Representative-based Protocol for Multiple Interdependent Issue Negotiation Problems

Abstract

Multi-issue negotiation protocols represent an important field in the multi-agent systems study. In fact, negotiation covers many aspects of our lives has led to extensive research in the area of automated negotiators, that is, automated agents capable of negotiating with other agents in a specific environment include e-commerce, large scale argument on worldwide problem(Malone & Klein, 2007), collaborative design for cars and so on. Even though there has been a lot of previous work in this area ((Bosse & Jonker, 2005; Faratin et al., 2002; Fatima et al., 2004; Lin & Chou, 2003)) most have dealt exclusively with simple negotiations involving independent multiple issues. Many real-world negotiations, however, are complex negotiation and involve interdependent issues. When designers work together to design a car, for example, the value of a given carburetor is highly dependent on which engine is chosen. We study on the multiple interdependent issues negotiation using automated agent with nonlinear utility function reflected on the real-world negotiations. The Bidding-based Negotiation Protocol is high performance on multi interdependent issues negotiation(Ito et al., 2007). In bidding based protocol, agents generate bids by sampling and searching their utility functions, and the mediator finds the optimum combination of submitted bids from agents. However, the bidding-based negotiation protocol has two main issues. 1) Privacy: Existing works have not yet been concerned with agents’ private information. Agents’ private information should not be revealed excessively because agents who reveal much utility information are brought to a disadvantage. For example, suppose that several companies collaboratively design and develop a new car model. If one company reveals more utility information than other companies, the other companies can know more of that company’s utility information. As a result, the company is brought to a disadvantage in the next negotiations. Furthermore, it is dangerous to reveal utility information explicitly as an aspect of security. 2) Scalability for the number of agents: The bidding-based negotiation protocol is not a high scalability for the number of agents. In the bidding based negotiation protocol, the mediator needs to find the optimum combination of submitted bids from agents. However,