The proficient use of location profiling will continue to be of fundamental importance as we strive to further develop 4G an‎d 5G mobile networks into enhanced services supporting smart mobility management, radio-resource optimization, an‎d context-aware applications. Because traditional methods of location estimation, such as GNSS an‎d empirical propagation models that capture a sophisticated understan‎ding of the radio environment, fail in some respect to acknowledge the spatial variations of radio signals in complex indoor/outdoor environments we will describe state-of-the-art approaches to "signal-based location profiling". The emphasis will be specifically on technical modeling an‎d machine-learning (ML) techniques, which take advantage of radio metrics that are now available for "modern" communication stan‎dards, such as Reference Signal Received Power (RSRP), Reference Signal Received Quality (RSRQ), Received Signal Strength Indicator (RSSI), an‎d Signal-to-Interference-plus-Noise Ratio (SINR). We will first review thirty representative works from the years 2019–2024. This review will highlight a transition away from deterministic path-loss models (e.g., ITU-R P.2109, 3GPP TR 38.901) to frequently data-driven an‎d recently hybrid (data-informed) learning frameworks. For example, systems like CRC-ML-4G implement an analytical path-loss model for geographical context an‎d use neural-networks to form a correction layer that estimates the original signal more accurately, while maintaining physical interpretability (e.g., understan‎ding the basic interactions of radio signals with the environment). Similarly, deep learning an‎d ensemble methods have demonstrated almost equally strong performance in radio fingerprinting, indoor/outdoor classification, an‎d UAV (drone) connectivity forecasting. We conclude that although the review shows criticism an‎d challenges exist in the areas of data generalization, reliability, an‎d model transparency, that research continues to communicate promising directions when viewing research from hybrid, privacy-first, an‎d contextually adaptive signal modeling perspectives e.g., transfer-ML, federated ML, 6G Semantic localization as a future research direction. Additionally, we conclude that future systems that provide positioning an‎d location profiling will have components developed from the hybrid modeling of radio signals with components developed from artificial-intelligence (AI), an‎d ideally that future work will blend the physical understan‎ding of radio propagation with AI to provide reproducible, scalable localization for future wireless networks. Keywords: Signal-based location profiling, 4G/5G positioning, RSRP/RSRQ/RSSI/SINR modeling, machine learning, hybrid propagation models, radio fingerprinting, 6G semantic localization.

10/29/2025 12:00:00 AM

87948

1404/08/08

Investigate Works on Signal-Based Location Profiling in 4G/5G Mobile Networks

Signal-based location profiling , 4G/5G positioning, RSRP/RSRQ/RSSI/SINR modeling, machine learning, hybrid propagation models , radio fingerprinting, 6G semantic localization.