Daniele Franci

In-situ RF measurements are essential for regulatory Electromagnetic Field (EMF) exposure assessment and radio coverage characterization. These tasks have historically relied on labor-intensive surveys based on discrete spot measurements, which provide limited spatial resolution and therefore often require support from simulation tools. These limitations call for a shift toward automated and scalable RF measurement campaigns. This paper formalizes a set of design guidelines for autonomous RF measurement campaigns using mobile robotic platforms and illustrates their implementation within a modular system architecture, showing howto integrate RF sensing and autonomous robotic navigation to enable consistent, geo-referenced data collection and adaptive acquisition strategies tailored to specific measurement objectives and driven by observed radio conditions. The proposed framework is validated through an outdoor measurement campaign on a private 5G Standalone network using a quadruped robot equipped with professional RF instrumentation. Experimental results show that the system supports the autonomous execution of measurement missions for both coverage characterization and EMF exposure assessment, demonstrating the potential of the proposed approach as a scalable and effective alternative to traditional RF measurement practices.

Surveillance of electromagnetic field exposure in 5G networks is crucial for regulatory compliance but currently relies on intrusive and expensive active traffic injection. We propose a hardware-agnostic, automated, and non-invasive framework for exposure assessment in real-world deployments. By leveraging a blind Downlink Control Information (DCI) decoding mechanism, our method extracts scheduling information to isolate user-specific Physical Downlink Shared Channel (PDSCH) resources. Experimental validation in frequency range (FR)1 and FR2 demonstrates that this approach enables a procedural mapping of radiated power while obviating the need for full carrier bandwidth saturation. By abstracting the acquisition hardware as an IQ sample provider, the pipeline is designed to operate across heterogeneous receivers, including low-cost Software-Defined Radios (SDRs). While absolute metrological accuracy remains dependent on specific device calibration, this methodology supports a streamlined, user-specific Maximum-Power Extrapolation (MPE) analysis workflow at the signal-processing level that is independent of network traffic conditions.

Multi-User (MU) Multiple Input Multiple Output (MIMO) operation at millimeter-wave (mmWave) frequencies enables 5G base stations (BS) to transmit simultaneously to several users by forming multiple beams in parallel. This feature, while boosting capacity, modifies the exposure landscape compared to the conventional Single-User (SU) mode, where resources are time-multiplexed. In this work, we investigate how MUMIMO operation affects maximum-power extrapolation (MPE) procedures used to assess electromagnetic field (EMF) exposure in accordance with IEC 62232. Building on the experimental framework and measurement methodologies validated for SUMIMO in a recently published work by the authors, we extend the analysis to signals configured with MU-MIMO transmission and variable numbers of Channel State Information Reference Signal (CSI-RS) antenna ports. Results show that concurrent user scheduling may reduce the received power per beam by several dB, leading to a systematic underestimation of maximum exposure when MPE procedures are applied without knowledge of the active mode. Furthermore, the number of CSI-RS ports influences the measured levels, resulting in reductions of up to 12 dB. These findings underscore the need to incorporate MUMIMO and CSI-RS port configurations into updated exposure assessment standards.

High-fidelity in-phase and quadrature (I/Q) signal traces are critical for a variety of wireless network applications, including spectrum monitoring, interference detection and mitigation, radio-frequency (RF) fingerprinting (RFFP), smart jamming detection, anomaly identification, and modulation classification. However, the number and scope of publicly available I/Q datasets are currently limited, as most datasets are either restricted to a single frequency or wireless technology, or collected in controlled laboratory environments. This paper introduces ITALYSIG, a comprehensive, high-definition, open-source I/Q database of diverse real-world radio signals, including cellular, radar, and other wireless technologies. ITALYSIG provides I/Q captures with up to 100 MHz bandwidth, collected across diverse urban and rural environments in Italy. The I/Q signals are stored in multiple formats, including raw binary files, standardized VITA Radio Transport (VRT, VITA-49), and visual formats such as Portable Network Graphics (PNG) and JavaScript Object Notation (JSON) for broader applicability. The data-acquisition setup is based on a CRFS RFeye SenS Portable recorder at the front end, which enables automatic long-term I/Q recordings on the order of hours, multi-terabyte storage, and real-time signal processing. In addition to releasing the dataset, this article provides a comprehensive overview and qualitative comparison of state-of-the-art datasets in terms of measurement setup, data format, wireless technology, frequencies, and time duration. Furthermore, this article provides a tutorial on the end-to-end measurement setup for automatic I/Q acquisition, multi-format export, and back-end storage, as well as real-time analysis via the DeepView software. Finally, we provide guidelines for integrating I/Q traces into deep learning pipelines and highlight specific use cases of the dataset.

The assessment of maximum exposure levels generated by 5G base stations via the Maximum-Power Extrapolation (MPE) procedure defined in international standards is challenged by the technical traits of next-generation signals. Dynamic resource allocation to multiple users produces highly non-uniform signal frames in both frequency and time, complicating worst-case exposure evaluation. Furthermore, massive MIMO beamforming reduces the reliability of control channel–based measurements, motivating traffic-forcing techniques. The wide frequency span of 5G necessitates measurement protocols applicable to both FR1 and FR2 bands. In this paper, we compare conventional methods–namely, Channel Power (CP) and Zero Span (ZS)–against vector-based techniques like Vector Channel Power (VCP) and PDSCH Power (PP), based on the demodulation of the received signal. Unlike most studies focused on FR1, our measurements target a 5G millimeter-wave MIMO signal under realistic network conditions, including multi-user setups and diverse traffic-forcing scenarios. The results show that decoding-based procedures analyzing the PDSCH traffic channel reliably estimate maximum exposure, even when network operation deviates from ideal scenarios. These findings support integrating PDSCH-based methodologies into future revisions of standards such as IEC 62232.

Despite the broad adoption of 5G terminals, public concerns regarding ElectroMagnetic Field (EMF) exposure are primarily focused on the nearby presence of Radio Base Stations (RBSs), which are currently considered the strongest source of EMF radiation. However, is the RBS exposure really higher than that generated by multiple terminals in the same location? To address this question, we design and evaluate an innovative narrow-band measurement methodology, called Electromagnetic field Measurement for Fixed Wireless Access (EM-FWA), tailored to the evaluation of the exposure from multiple terminals vs. the one radiated by the RBS in a operative 5G StandAlone (SA) Fixed Wireless Access (FWA) network. Our approach, which includes a new narrow-band measurement algorithm, involves exposure assessments upon the generation of Downlink (DL), Uplink (UL) and combined DL/UL data transfers from the pool of terminals. Results, obtained from multiple locations subject to variegate propagation conditions, indicate that almost 100% saturation of the radio link resources in the DL direction is only achieved when three distinct terminals generate traffic in proximity to the RBS, thus justifying the adoption of our methodology. Conversely, terminal exposure significantly exceeds the RBS one at the cell edge, even when a single terminal is employed, with more than 1.4 V/m from CPE and less than 0.2 V/m from RBS. Furthermore, the RBS output power is influenced more by the volume of DL traffic than by the terminal propagation condition. Finally, we demonstrate that the overall exposure cost is minimized when the pool of terminals is placed close to the RBS. We believe that our work can pave the way to a new family of EMF compliance assessment procedures for 5G networks, adopting multiple terminals that generate traffic during the exposure measurements.

This work presents the results of a measurement campaign conducted on an LTE Active Antenna System (AAS) aimed at defining effective operational procedures for determining the maximum exposure generated by such systems. Four distinct scenarios, in which data traffic conditions were varied using three mobile terminals that allowed for saturating the signal frame, were defined. For this reason, it is possible to apply an exposure assessment procedure to LTE AAS similar to that proposed for 5G NR systems, namely based on forcing data traffic towards the receiving antenna and measuring the corresponding received power.

Dynamic Spectrum Sharing (DSS) allows the transmission of 5G services on existing 4G LTE networks to deploy 5G networks rapidly. It is a straightforward solution requiring only software updates, thus enabling operators to initiate nationwide coverage with minimal expenses. According to the IEC 62232 standard, measurement of human exposure to electromagnetic fields (EMFs) produced by a DSS system can be accomplished by a vector spectrum analyzer (VSA) or a method referred to as Maximum Power Extrapolation (MPE). The vector evaluation entails measuring the broadcast channel of the 5G or LTE technology while assuming that both broadcast and traffic channels are transmitted on the same beam and employing a suitable extrapolation method. Instead, the objective of the MPE procedure is to estimate the potential maximum field level at the measurement point. Both measurement methodologies rely on the knowledge of the implemented system. In this paper, we introduce an experimental study to evaluate the influence of these methodologies in a complex scenario like DSS, where multiple technologies operate within the same bandwidth.