: A mobile ad hoc network (MANET) is a self-organizing network comprised of mobile nodes that communicate with each other via wireless links without any fixed infrastructure. Securing the exchanges in such network is required to guarantee a widespread development of services for this kind of networks. Inherent vulnerability of mobile ad hoc networks introduces new security problems, which are generally more prone to physical security threats. The possibility of eavesdropping, spoofing, denial-of-service, and impersonation attacks increases. Similar to ?xed networks, security of mobile ad hoc networks is considered from different points such as availability, con?dentiality, integrity, authentication, non-repudiation and access control. However, security approaches used to protect the ?xed networks are not applicable due to the salient characteristics of mobile ad hoc networks, such as weak-secure wireless links, unpredictable and freely movement of mobile devices and dynamic topology of network. The deployment of any security service requires the de?nition of a trust model that de?nes who trusts who and how. In wired networks such as Internet, a centralized trusted third party (TTP) is responsible for determine security policies. However centralized TTP can cause some problems such as single point of failure and some limitations for accessibility. Therefore it is proposed use of distributed TTP. The threshold cryptography is a good solution to the problem of deployment a TTP in MANET. In a distributedTTP the functionality of TTP is distributed between some special nodes using a threshold secret sharing scheme. In threshold secret sharing schemes a secret is distributed to multiple nodes in a manner that the secret is disclosed only if a threshold number of shareholders are compromised. Secret sharing alone does not defend against mobile adversaries which attack, compromise, and control one server for a limited period before moving to another. The defense here is proactive secret sharing, whereby servers periodically create a new, independent secret sharing and then replace old shares with new ones. In this thesis, at first, we indicate a specific attack to the pervasive proactive secret sharing scheme proposed by Herzberg in 1995 and show that this scheme and its successors are inapplicable for using in MANETs. Then a secure and scalable proactive secret sharing scheme is presented to eliminate the weaknesses of Herzberg scheme. In comparison to Herzberg scheme and its successors, in the proposed scheme server nodes in every subset comprised of threshold number of server nodes can refresh their shares independently. Moreover, in the proposed proactive secret sharing scheme server nodes can verify the correctness of received shares without requirement to broadcasting witnesses. It leads to lower communication and computation overhead, no need to knowledge of global network topology and compatibility with dynamic network topology. Therefore our proposed proactive secret sharing scheme will be scalable and efficient. Then a robust and distributed trust model for mobile ad hoc networks is proposed, in which the proposed proactive secret sharing is used to defend against mobile adversary. Keywords : Mobile ad hoc networks, Distributed trust model, Distributed certificate authority, T hreshold cryptography, Proactive secret sharing.