In the era of 5G and big data transmission, heat dissipation challenges for critical components such as CPUs and bi-directional optical sub-assembly (BOSA) modules must be addressed. In this study, we developed an intelligent, sensor-integrated cooling design that combines high thermal conductivity materials with topology-based flow field optimization to enhance thermal reliability. A hybrid structure of natural and forced convection was employed, regulated by embedded temperature sensors and a microcontroller. When component temperatures approached their operational limits (90 ℃ for BOSA and 105 ℃ for CPU), the system activated a 6.0 mm fan to ensure safe operation. Experimental validation tests and high-fidelity computational fluid dynamics simulations were conducted using calibrated K-type thermocouples and the SIMPLEST algorithm, respectively. The mesh-independent simulation results confirmed numerical accuracy, with only 0.7% error relative to experimental data. Topology optimization was performed to identify the optimal CPU heat sink parameters (fin height = 20 mm, fin width = 1 mm, and number of fins = 9), reducing CPU temperature to 93 ℃. The novelty of this study lies not only in extending prior research on laptop thermal management (primarily focused on CPU-only systems), but also in further applying the control strategy to the thermal design of high-heat-generation routers that incorporate both CPUs and bi-directional optical sub-assembly (BOSA) modules in practical applications. The results demonstrate that shifting sensors from passive monitoring to active regulation significantly improves cooling efficiency, offering a robust framework for next-generation portable electronics under high thermal stress. Future research may further advance this approach by developing optimal fan cooling strategies that dynamically adjust fan speed in response to real-time temperature variations. In addition, given the impact of fan noise on user comfort, subsequent studies should integrate acoustic modeling with thermal performance simulations to achieve a balanced design that simultaneously satisfies both noise constraints and cooling efficiency.