Experimental evaluation of service negotiation protocols in agent systems

In this paper we provide experimental results concerning the impact of the negotiation protocol onto the quality of the negotiation outcome as well as onto the communication complexity of interactions incurred during one-to-many automated negotiations. We evaluate experimentally three negotiation protocols (Direct Task Assignment, Contract Net and Iterated Contract Net) with respect to two performance measures: negotiation outcome (i.e. utility) and communication complexity (i.e. number of exchanged messages) for different profiles of the contractor agents. We find that the Direct Task Assignment protocol delivers the worst average outcome, but at the same time it consumes the lowest number of messages. The Contract Net and Iterated Contract Net deliver much higher utility on average, while the Iterated Contract Net obtains the highest outcome for some configurations at the cost of the highest number of exchanged messages.