A one-to-many Paralleling Negotiation Model and its Paralleling Negotiation Course in Speculative Computation

Paralleling negotiation can improve efficiency of negotiation on a large scale. Speculative computation is a course for making decision under incomplete information in multi-agent systems. This paper combines the paralleling negotiation with speculative computation for obtaining resource paid service in advance. One-to-many paralleling negotiation can provide making decision for resource in advance in speculative computation with data. Extension of speculative computation model into covering multi-process paralleling negotiation is presented. Multi-issue negotiation model based on belief-revision is presented, and multi-process paralleling negotiation course based on the two models is presented. Buyer can perform speculative computation based on negotiation results and utility distance of active processes. Finally, the experiment proves the models and paralleling negotiation course are valid.