Next-Gen Smart Healthcare Using UAV Assisted Cooperative Communication: A Deep Learning Approach

Unmanned aerial vehicles (UAVs) are evolving rapidly and are revolutionizing the operation of several applications. An important application that can be significantly improved with UAV-assisted communication is healthcare. Wearable devices can be worn on the body and are equipped with sensors that can detect and monitor various physiological parameters. The data collected by these devices can be used to track changes in health status over time, identify potential health risks, and inform clinical decision-making. UAV-assisted non-terrestrial wireless communication networks can play a significant role in the reliable transfer of this data. The novelty of the work is that it considers multi-antenna UAVs for communicating patient data collected from various wearable devices. It also proposes a hybrid relaying scheme incorporating deep learning (DL) for improved performance. The UAVs have pre-trained deep neural networks (DNNs) to process patient data and make accurate decisions without forwarding it to the healthcare facility. A two-layer deep neural network has been used at the UAV with a training accuracy of 99.6%, validation accuracy of 96.6%, and test accuracy of 96.62%. The UAV relays the data to the healthcare facility for expert opinion if it cannot make decisions with the desired accuracy. In this scenario, the optimum UAV and transmit antenna are selected, and a hybrid relaying scheme is used to minimize the network outage and maximize the throughput. The communication model is built, and extensive simulations are performed using patient data to demonstrate the approach. The work will significantly impact the use of UAVs for smart healthcare.