A Heterogeneity-Aware Semi-Decentralized Model for a Lightweight Intrusion Detection System for IoT Networks Based on Federated Learning and BiLSTM

. 2025 Feb 9;25(4):1039. doi: 10.3390/s25041039

Algorithm 5: Semi-decentralized FL client training.

Input: $W_{G}$ : Global Model Weights,
$N u m b e r O f E p o c h s$ : Number of Local Epochs
Output: $W_{i}$ : Client Model Weights
₁
Receive $C l i e n t M o d e l$ from the $C l u s t e r H e a d$
₂
Receive weights and parameters from the $C l u s t e r H e a d$ ( $W_{G}$ , $N u m b e r O f E p o c h s$ )
₃
Train each $C l i e n t M o d e l$ with received weights and parameters
₄
$W_{i} \leftarrow$ Get weights from $C l i e n t M o d e l$
// Return updated client weights $W_{i}$ to the $C l u s t e r H e a d$ for aggregation
₅
return $W_{i}$