OFFLINE RADIO SIGNAL PROCESSING AND VISUALIZATION SYSTEM

Abstract

The paper presents the design and implementation of an autonomous software-defined radio (SDR) scanner intended for reproducible radio engineering experiments in environments with limited network connectivity and constrained computing resources. The primary objective is to ensure strict causality and full reproducibility of the signal processing chain “from receiver to viewing,” while eliminating dependencies on external message brokers, time-series databases, or cloud-based infrastructure.

The proposed architecture forms signal frames directly at the capture node using HackRF hardware and FFTW-based spectral processing with a fixed transform size of N = 4096. Each frame is serialized as newline-delimited JSON (NDJSON), following the principle of “one event per line,” and transmitted to a lightweight backend implemented with FastAPI and SQLite operating in write-ahead logging (WAL) mode. This backend simultaneously supports persistent storage and real-time streaming to a web-based client via Server-Sent Events (SSE) or WebSocket (WS) interfaces.

Receiver parameter control, including center frequency, sample rate, LNA and VGA gains, and attenuation, is implemented through a Unix Domain Socket interface. A transactional control mechanism based on a txid↔ack handshake guarantees deterministic application of commands strictly between frames, thereby preserving temporal causality. Frame integrity is maintained using precise timestamps, while frequency-domain invariants are ensured through fixed bin spacing (bin_hz), enabling accurate reconstruction and “as recorded” playback of measurement sessions.

The paper also demonstrates a fully offline deployment on a Raspberry Pi (ARM64), confirming the feasibility of autonomous operation in field conditions. In addition, an approach to aggregating observation intervals based on time ranges is introduced, along with lifecycle management of stored intervals for efficient long-term use.

The proposed solution minimizes architectural complexity while remaining scalable and extensible, making it well suited for field measurements, autonomous spectrum monitoring, rapid prototyping of radio control methods, and the development of proprietary SDR-based monitoring and analysis systems.