Algorithm for signal detection based on adaptive threshold calculation

Problem statement. Primary information processing in modern radio-electronic equipment faces the fundamental problem of noise and interference affecting useful signals. This is especially acute in a complex radio-electronic environment (REE). High density of radio emissions, complex terrain, and intentional jamming make it difficult or impossible to detect signals using existing methods (Bayesian filtering, multi-hypothesis classification, etc.), whose effectiveness decreases with an increasing number of emission sources and changes in their parameters.

Objective. Development and experimental testing of a signal detection algorithm based on an adaptive threshold calculated by statistical processing of the amplitude-frequency spectrum obtained from an SDR receiver.

Results. An algorithm is proposed that performs multiple spectrum averaging, estimates the mathematical expectation and standard deviation (STD) of the noise level, and forms a detection threshold as, where n is an adaptive coefficient. Based on the algorithm, C++ software was developed for the HackRF One SDR receiver. Experimental testing on detecting control signals from various quadcopter models confirmed the operability of the system. The detection range of a –58.9 dB signal was at least 75 m in open terrain and at least 45 m in difficult terrain conditions (operator below ground level).

Practical significance. The developed software and hardware system can be used for a wide range of tasks requiring radio frequency spectrum analysis: spectrum monitoring, ensuring electromagnetic compatibility, radio channel control in various industries, and in the future, for enhancing situational awareness of military personnel.