Research Article
A Homomorphic Signcryption-Based Privacy Preserving Federated Learning Framework for IoTs
Algorithm 1
Model training phase of client
. | Input: Training round , model parameters , dataset ,quantization bit width , blinding factors , , random seed , Paillier key pairs , secret value | | Output: Signcryption of the masked gradients | | functionMODELTRAINING | | Compute gradients based on and | | Send layer-wise gradients Max-Min values and sizes to the aggregation server | | Clip with corresponding threshold (Advance Scaling) and quantize them into bits | | Batch the quantized gradients layer by layer into | | Blind with to compute | | Sign each blinded gradient with , use the PRG to generate the synchronizing random number , compute the signcryption . | | Send to the aggregation server | | end function |
|