The common phenomenon of urban traffic congestion leads to the loss of time as well as pollution, which consequently affects the environment. We introduce here a new approach, which is called Discrete Structure – based Traffic Optimization (DSBTO), which uses network algorithms and high resolution traffic information to reduce congestion. Road network is modeled as a discrete graph, and optimization strategies are used to optimize traffic flow across the graph. In order to verify our method’s effectiveness, we use the Glasgow High-Resolution Traffic Flow Dataset, which contains fine-grained traffic measurements over several years for calibration and validation under real traffic conditions. Our approach employs DSBTO to identify significant traffic bottlenecks and identifies optimal flow modifications or control mechanisms that reduce congestion times without compromising the capacity of the network. During our method testing, we discovered that average travel times were reduced by about 15%, and network throughput increased compared to normal traffic. The findings demonstrate the potential of using granular transportation data as well as discrete optimization tools to have a favorable effect on urban traffic management. Theoretically, DSBTO might translate to improved traffic flow and reduced surge and stop congestion thus resulting to reduced emissions from vehicles in cities. The suggested system can easily be integrated into the existing traffic control infrastructure to offer data-based support process with the purpose of reducing congestion and increasing mobility.