[Paper Review] An Efficient Protocol for Negotiation over Combinatorial Domains with Incomplete Information

Minyi Li, Quoc Bao Vo+1 more|Swinburne Research Bank (Swinburne University of Technology)|Feb 14, 2012

Constraint Satisfaction and Optimization8 references3 citations

TL;DR

This paper proposes a novel negotiation protocol for combinatorial domains under incomplete information, enabling rational agents to converge on optimal agreements through efficient distributed search over a small subspace of outcomes. The protocol is general across preference models and demonstrates computational efficiency and feasibility in experiments.

ABSTRACT

We study the problem of agent-based negotiation in combinatorial domains. It is difficult to reach optimal agreements in bilateral or multi-lateral negotiations when the agents' preferences for the possible alternatives are not common knowledge. Self-interested agents often end up negotiating inefficient agreements in such situations. In this paper, we present a protocol for negotiation in combinatorial domains which can lead rational agents to reach optimal agreements under incomplete information setting. Our proposed protocol enables the negotiating agents to identify efficient solutions using distributed search that visits only a small subspace of the whole outcome space. Moreover, the proposed protocol is sufficiently general that it is applicable to most preference representation models in combinatorial domains. We also present results of experiments that demonstrate the feasibility and computational efficiency of our approach.

Motivation & Objective

• To address the challenge of reaching optimal agreements in bilateral or multi-lateral negotiations when agents' preferences are not fully known.
• To design a protocol that enables rational agents to identify efficient solutions despite incomplete information.
• To ensure computational feasibility by restricting search to a small subspace of the full outcome space.
• To develop a protocol general enough to be applicable across various preference representation models in combinatorial domains.
• To empirically validate the feasibility and efficiency of the proposed approach through experimental evaluation.

Proposed method

• The protocol employs a distributed search mechanism that explores only a small, strategically selected subspace of the entire outcome space.
• It enables agents to iteratively refine their offers based on incomplete preference information, using heuristics to guide search toward efficient agreements.
• The protocol is designed to be compatible with multiple preference representation models, such as CP-nets, additive utilities, and others commonly used in combinatorial domains.
• Agents use local utility estimation and negotiation rules that promote convergence toward Pareto-efficient outcomes.
• The protocol incorporates mechanisms to prevent infinite negotiation loops and ensures progress toward agreement.
• It leverages iterative improvement and constraint propagation to reduce the search space while preserving optimality guarantees under rational behavior.

Experimental results

Research questions

• RQ1 Can a negotiation protocol achieve optimal agreements in combinatorial domains when agents have incomplete knowledge of each other’s preferences?
• RQ2 How can agents efficiently search for efficient agreements without exploring the entire outcome space?
• RQ3 To what extent is the proposed protocol generalizable across different preference representation models in combinatorial domains?
• RQ4 What is the computational efficiency of the protocol in terms of search space reduction and convergence speed?
• RQ5 How do rational agents behave under the protocol, and does it lead to stable, Pareto-efficient agreements?

Key findings

• The protocol successfully enables rational agents to reach optimal agreements even under incomplete information, avoiding inefficient outcomes.
• The search space is significantly reduced, with the protocol exploring only a small subspace of the full outcome space, enhancing computational efficiency.
• The approach is general and applicable to a wide range of preference representation models used in combinatorial negotiation domains.
• Experimental results confirm the feasibility and efficiency of the protocol, demonstrating practical convergence within reasonable time and communication costs.
• The protocol maintains strong convergence properties and avoids infinite negotiation loops, supporting stable agreement formation.
• The protocol outperforms baseline methods in terms of solution quality and search efficiency in tested scenarios.