1: \begin{abstract}

2: 

3: Achieving a balance of supply and demand in a multi-agent system with

4: many individual self-interested and rational agents that act as

5: suppliers and consumers is a natural problem in a variety of real-life

6: domains---smart power grids, data centers, and others.  In this paper,

7: we address the profit-maximization problem for a group of distributed

8: supplier and consumer agents, with no inter-agent communication. We

9: simulate a scenario of a market with $S$ suppliers and $C$ consumers

10: such that at every instant, each supplier agent supplies a certain

11: quantity and simultaneously, each consumer agent consumes a certain

12: quantity.  The information about the total amount supplied and

13: consumed is only kept with the center.  The proposed algorithm is a

14: combination of the classical additive-increase multiplicative-decrease

15: (AIMD) algorithm in conjunction with a probabilistic rule for the

16: agents to respond to a capacity signal.  This leads to a

17: nonhomogeneous Markov chain and we show almost sure convergence of

18: this chain to the social optimum, for our market of distributed

19: supplier and consumer agents. Employing this AIMD-type algorithm, the

20: center sends a feedback message to the agents in the supplier side if

21: there is a scenario of excess supply, or to the consumer agents if

22: there is excess consumption. Each agent has a concave utility function

23: whose derivative tends to 0 when an optimum quantity is

24: supplied/consumed.  Hence when social convergence is reached, each

25: agent supplies or consumes a quantity which leads to its individual

26: maximum profit, without the need of any communication.  So eventually,

27: each agent supplies or consumes a quantity which leads to its

28: individual maximum profit, without communicating with any other

29: agents.  Our simulations show the efficacy of this approach.

30: 

31: \end{abstract}

32: