Nowadays, there is an increasing need for advanced technological solutions due to concerns about the energy crisis and environmental protection. In particular, there has been a growing focus on developing green Information and Communication Technologies (ICT) solutions, as the energy consumption of ICT devices has been rapidly increasing in recent decades. This paper introduces EPSO-based flow-aware routing mechanisms that aim to minimize Queueing Delay, Power Consumption (Energy Dissipation), and Load Deviation. Two flow-aware mechanisms, namely the Flow aware load adaptive scheme (FA-LAR) and the flow-aware distance adaptive scheme (FA-DAR), are utilized in this study. The goal is to avoid traffic congestion across the routes by selecting the most direct and efficient paths for the flows. These routed paths are optimized using the MEPSO-based FA-DAR routing algorithm, resulting in significant energy savings. Extensive numerical simulations are conducted to evaluate the proposed algorithms using metrics such as throughput, routing efficiency, queuing delay, load deviation, and energy efficiency. The simulation results demonstrate that the proposed method outperforms existing benchmark methods in all metrics. Furthermore, it ensures lower power consumption, queuing delay, and load deviation. The method exhibits a 30% improvement in throughput and a 24% improvement in energy efficiency compared to existing benchmark methods.