Paper published in Franklin Open

A paper titled “Communication-efficient ADMM using Gaussian process regression and fully adaptive uniform quantization” by Aldo Duarte, Binh Nguyen, Truong X. Nghiem, and Shuangqing Wei has been published in Franklin Open. This paper presents a communication-efficient distributed optimization framework for ADMM with adaptive quantization strategies that adjust mid-value, window length, and resolution at each iteration. It also introduces uncertainty-aware communication and bit-allocation mechanisms and a hybrid switching approach to maintain convergence near stagnation. Simulations on a distributed quadratic cost-sharing problem show a strong trade-off between communication reduction and solution accuracy, with the hybrid method providing the most robust and consistent performance.