1: \begin{abstract}

2: In this paper we propose and analyze two dual  methods based on

3: inexact  gradient information and averaging that generate

4: approximate primal solutions for smooth convex optimization

5: problems. The complicating constraints are moved into the cost

6: using the Lagrange multipliers. The dual problem is solved by

7: inexact first order methods based on approximate gradients and we

8: prove sublinear rate of convergence for these methods. In

9: particular, we provide, for the first time,  estimates on the

10: primal feasibility violation and primal and dual suboptimality of

11: the generated approximate primal and dual solutions. Moreover, we

12: solve approximately the inner problems with a parallel coordinate

13: descent algorithm  and we show that it has linear convergence

14: rate. In our analysis we rely on the Lipschitz property of the

15: dual function and inexact dual gradients. Further, we apply these

16: methods to distributed model predictive control for network

17: systems. By tightening the complicating constraints we are also

18: able to ensure the primal feasibility of the approximate solutions

19: generated by the proposed algorithms. We obtain a distributed

20: control strategy that has the following features: state and input

21: constraints are satisfied, stability of the plant is guaranteed,

22: whilst the number of iterations  for the suboptimal solution can

23: be precisely determined.

24: \end{abstract}

25: