Collaborative networks: From network analysis to fair network coordinatio

Globalization has not only reduced several product and service prices but also increased product and service availability. As a result, demand and inventory re-allocation between collaborative enterprise network (CEN) members has become essential in improving such performance to be more competitive. This research proposes two centralized coordination strategies, Proportional Inventory Deficit Satisfaction Algorithm (PIDSA) and the Minimal Inventory Allocation Algorithm (MIAA). PIDSA supports more effective inventory and demand sharing among all members through balanced collaboration based on single enterprise and network characteristics. MIAA first re-allocates inventory or demand based on minimal requirements within the network needs. Partner Selection Protocol (PSP) is a decentralized coordination strategy that was proposed to compare the two strategies. The performance of the coordination strategies has been validated and analyzed by comparing three models: No Collaboration (NC), Partial Collaboration (PC) and Complete Collaboration (CC). The difference between PC and CC is that CC has considered enterprise inventory adjustments through gained knowledge by network information sharing. The aspect of network characteristics by comparing homogeneous and heterogeneous collaborative enterprise networks (HoCENs and HeCENs, respectively) has been considered. Network homogeneity is assumed when all characteristics of each enterprise in the CEN are identical; on the other hand, network heterogeneity is modeled through variations of each enterprise's characteristics such as demand parameters, inventory levels and network member composition. The experimental results show that PSP is outperformed by MIAA and PIDSA in terms of average inventory utilization, demand fulfillment and total profit in any network- and collaboration type. PSP only shows optimal performance for complete order fulfillments under CC in larger HoCENs. Fairness was another performance measure and in contrast to other measures, experiences optimal results under PIDSA in PC and HoCENs. It was also observed that smaller enterprises prefer collaborating in HeCENs with larger members.