Proportional-integral-based distributed nesterov gradient methods for distributed optimization
Author Identifier (ORCID)
Abstract
This paper investigates the distributed constrained optimization problem, where participating agents cooperatively minimize the average of individual cost functions through local computation and information exchange. To address this problem, we develop a proportional-integral-based distributed algorithm from a new dynamic average consensus perspective, and further incorporate Nesterov momentum to accelerate convergence. The algorithm requires communication of only a single intermediate variable. Based on a spectral-radius-based analysis framework, we demonstrate a linear convergence of the proposed algorithm, and numerical simulations validate its effectiveness.
Keywords
distributed constrained optimization, Nesterov gradient methods, proportional-integral strategy
Document Type
Conference Proceeding
Date of Publication
1-1-2026
ISSN
15503607
Publication Title
ICC 2026 - IEEE International Conference on Communications
Publisher
IEEE
School
School of Engineering
RAS ID
102052
ISBN
[9798319542090]
Copyright
free_to_read
Recommended Citation
Yang, Z., Li, K., Pourkabirian, A., Ni, W., & Guizani, M. (2026). Proportional-integral-based distributed nesterov gradient methods for distributed optimization. In ICC 2026 - IEEE International Conference on Communications (pp. 1-6). IEEE. https://doi.org/10.1109/ICC59461.2026.11588109