Efficient and secure information sharing in wireless and distributed networks

Abstract

Collaboration is the process of two or more participants working together to achieve a common goal. For productivity improvement in a during collaboration process, all participating in the collaboration have to share the related information and finite resources efficiently required to accomplish the goal. However, it is hard to accomplish efficient collaboration due to several reasons

geographical distance between collaborating participants, diverse range of working tools, different job method, and evasion of information sharing for the fear of information leakage. Therefore, the needs are growing for the study of the methods to transmit the information effectively over wireless and distributed networks and to share the information securely. In this thesis, three approaches are proposed to maximize the collaboration effectiveness by improving the efficiency of information sharing between the participants or the automated working tools and by ensuring confidentiality of the information. First, a video streaming scheme considering video complexity over a cooperative multi-hop relay network is proposed. Video stream is a kind of representative large-volume information usually shared during collaboration. The proposed model can improve the network performance using an adaptive routing scheme taking into account the motion level of each video stream when video streams with heterogeneous motion levels are transmitted over multi-hop cooperative relay network. A video distortion model is developed taking into account video complexity of the video sequences to measure the improvement of the transmission efficiency, The experimental results show that network performance improves when video sequences are routed considering heterogeneous motion levels with the object of maximizing the minimum peak signal-to-noise ratio for the video streams. The adoption of cooperative relays and hop-count limitation can also improve the routing performance of video streams. Second, the detection mechanism of black-hole attacks attempted by malicious participants is proposed, when participants are sharing information over mobile ad-hoc network with the form of wireless and distributed network. The proposed mechanism uses a tandem nodes, special nodes for the black-hole attack detection and the tandem nodes consistently sniff all data traffic transmitted between the neighboring nodes. When a suspicious node with abnormal amount of incoming or outgoing traffic is detected, the tandem nodes certainly determine the black-hole node cooperating with other tandem nodes. This mechanism can minimize the change of the existing network topology since it does not require any routing protocol modifications or additional deployment of a detection system. Instead, only a few tandem nodes can get a high detection performance with only very low false positive. Third, a framework is proposed which enables all the members participating large-scale collaboration over distributed network to access and mange product information conveniently and securely. The proposed framework can easily to delegate the right of access to information to others. In the proposed system, permission-based access control scheme with lazy revocation is used where each set of product data is encrypted and stored in data storage. The existing mechanisms are not suitable to be applied them to a large-scale environment since they usually generate high expense, and require many experts. In comparison with the existing ones, the proposed mechanism can operate smoothly in a large-scale collaboration environment.