Multi-agents and GIS framework for collaborative supply chain management applications

In this paper we present an agent-based framework for modeling supply chain management business flows. A supplier search algorithm that takes geographic data into consideration is introduced as well as an ontology that extends the Network Description Language (NDL) with additional information related to the internet provider's network geographic area and costs. We also present a circuit discovery method (flow) similar to the Dijkstra algorithm but using a different distance metric. The cost function takes into account a series of parameters such as: committed information rate, current segment link load, current segment total transfer capacity, latency measured, reliability of a link, additional costs.