Combinatorial Online Reverse Auction: A Framework for Application in the Telecom Industry

Multi Online Reverse Auctions (MORAs) are one of the most effective approaches used in the management of supply chains to reduce costs and improve the efficiency of the procurement process. Despite the various advantages of this approach, the application of (MORA) is considered low by procurement professionals in the telecom industry due to the perceived computational complications associated to it. Accordingly, in this thesis, a framework for combinatorial online reverse auction is proposed for the group purchasing of scratch cards for multiple lots. The models provided in this thesis are mainly focused on the evaluation methodologies consisting of the Winner Determination Problem for multiple scenarios. The first model involves a sealed bid approach where each seller submits multiple packages, while the buyer’s objective is to determine the winning bids that minimize the total purchasing cost while fulfilling the requirements for all of the lots. The second model allows further flexibility as it follows a free disposal approach and permits partial bidding for each lot. One of the main features of both of the models is the introduction of the penalty which is applied for excluding certain predefined small lots with lower demand. The results show that the introduction of penalties force sellers to include small lots in their bids, which reduces the number of awarded seller even though the objective value increases. In addition penalties have only proven to be effective up to a certain value. The results of the second model show that applying the free disposal approach contributes towards lowering the total procurement cost.