Software Project Scheduling Using a Multi-Agent System

Multi-agent system can be implemented to exploit parallel processing of project scheduling. Since mathematical approaches cannot describe the complexity of real problems, and have difficulties to adapt the analysis to dynamics changes, therefore Monte Carlo simulation can be used to quantify the effects of risk and uncertainty in both project schedules and budgets. Also, genetic algorithms can be used to provide a prioritization of tasks. Within a multi-agent system, a broker Agent can assign project scheduling tasks to different worker Agents using different project scheduling algorithms.