BFLP: An Adaptive Federated Learning Framework for Internet of Vehicles

Algorithm 4

Federated transfer learning.

	Input:, ST
	Output: lServer−RSU
	1: V and T do
	2: Initialize required by ;
	3: Distribute the public keys τV and τT;
	4: Use fully homomorphic encryption to encrypt ;
	5: Send to Server-RSU;
	6: Server-RSU do:
	7: Distribute the public keys τRSU;
	8: while the model is not convergent do
	9:     Server-RSU Execute:
	10:  Decrypt ;
	11:  Calculate , , ∂T;
	12:  Use fully homomorphic encryption to encrypt , ∂T;
	13:  Send [] to V and send [∂T] to T;
	14:  Send to Server-BC;
	15:  V do:
	16:  Decrypt [∂V] locally;
	17:  Calculate γV;
	18:  Use fully homomorphic encryption to encrypt γV;
	19:  Send [γV] to Server-RSU;
	20:  T do:
	21:  Decrypt [∂T] locally;
	22:  Calculate γT;
	23:  Use fully homomorphic encryption to encrypt γT;
	24:  Send [γT] to Server-RSU;
	25: end while