One of the significant challenges in the daily operation of train dispatching is making the rightrndecision upon unplanned conflict occurrence. The resolution process by itself will introduce anrnadditional delay on the railway network unless well resolved. Thus, it is a decisive and challengingrnissue for train dispatchers and railway operation planners to decide which of the trains to stop orrnto pass from the trains involved in the conflict to bring minimum propagated delay. Such anrnoperation with effective conflict resolution requires an intelligent decision support system thatrnconsiders minimization of future dwelling time. rnTo this end, this thesis developed a decision support system that provides an intelligent decisionrnto the train dispatcher by detecting a conflict on a rail network with an optimal resolution of thernconflict and cost. The approach addressed the minimization of overall delay due to the conflictrnresolution in addition to detection and resolution. A mixed integer linear programming approachrnhas been implemented to find optimal combinations of arrival and departure events that bringrnminimum propagated delay. Optimization toolbox of the commercial software, MATLABrnR2015a, was used to develop the solving algorithm and obtain the result. The solution procedurernis also clearly illustrated using practical and hypothetical applications. The model has been appliedrnto Ethio-Djibouti Railway enterprise railroad from DIRE-DAOUA to DAOUENLE, whichrnconsists of eight stations. The program was able to reach an optimal solution with minimum costrnwhen compared to the manual (heuristic) approach especially for an increased number of trainsrnand stations. The model was also tested based on various hypothetical assumptions and showedrnthat it is a powerful tool to be used for train dispatchers for ensuring operational optimality andrnsafety of the railway line.