Abstract: Data forwarding in intermittently connected wireless networks requires multiple nodes to complete in a collaborative manner. However, malicious nodes can spread false recommendation information, which is able to seriously affect the choice of relay nodes. To address this issue, this paper proposes a data forwarding mechanism that can detect the forged data produced by malicious recommendation nodes by using the consistency of interactive list. Furthermore, to aim at differences of the recommendation information, this paper filters the error recommendation information to accurately evaluate the trust relationship between two nodes by the outlier detection method. Meanwhile, this paper also achieves the efficient data forwarding by selecting the relay nodes accurately based on similarities between a meeting node and a destination node. At last, the simulation contrastive analysis shows that the proposed mechanism can resist the malicious recommendation information effectively and improve the data delivery rate compared with other typical algorithms.

Key words: intermittently connected wireless networks, trust evaluation, bad mouthing attack, recommendation information