Forecast-Optimized Selection of Time Segment Length for Improving the Interference Robustness of Complex Signal Ensembles
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
This paper addresses the problem of coordinated selection of the time-domain segmentation length in the formation of ensembles of complex signals using time-domain permutations. While time permutations are known to reduce mutual correlation between ensemble signals, the segmentation length is typically predefined or selected empirically, without accounting for its simultaneous impact on correlation attenuation and spectral stability. To overcome this limitation, a predictive–optimization-based method for selecting the segmentation length is proposed. The method treats the segment length as a controllable parameter and evaluates its influence on ensemble properties through two quantitative metrics: a correlation-based metric characterizing the maximum mutual correlation and a spectral deviation metric assessing structural–spectral distortions caused by time permutations. These metrics are combined into an integrated objective function, enabling algorithmic determination of the optimal segmentation length that ensures a balanced trade-off between correlation reduction and spectral preservation. Experimental modeling using stochastic complex signal ensembles demonstrates that intermediate segmentation lengths provide the most favorable operating regime, yielding a significant reduction in mutual correlation while limiting spectral distortions. The obtained results confirm the effectiveness of the proposed approach and its applicability to the design of interference-robust signal ensembles in modern telecommunication systems
Fields of Study & Subjects
Radar Systems and Signal Processing
Advanced Signal Processing Techniques
Wireless Communication Networks Research
Publication details
- Journal / source
- COMPUTER-INTEGRATED TECHNOLOGIES EDUCATION SCIENCE PRODUCTION
- Issue
- 62
- Pages
- 195–205
- Publication date
- 28 March 2026
- Language
- EN