Hardening Trust Models against Slandering Attacks in Relayed Content Delivery Services

There are a number of application contexts in which services are delivered by a given provider to some clients through other relay clients in a collaborative fashion. This is, for example, the case of sensor networks, vehicular networks, D2D, and so on. In this case, a security problem arises. Indeed, when a service is relayed by a client, it is not sure that it is relayed correctly. Therefore, the final client could be deceived if malicious relay clients exist. The classical way to contrast this problem is to use a trust mechanism, managed by the provider, based on the feedback returned by the clients. Thanks to this mechanism, the trust of malicious relay clients can be decreased, then reducing (even cancelling) the negative effects of these clients in the community. The trust mechanisms of this type often suffer from weakness against slandering attacks. Dishonest final clients can fraudulently decrease the trust of relay clients, by reporting a false feedback. In this paper, we propose a general approach to fortify the trust mechanism against this kind of attacks.