Notably, the model parameters are reduced to just 0.306M, representing an 86.3% reduction compared to current FL methods. By enabling privacy-preserving and efficient intrusion detection, FL-TET has ...