Federated Intrusion Detection in Blockchain Simulation | Neural Network for Edge Assisted 6G - IoT

We create a network it consists of 100 IoT devices, 4 edge servers, 1 global server and 1 blockchain node. We perform the efficient client selection process based on the trust,energy,bandwidth and network conditions using Auction game theory algorithm. Next, We perform the process of base criterion method (BCM) to select the optimal secure channel based on the noise, path loss, channel quality, stability, trust and fading parameters. Next, the gathered data from the clients are transmitted to the edge server to train the data for creating local models.The optimized back propagation based deep belief network algorithm is used to train both local and global models. The deep learning parametrs are optimally selected using pelican optimization algorithm (POA) in which the POA algorithm has a high convergence rate that provides rapid performance. Finally we plot the results graph for, Number of Epochs vs Accuracy, Number of IoT Devices VS Accuracy.

Видео Federated Intrusion Detection in Blockchain Simulation | Neural Network for Edge Assisted 6G - IoT канала PhDprojects. org - Ideas For Growing Your Career