1: \begin{abstract}

2: We study the convergence of the Augmented Decomposition Algorithm (ADA) proposed in \cite{rockafellar2017PSPA} for solving multi-block separable convex minimization problems subject to linear constraints. We show that the global convergence rate of the exact ADA is $o(1/\nu)$ under the assumption that there exists a saddle point.

3: We consider the inexact Augmented Decomposition Algorithm (iADA)  and  establish global and local convergence results under some mild assumptions, by providing a stability result for the maximal monotone operator $\mathcal{T}$ associated with the perturbation from both primal and dual perspectives. This result implies the local linear convergence of the inexact ADA for many applications such as the \textit{lasso}, total variation reconstruction, exchange problem and many other problems from statistics, machine learning and engineering with $\ell_1$ regularization. 

4: \keywords{  Separable convex minimization \and

5: 	 convergence rate \and

6: 	   augmented decomposition algorithm \and

7: 	   distributed computing}

8: % \PACS{PACS code1 \and PACS code2 \and more}

9: % \subclass{MSC code1 \and MSC code2 \and more}

10: 

11: \end{abstract}