A network coding transmission strategy for preventing eavesdropping attacks

For secure network coding to prevent eavesdropping attacks, a general design is used that puts restrictions on the eavesdropping channel set or encryption technology is embedded, which can cause problems in practical applications. Based on random linear network coding technology, in this paper, a transmission strategy to prevent eavesdropping attacks is proposed for single-source multicast networks, without either precondition of restricting the eavesdropping channel set or embedding encryption technology. The data transmission process is divided into three phases, and the global encoding vectors are separated from the encoded data on the channel. The proposed strategy can enable the sink to decode successfully, making it very difficult for eavesdroppers to meet the conditions for successful decoding by randomly selecting the eavesdropping channel set without obtaining network topology knowledge. The feasibility of the proposed method is strictly demonstrated using layered network technology, and its security is analyzed. The simulation results confirm the validity of the theoretical analysis.