1: \begin{abstract}

2: We study the average consensus problem of multi-agent systems for

3: general network topologies with unidirectional information flow. We

4: propose two (linear) distributed algorithms, deterministic and

5: gossip, respectively for the cases where the inter-agent

6: communication is synchronous and asynchronous. Our contribution is

7: that in both cases, the developed algorithms guarantee state

8: averaging on arbitrary strongly connected digraphs; in particular,

9: this graphical condition does not require that the network be

10: balanced or symmetric, thereby extending many previous results in

11: the literature. The key novelty of our approach is to augment an

12: additional variable for each agent, called ``surplus'', whose

13: function is to locally record individual state updates.  For

14: convergence analysis, we employ graph-theoretic and nonnegative

15: matrix tools, with the eigenvalue perturbation theory playing a

16: crucial role.

17: \end{abstract}