Seamless Gbps Hybrid FSO/THz Communication With Intelligent Switching

The advancement of hybrid free-space optical/radio frequency (FSO/RF) communication holds promise in improving link reliability and resilience against diverse weather conditions. However, the limited data rates of conventional RF bands, typically in the range of several Mbps, restrict the performance of high-speed wireless systems. To address this challenge, a Terahertz (THz) communication link is used to replace the conventional low-data-rate RF link and hybridize it with an FSO communication system to achieve fast data transmission. Specifically, a hybrid FSO/THz ($0.14\,\text {THz}$/$1550\,\text {nm}$) communication system is experimentally conducted and analyzed. Through experiments in a controlled foggy environment, a seamless data transmission of $1.6\,\text {Gbps}$ indoor environment over a distance of $20\,\text {m}$ is demonstrated. Moreover, to facilitate efficient decision-making between FSO and THz links, an intelligent switching strategy based on a deep neural network is implemented on field-programmable gate arrays (FPGAs) by incorporating the PYNQ framework. FPGAs offer a compact and energy-efficient solution, making them suitable for deploying machine learning models in resource-constrained environments. This approach ensures efficient link selection and maintains continuous connectivity without any outage.