We present dual polarization and flux calibrated measurements of Radio Frequency Interference (RFI) in the radio astronomy protected band between 1400 and 1427 MHz. These observations suggest that a subset of satellites observed likely produce ground level RFI in excess of ITU thresholds for interference to radio astronomy. They also highlight the effects and challenges associated with modeling and measuring RFI. Such challenges include limitations in measurement timing from errors in predicted satellite orbits using publicly available ephemeris information, the difficulty of distinguishing individual satellites in an increasingly crowded orbital environment, dynamic satellite platform behaviors such as beamforming, and contributions from other signals reflecting off of the satellites being observed.
Dynamic spectrum sharing and coexistence in shared and unlicensed spectrum call for receivers that can retune quickly while withstanding strong in-band and adjacent-channel interference from incumbent users or nearby networks. This paper introduces a digitally assisted RF-domain interference canceller for wide-dynamic-range, frequency-agile receivers, realized on a direct-RF system-on-chip (RFSoC) platform. The RF front-end is divided into a high-sensitivity path, where a DAC-generated cancellation signal is injected ahead of the LNA/ADC, and a low-sensitivity path that supplies an interference-rich internal reference for autonomous blocker detection. Welch-based PSD estimation combined with peak detection identifies the dominant interferers and configures per-blocker down/up-conversion along with reconfigurable low-pass filtering. A deterministic multitone probing scheme estimates a compact two-path leakage model, whose parameters control a Farrow fractional-delay engine to achieve sub-sample timing alignment at multi-GS/s sampling rates. Measurements on an RFSoC prototype using OFDM-like interferers demonstrate deep RF-domain notches and a further 10–15 dB suppression from fractional-delay alignment, with the desired band largely preserved. These results indicate that modest levels of digitally aided RF cancellation can recover ADC headroom and enhance blocking robustness, offering a practical building block for spectrum-sharing radios and experimental platforms.
Spectrum occupancy analysis is essential for enabling dynamic sharing of the scarce RF spectrum among coexisting wireless technologies. Recent works typically address wideband spectrum analysis by applying deep learning (DL) based object detection networks on spectrogram images of the monitored frequency bands to perform signal detection, localization, and classification. While effective, such approaches have certain limitations, including additional latency, information loss due to the transformation into image format, and reliance on pre- and post-processing steps. In this work, we propose an alternative DL framework that is able to detect, localize, and classify signals directly from frequency domain I/Q samples of wideband recordings. Following a direct set prediction problem formulation, our approach employs a transformer based network to generate a set of signal predictions in an end-to-end manner, eliminating the need for image generation and post-processing steps. We evaluate the proposed framework on a public dataset containing real-world wideband measurements covering five wireless technologies. Experimental results demonstrate that our approach achieves high signal detection performance, competitive spectrum segmentation results, and improved robustness to noise.
The rapid growth of low power wide area networks (LPWANs) in unlicensed industrial, scientific, and medical (ISM) bands has significantly increased spectrum congestion, raising new challenges for coexistence monitoring and regulatory compliance. Existing datasets and simulators are largely designed for communication performance evaluation and lack the temporal continuity, regulatory context, and non-compliant behaviors required for spectrum enforcement tasks. This paper introduces a regulatory-aware, cross-layer spectrum simulator for the coexistence of long range (LoRa) and IEEE 802.15.4g systems in the 868 MHz band. The proposed framework generates continuous-time complex baseband I/Q sequences driven by stochastic traffic models and constrained by ETSI regulations. It combines PHY-accurate waveform synthesis with realistic channel effects and explicitly injects regulatory violations, including bandwidth, duty cycle, out-of-band emissions, and power exceedance. Our proposed simulator, SynSpec-EU868, produces labeled spectrum snapshots suitable for training data-driven spectrum monitoring and compliance detection models, providing a practical tool for next-generation regulatory-aware sensing systems.
Alamouti coding provides one of the simplest and most elegant ways to enable and collect full diversity, but its benefits hinge on reliable synchronization and channel estimation, as well as negligible interference. In spectrum-sharing, industrial/hospital and tactical deployments, strong co-channel interference can easily break down Alamouti decoding if the interference is not properly mitigated a priori - which is hard to do. In this paper, we address this challenge by marrying Alamouti coding and signal alignment – a technique that exploits repetition and multiple receive antennas to form two received data views that contain the same signal but different interference. This enables simple interference suppression and signal estimation via subspace intersection, implemented using Canonical Correlation Analysis (CCA). However, CCA alone does not exploit the orthogonal structure induced by Alamouti coding and therefore cannot harvest the associated diversity gains. Towards this end, we develop a practical two-stage receiver which first estimates and suppresses the interference subspace using CCA, and then applies standard Alamouti combining on the projected streams. Our analysis and experiments show that the combination offers remarkable interference resilience from low to moderate SINRs at thin SVD-level complexity. Our work shows that the Alamouti signal structure, which was originally designed without interference in mind, can also be leveraged to reject unpredictable interference.
