COMPARISON OF ALGORITHMS FOR DETERMINING THE FREQUENCY COMPONENTS OF A SIGNAL IN MOBILE SYSTEMS
DOI:
https://doi.org/10.35546/kntu2078-4481.2025.4.2.20Keywords:
software engineering for mobile applications; signal spectral analysis; performance optimization; Android NDK; real-time systems.Abstract
The article presents a systematic analysis of algorithms for determining the frequency components of a digital signal and their efficient implementation in the Android mobile environment. Classical approaches to spectral analysis, including the Fast Fourier Transform (FFT), Goertzel algorithm, and autocorrelation methods (YIN, MPM), are considered in terms of their suitability for real-time operation on devices with limited computational resources. It is shown that within the constraints of mobile architecture, the main challenge lies in balancing accuracy and computational speed. To overcome these limitations, a native implementation of computationally intensive stages of spectral analysis was developed in C++ using the Android NDK. The key architectural elements include the KissFFT library for performing fast Fourier transforms, ARM NEON technology for vector optimization of multiplication and addition operations, and the Oboe library, which provides low-latency audio exchange without involving the virtual machine. The paper describes the principles of integration between Java and C++ via JNI, minimization of data copying, and the creation of a custom buffer pool to avoid garbage collector pauses during signal processing. The experimental part included both functional and stress testing. The obtained results confirmed that the system can operate in real-time mode with a total latency not exceeding 12 ms, which meets the requirements of professional music applications. General recommendations have been formulated for selecting the appropriate algorithm depending on the target task: FFT for spectral analysis and visualization, Goertzel for selective frequency detection, and autocorrelation methods for precise fundamental tone estimation. It has been demonstrated that combining native signal processing, SIMD optimization, and algorithmic flexibility ensures an optimal balance between accuracy, stability, and energy efficiency. The proposed approach can be further applied to the development of mobile audio tuners, spectrum analyzers, automatic sound recognition systems, and educational tools for musicians.
References
Khadem A., Fujiki D., Talati N., Mahlke S., Das R. Vector-Processing for Mobile Devices: Benchmark and Analysis // Proceedings of the 2023 IEEE International Symposium on Workload Characterization (IISWC 2023). New York : IEEE, 2023. P. 1–12. DOI 10.1109/IISWC59245.2023.00036.
Lee J., Park H., Kwon S. та ін. Energy-Efficient Low-Latency Audio on Android // Journal of Systems and Software. 2019. Vol. 157. P. 110–123. DOI 10.1016/j.jss.2019.03.013.
Bloessl B., Baumgärtner L., Hollick M. Hardware-Accelerated Real-Time Stream Data Processing on Android with GNU Radio // Proceedings of the 14th ACM International Workshop on Wireless Network Testbeds, Experimental Evaluation and Characterization (WiNTECH ’20). New York : ACM, 2020. P. 35–40. DOI 10.1145/3411276.3412184.
Lota J., Demosthenous A. Low Computational Sensing with Goertzel Filtering for Mobile Industrial IoT Devices // Proceedings of the IEEE International Symposium on Circuits and Systems (ISCAS 2020). Seville : IEEE, 2020. P. 1–5. DOI 10.1109/ISCAS45731.2020.9180896.
Maity A., Prakasam P., Bhargava S. Robust Dual-Tone Multi-Frequency Tone Detection Using k-Nearest Neighbour Classifier for a Noisy Environment // Applied Computing and Informatics. 2025. Vol. 21, No. 2. P. 45–59. DOI 10.1108/ACI-10-2020-0105.
Hoopes J., Chalmers B., Zappi V. Neural Audio Processing on Android Phones // Proceedings of the 27th International Conference on Digital Audio Effects (DAFx 2024). Huddersfield : University of Huddersfield Press, 2024. P. 33–40. URL: https://www.dafx.de/paper-archive/2024/papers/DAFx24_paper_78.pdf.
Liu H., Wei Z., Chen X. Integrated Sensing and Communication Signal Processing Based on Compressed Sensing Over Unlicensed Spectrum Bands // IEEE Transactions on Cognitive Communications and Networking. 2024. Vol. 10, No. 3. P. 1801–1816. DOI 10.1109/TCCN.2024.3391307.
Danielsson A. Comparing Android Runtime with Native: Fast Fourier Transform on Android : Master’s Thesis. – Stockholm : KTH Royal Institute of Technology, 2017. 78 p. URL: https://www.diva-portal.org/smash/get/diva2:1109257/FULLTEXT01.pdf.
Bloessl B., Baumgärtner L., Hollick M. Accelerated Processing for Wireless Applications on Android Using GNU Radio // ACM WiNTECH ’20 Companion Proceedings. New York : ACM, 2020. P. 41–47. DOI 10.1145/3411276.3412190.
Sugawara K. Efficient FFT Algorithms for Mobile Devices : Master’s Thesis. Espoo : Aalto University, School of Electrical Engineering, 2016. 72 p. URL: https://aaltodoc.aalto.fi/handle/123456789/19783.
