| Tactical Internet plays an important role in digitized battle. However, due to its dynamic topology, self-organization, wireless communication, deployment of the battlefield frontier and so on, Tactical Internet is confronted with much more serious security threats than other military networks. Clustered network is a typical network structure of Tactical Internet, therefore, the research on secure clustering algorithm in Tactical Internet possesses an important practical significance.According to the characteristics of Tactical Internet and security challenges it faces, for effective managing nodes, reducing cost of control, and ensuring the security of communication environment, this dissertation researches, designs and simulates a new secure clustering algorithm in Tactical Internet. The main contributions are summarized as follows:1. Basing on investigating typical trust evaluation models and analyzing the trust demand in Tactical Internet, for the sake of synthetically describing the randomicity and fuzziness of trust, the cloud theory is introduced into trust evaluation, and definition of trust space, trust degree space, trust cloud, as well as trust grading cloud are given, a trust evaluation model is proposed based on trust cloud. For a two layer clustering network structure, the mechanisms of trust evidence gathering, trust computing, trust updating and trust grade determining intra and inter cluster are proposed, the proposed trust evaluation model is analyzed from aspects of veracity, practicability and efficiency.2. Aiming at the characteristics of military organizational system and the shortages of existing clustering algorithms in lack of security, a secure clustering algorithm applying in Tactical Internet is presented, into which the trust cloud based trust evaluation model is introduced. The algorithm consists of the original cluster formation phase and the cluster maintenance phase. The original cluster formation phase mainly includes cluster head selection, converge-cast and distributing gateway verdiction. According to different events which trigger the maintenance of cluster structure, the cluster maintenance phase is divided into trust based cluster maintenance and movement based cluster maintenance.3. Adopting network simulation tool NS-2, the trust evaluation based secure clustering algorithm for Tactical Internet (TEBSCA) is simulated and the performance is evaluated by comparing with some correlative clustering algorithms in stability, load balancing and network cost and compared with Beth trust model in security, the experiment results shows that the proposed algorithm is effective. |
