A Research On Design And Formal Construction Of Trusted Routing Protocols In Mobile Ad-hoc Networks

Mobile Ad hoc Network (MANET) is a new kind of wireless mobile network, andalso a new mobile computing platform. It is a multi-hop temporary autonomous systemwhich consists of mobile terminals with transceivers but no stations, and mainly usedin special or emergency circumstances, such as military battlefield, fire, flood fighting,etc. It is different from fixed computing platform that the platform is self-organized,the topology always changes, the terminal has the same role, the bandwidth is limited,controls are distributed, the safety is not guaranteed, and so on.With the growing dependence of security and reliability in networks, especiallythe mobile ad hoc networks, a new research field, Trusted Computing, was given birth.TrustedComputingisconsideredtobearevolutionarytechnologywhichfundamentallysolves the security problem. The core is to use a security chip, called Trusted PlatformModule (TPM), as a trusted root implanted into a variety of computer platforms, to cre-ate a security architecture for the application to provide a safe computing environment.Atpresent, domesticandinternationalresearchintrustedcomputinghasmadepre-liminary results, but the studies of trusted protocol and trusted software in MANET hasjust started. Representative is the security of MANET routing protocols. As MANETis open-ended and mobile platform is not trusted, related malicious attacks can not beavoided, such as tunnel attack, selfish node attack and vertex cut attack, etc. Therefore,we need to adopt a new strategy and approach to ensure that the routing protocol inMANET is secure and trusted.In this dissertation, we conduct research on trusted routing protocols and formalmodeling of trusted software in mobile ad hoc networks.Firstly, we propose a trusted routing protocol in MANET, named DAAODV. InDAAODV, the integrity of local routing software is protected by trusted computingtechnology, to restrict the behavior the routing nodes, and is attested to remote nodesby Directed Anonymous Attestation (DAA) and Property-based Attestation (PBA), inorder to solve the problem of the previous malicious attacks. As the using DAA forattestation, which does not need a Trusted Third Party (TTP) for authentication, therouting protocol avoids the bottleneck caused by TTP. We tag the node with PseudoName(PN)asitsidentity,andfortheanonymityofDAAandPBAprotocols,theroutingprotocolsisanonymous, whichiscapableofpreventingtrafficanalysisattacks. Wealsopropose a light-weighted signing and verifying algorithm for authentication of PN, to ease the overhead of DAA and PBA protocols, so as to avoid the related Denial ofService (DoS) attacks.Secondly, based on the work of DAAODV, we propose a formal method of mod-eling the mobile trusted software. By using the method, we can accomplish the designof the formal specification, refinement and implementation of the software, to ensurethe trustworthy on the software level of mobile protocols. In particular, we propose aformal specification of mobile trusted platforms, named TCSpec. TCSpec provides acall interface for security, and confidence to meet various of functional needs of mo-bile software. Then, a mobile trusted routing software is given as a case to show themethod of calling TCSpec and constructing the up-level software. By mathematicalproof, we conclude that the security of private information are preserved by using theformal specification in design of mobile software.Finally, we make further improvement on DAAODV, and propose a new trustedrouting protocol, FuzzyAODV, based on fuzzy set theory. The protocol is also basedon trusted computing technology, and uses DAA and PBA protocols for authenticationbetween nodes. But the format of messages is redefined, and the authentication processissimplified, whichimproves theefficiencythe trusted routing protocol. Meanwhile, inFuzzyAODV, we make extra consideration of malicious nodes impact on efficiency oftheroutingprocess. Weanalyzeavarietyofattacksfacedbytheprotocol. Thenbyusingthe fuzzy set theory, the nodes analyze and identify the behavior of malicious nodes,and use the corresponding preventive strategies to deal with these attacks. Finally, wemitigate the impacts of malicious nodes on routing efficiency.