Using deep learning to secure mobile crowdsensing

Researchers propose using deep learning (SAE, DNN, CNN) to enhance the security of mobile crowdsensing (MCS) systems, countering jamming, deception and forged sensing attacks. These methods are used to improve identity authentication, privacy protection, countermeasures against forged sensing, intrusion detection and anti-jamming transmission of MCS in IoT scenarios such as healthcare and traffic monitoring. The authors state that it achieves performance improvements over traditional security schemes, and note that practical deployment still requires addressing several challenges.