Security is one of the most important issues [21, 77] to be resolved in MSNs. There exist many kinds of attacks such as the Sybil attack, the flooding attack, etc. In addition, users may be cheated by malicious users who forge some contents in the system. For example, service evaluation applications enable users to share their experiences or reviews about services they are given. However, Sybil attackers may exist in this application and frequently change pseudonyms or identities to repeatedly broadcast the same or similar information [30, 78]. They may mislead the user’s opinions and preferences [79]. Besides, the flooding attackers can exhaust the resources of users by transmitting mass data packets. As users have the limited resources such as the battery power, memory space, bandwidth and so on, how to propose security strategies against these attacks should be studied.

Zhang et al. [30] classify the Sybil attackers into four levels according to their attacking capabilities and propose a Sybil detection scheme to distinguish attackers in MSNs. Abbas et al. [80] present a lightweight Sybil attack detection method for MSNs, in which the entry and exit behaviors of all identities are studied to detect the Sybil identities. Yu et al. [81] propose a defense method against Sybil attacks called SybilGuard in which the social graph between the identities is used to detect the fake identities. Quercia et al. [82] identify Sybil attackers by allowing each mobile user to manage two small networks which are network of friends and network of foes, respectively. And the network information can be shared when two users have a contact.

Zhang et al. [83] study a flooding attack in MSNs where the malicious users broadcast mass data packets to other users and propose a generic defense with a threshold against the attacks. Yi et al. [84] discuss the influence of flooding attacks and analyze the network’s performance under different degree of flooding attacks.

Kim et al. [85] propose a period based defense mechanism against the flooding attacks and each user in MSNs has a blacklist to prevent the attacks. Fallah et al. [86] consider the resources of both the normal users and the malicious users. They propose an optimum defense mechanism which can provide the maximum possible payoff for normal users based on the game theory.

< Prev   CONTENTS   Source   Next >