Stable Trade Coordinations in Smart Agriculture

Abstract

This work enhances a previous work and focus on an efficient parallel and decentralized coordination mechanism dealing with: - uncertainties on dependencies and on conflicts between farmer, trader, custumers, freight carrier, … - uncertainties on preferences and teamworks (coalition) in order to achieve the stochastics events (e. g. the rain blocks the delivery while the customer has deadlines to meet). Usually, when a farmer wants to sell his products, he faces with a heterogeneous and distributed real-world context in order to anticipate the coordination issues. We assume no prior knowledge on stable coalitions to form and it is not possible to compute in a centralized manner the stable coalitions to form before the task achievement due to parteners’ uncertainties, stochastic events and time constraints. To handle these issues, we propose a coalition formation mechanism named TSC (Trade in Stochastic Context). Its main properties are: -the coalitions are Nash-stable, -they maximize the utilitarian social welfare and -they are auto-stables. TSC combines the formalism of the MDP (Markov Decision Process) and the laws of probability. The analysis and the experiment of our method show how we overcome these uncertainties in order to reach the required coalitions.