"Surface EMG and Synthetic Signal Dataset for High-Resolution Spectrum and Fractional Frequency Analysis"

"High-resolution frequency analysis, including mixed integer and fractional frequency characterization, is essential in biomedical signal processing for extracting clinically meaningful diagnostic information. The validation of such advanced spectral analysis methods requires datasets containing both controlled synthetic signals and real biosignal recordings.This dataset provides 40 synthetically generated discrete-time signals and 10 surface electromyography signals (sEMG) recordings collected during dynamic lower-limb activities, including normal walking and cycling. The synthetic signals were generated across ten frequency bands (1\u201325 Hz, 1\u201350 Hz, 1\u201375 Hz, 1\u2013100 Hz, 1\u2013150 Hz, 1\u2013200 Hz, 1\u2013250 Hz, 1\u2013300 Hz, 1\u2013400 Hz, and 1\u2013500 Hz) using fractional frequency resolution of 0.5 Hz and integer frequency resolution of 1 Hz, synthesized at 2fmax and 4fmax sampling rates. These signals enable detailed benchmarking of spectral analysis techniques that aim to capture mixed integer and fractional frequency components.The real-world dataset includes the gastrocnemius sEMG signals recorded at 2148 Hz during normal walking, cycling, and post-cycling walking segments. The normal walking before cycling, cycling and normal walking after cycling segments are time-annotated to support accurate validation of biomechanical and neuromuscular analysis algorithms.This dataset is intended to support researchers working on high-resolution spectral analysis, graph-based signal processing, neuromuscular modelling, and advanced biomedical signal processing methodologies."