

Key research themes

1. How can machine learning and deep learning models improve path loss prediction accuracy across varied environments and network parameters?

This theme investigates the application of machine learning (ML) and deep learning (DL) techniques to enhance the accuracy and versatility of path loss prediction models. Traditional empirical and deterministic models often fail to capture complex environmental variability or require extensive computations and measurements. ML/DL models seek to provide accurate, computationally efficient, and adaptable path loss predictions across multiple frequencies, antenna heights, and diverse propagation environments by leveraging data-driven approaches and environmental feature extraction.

Key finding: Developed an Extreme Gradient Boosting (XGBoost) path loss prediction model using numeric features alongside environmental features extracted from tiled satellite images. Demonstrated that incorporating satellite... Read more
Key finding: Demonstrated that ML models (ANN, SVR, RF) consistently outperform traditional log-distance empirical models in path loss prediction accuracy. Proposed two data expansion approaches—reusing old measurement campaigns and... Read more
Key finding: Showed that a feed-forward neural network (FFNN) with 48 hidden neurons and tangent activation function delivers optimal accuracy (MAE ~4.21 dB) and generalization (MAE ~4.74 dB) on real path loss measurement data from... Read more
Key finding: Addressed the challenge of ML path loss models lacking empirical models’ versatility across varying frequencies, antenna heights, and environment types. Developed an adaptive ML framework with dynamic ensemble selection... Read more
Key finding: Utilized random forest, a machine learning method, to predict path loss in an urban wireless environment. Achieved a low 9.24% mean absolute percentage error (MAPE) and RMSE of 13.6 dB, outperforming several empirical models... Read more

2. What roles do the underlying loss landscapes and optimization properties of neural networks play in developing effective path loss prediction models?

This theme explores theoretical and empirical insights into the optimization landscape of neural networks used for path loss and related prediction tasks. Understanding the loss surface geometry, presence of saddle points, local minima, and training dynamics is crucial for developing robust machine learning models for path loss prediction. Improved knowledge about neural network convergence behavior can guide selection of architectures, regularization parameters, and training strategies to obtain models that generalize well to complex wireless environments.

Key finding: Analyzed the non-convex loss surface of artificial neural networks and its implications for training stability. Demonstrated that the number of local minima and saddle points grows rapidly with network size, while higher... Read more
Key finding: Identified that current critical point-finding optimization methods in deep neural networks often converge to gradient-flat regions where the gradient norm is stationary, rather than true minima or saddle points. These flat... Read more
Key finding: Proposed a unified phenomenological model of the neural network loss landscape as a union of high-dimensional wedge-shaped manifolds. Demonstrated experimentally that hyperparameters like learning rate, network width, and L2... Read more
Key finding: Introduced multi-path neural network architectures with data-dependent routing between parallel convolutional/dense computations at each layer. This structure allows efficient feature extraction without quadratic parameter... Read more
Key finding: Investigated the effects of different loss functions (generalized cross-entropy vs specialized losses like Dice and focal loss) on the learning capability and robustness of neural networks. Found that appropriate loss choice... Read more

3. How can domain-specific empirical and hybrid modeling approaches be developed for path loss prediction in complex propagation environments such as vegetated agricultural and urban macro settings?

This theme focuses on designing specialized models for path loss prediction tailored to complex real-world propagation environments, including vegetated farmlands and urban macro environments with diverse obstacles. It highlights the need to capture distinctive physical characteristics and environmental interactions affecting signal attenuation by combining empirical measurements with advanced model fitting, hybrid machine learning structures, and domain-informed parameterizations to improve prediction reliability and support practical wireless network design in challenging environments.

Key finding: Applied artificial neural networks (ANN) to model path loss and identify impulsive noise in power delay profiles for 2.4 GHz signals in vegetated urban environment. Found that ANN outperformed nonlinear least squares... Read more
Key finding: Developed a dual-slope log-distance path loss model parameterized through extensive radio measurements in cassava farms with 1.8 m vegetation height. Found significant differences in path loss exponent between line-of-sight... Read more
Key finding: Proposed a hybrid model combining exponential path loss functions capturing general distance decay with LSTM deep networks that learn complex temporal dependencies, applied to urban macrocellular (UMa) propagation data.... Read more
Key finding: Developed and combined empirically derived close-in (CI) dual path loss model and wireless body area network (WBAN) energy models to simulate and relate path loss and transmitted energy on robotic wireless sensor networks.... Read more

All papers in Path Loss

In this paper we present theoretical data in support of the unified indoor geolocation channel model namely: path loss and multipath distribution models. First, the path loss model is currently accepted to be a function of the transmitter... more
Millimeter wave (mmWave) frequencies have been selected to provide the high data rates and capacity targeted by the next generation (5G) of mobile communication networks. Compared to the sub 6-GHz bands that have been historically... more
Investigation of the effect of beam alignment for milimeter wave (mmWave) transmission in the case of Vehicle-to-Infrastructure communication (V2I) is carried out. The investigation covered varying transmission-reception (TX-RX)... more
LoRa, or Long Range, is an increasingly popular wireless communication technology for Internet of Things (IoT) applications, offering long-distance connectivity with low power consumption. Radio channel propagation plays a critical role... more
In this study a comparative analysis of various empirical models for estimating radio wave propagation path losses with those measured experimentally for the rural area in Erbil city is presented. In Gazna village near the center of the... more
The ability to interconnect and communicate wirelessly in and around the built environment with sufficient aptitude is becoming progressively important as the array of consumer-grade portable devices with broadband wireless functionality... more
A novel, multi-slope dual breakpoint model for predicting path-loss in Ultra-Wideband (UWB) off-body communication channels, is proposed. This model is based on real-body measurements, carried out in the frequency range between... more
In this contribution, a narrowband radio channel model is proposed for rural scenarios in which the radio link operates under near-ground conditions for application in wireless sensor networks dedicated to smart agriculture. The received... more
The characterization of different vegetation/vehicle densities and their corresponding effects on large-scale channel parameters such as path loss can provide important information during the deployment of wireless communications systems... more
In this contribution, we present a narrowband radio channel model for a scenario wherein the radio link operates under near-ground conditions, occurring on a ZigBee wireless sensor networks applied to smart agriculture. A near-ground... more
The characterization of different vegetation/vehicle densities and their corresponding effects on large-scale channel parameters such as path loss can provide important information during the deployment of wireless communications systems... more
In this contribution, a narrowband radio channel model is proposed for rural scenarios in which the radio link operates under near-ground conditions for application in wireless sensor networks dedicated to smart agriculture. The received... more
The attenuation due to vegetation can limit drastically the performance of Wireless Sensor Networks (WSN) and the Internet of Things (IoT) communication systems. Even more for the envisaged high data rates expected for the upcoming 5G... more
Machine learning, which is a branch of Artificial Intelligence (AI), is a method where computers or learning machines can learn from and get information from a large amount of data. The machine learning approach involves data collection... more