1: \begin{abstract}                          % Abstract of not more than 200 words.

2: In this work, we consider solving a distributed optimization problem (DOP) in a multi-agent network with multiple agent clusters. In each cluster, the agents manage separable cost functions composed of possibly non-smooth components and aim to achieve an \emph{agreement} on a common decision of the cluster. The global cost function is considered as the sum of the individual cost functions associated with affine \emph{coupling} constraints on the \emph{clusters'} decisions. To solve this problem, the dual problem is formulated by the concept of Fenchel conjugate. Then an \emph{asynchronous distributed dual proximal gradient} (Asyn-DDPG) algorithm is proposed based on a cluster-based \emph{partial} and \emph{mixed} consensus protocol, by which the agents are only required to communicate with their \emph{neighbors} with \emph{communication delays}. An ergodic convergence result is provided, and the feasibility of the proposed algorithm is verified by solving two motivating applications in the simulation.

3: \end{abstract}