Large-Scale Multiobjective Federated Neuroevolution for Privacy and Security in the Internet of Things

With the development of communication techonologies, Internet of Things (IoT) devices will be deployed in more and more places and generate large amounts of data. The corresponding analysis of these newly available data will greatly improve our daily lives. However, these IoT devices can be attacked. Traditional intrusion detection systems (lDSs) usually use a centralized approach to transmit data to the cloud or a central server for analysis. This method has a great risk of privacy leakage. Federated learning (FL) is an excellent distributed learning stategy, which will have outstanding performance in preventing privacy leakage for lDSs in IoT. However, the FL-driven IDS is still in the infancy stage and needs further exploration. Therefore, we propose a large-scale multiobjective federated neuroevolution framework based on a deep fuzzy rough convolution neural network for IoT privacy and security.