Dynamic Trust Mechanisms In Ubiquitous Computing Environment

With the continuous development of the computer, network, communication and integrate circuit technologies, both the software and hardware environment of the information technology has been changed a lot. Ubiquitous computing, as the third generation of the computing pattern after the Host Computing and PC Computing, has been one of the most important technologies in modern information science. It makes the information space based on the communication and computer gradually integrated with the physical environment in which people live and work. The emerging computing model has huge difference with the traditional closed and centralized computing model. In the open network environment, applications are open, dynamic and unknown object oriented. At the same time, entities in the network have more and more mobility and diversity, which makes the interactions between these entities with more uncertainty and uncontrollability. All these features have brought several problems, such as how to effectively identify the true and false among the numerous resources and entities, how to evaluate the trustworthiness of the counterparty in an interaction and how to make a decision when you have lots of optional objects and so on. All these problems involve the trust related issues in the open network environment which need to be deeply studied to design corresponding models and criteria.Therefore, this dissertation mainly focus on the key problems related to trust in ubiquitous computing environment。The ubiquitous computing environment is made up of ubiquitous computing device, ubiquitous network, and system software and so on, so we research the trust issues in ubiquitous devices(including wireless sensors and mobile devices), ubiquitous network(Ad Hoc) and resource scheduling in virtual systems respectively. The main results are shown in the following:We first work on the trust management scheme for clustered wireless sensor networks(WSNs). By studying and analyzing the existing literatures, we find out that in clustered WSNs, a cluster head’s(CH’s) trust value depends mostly on other CHs’ trust ratings based on its performance as a service provider, while ignores the sensor nodes’(SNs’) ratings within the cluster considering the cluster head as a service provider and supervisor. In fact, each CH not only collaborates with other CHs, but also with the sensor nodes within the cluster. Considering this feature, we establish a mutual evaluation mechanism between the CHs and SNs. We construct the Petri Nets(PNs) model for the sensor nodes’ possible actions in our scheme, and refine the model to obtain a deadlock-free, bounded, and reversible model. This mechanism can not only evaluate each SN’s trust value to increase the interaction success rate, but also assure the fairness of the recommendation of a CH to its SNs. In addition, we give a trust based method to select a new CH to replace the bad one and reconfigure the SNs belonging to the excluded one.Mobile devices have been becoming more powerful and are increasingly integrated in our daily life, so the security of a mobile device is therefore increasingly linked to its context. Device comfort is a new concept which enhances the concept of trust. It evaluates the mobile devices security from the device’s perspective, which enables mobile devices smart enough to perceive its current context and synthesize the cognized cues, then use the internal models to reason about its security status under the cognized context(including its user). We propose the computational method to assess the sensitivity of a specific action running on the mobile device to the context. In addition, we propose two methods for evaluating the situational comfort level of a mobile device. The first proposed method uses the predefined ideal context as a standard to assess the comfort level of a device in the current context. The second evaluation method can monitor the status of the mobile device continuously rather than enforcing a static security policy used in traditional access control methods, which allows better reasoning about the risk of running an action in a certain context.And we work on the selfish phenomenon in the indirect trust evaluation process for mobile ad hoc networks(MANET). A game model is proposed for the indirect trust information obtaining process. We give a suitable punishment mechanism and design the utility function for nodes in a MANET. Moreover, with evolutionary game theory, we analyze the trust evolution of nodes using replicator dynamic equation and propose several theorems to illustrate preconditions under which each evolutionarily stable strategy can be achieved. Experiments verified the correctness and effectiveness of our method and simulated the dynamic evolutionary trend of the nodes under different parameters, which experimentally proved the proposed theorems.We also work on trust based recommendation mechanism for the virtual community in the Internet of Things. We proposed a recommendation algorithm which takes the community benefit and the stability of the entities’ trust value into account. The community’s control center adopts the Lyapunov optimization method to make decision which can give it maximum benefit on the premise of satisfying the desired constraint of the interaction requester, meanwhile guaranteeing the stability of the community nodes’ trust value.