Research On Three Common Features Of Information Networks

Although every information network behaves quite differently from each other in many aspects, such as data traffic, node property, and connect relations, and meantime they suffer from diverse technical problems, one will find some common features behind those related or unrelated problems existed in different information networks when reviewing them from a new perspective. These common features drive technologies to be updated endlessly and drive puzzles to be solved constantly. When the existence of these common features is recognized and the mechanisms leading to these common features are understand, it is expected to get some inspirations from other solutions of the similar problems in other information networks. It is helpful to comprehend the reasons resulted in technical problems more clearly. Following this idea, this study focuses on Internet, mobile ad-hoc network, vehicular ad-hoc network, software network, e-commerce network, and other information networks with different applying domains. After analyzing some related problems existed in these networks such as route, robust, topology modeling, and security, three common features are concluded, furthermore, some researches on the problems related with these common features are carried out.Many propagation behaviors can be seen as obeying a certain rule, such as congestion spreading in Internet, failures extending in software network, and worms disseminating in vehicular ad-hoc networks. The scientific researchers from different domains including mathematics, physics, and computer science are seeking to describe such propagation, uncover the factors affecting such propagation, and find a reasonable controlling method on such propagation. This research is important for understanding propagation behaviors and controlling methods on congestion, failures, and worms spreading in different information networks. The finding of characteristics of complex networks provides a new theoretic tool and technical method to investigate propagation behaviors. Many empirical studies show many social networks and technical networks have small-world (big clustering coefficient and small characteristic path length) and scale-free (degree obeys power-law distribution) characteristics. This work focuses on researching on congestion in Internet, failures in software network, and worms in vehicular ad-hoc networks, and builds their propagation models, respectively.Many evolution processes are dominated by a certain rule, such as Internet extending constantly, e-commerce network changing continuously, so these processes are predicted in a extent not random (random is just a appearance )。Only those network structures with some special pattens can hold on for a long time in order to spontaneously evolve to a new pattern (one of a group of special patterns) when receiving outside force. The complexity of network structure intensifies the non-linear affections between nodes, resulting in the research on the evolution of network structure converting from particle, simplicity, and determination to holistic, complexity, and uncertainty. Modeling the evolution of network structures can capture the dynamics of network development, understand the effects of every micro factor on changes of network structures, and help human grasp network functions and dynamics characteristics. Therefore, this work pays attentions to the evolutions of the router-level topology of Internet and trust network between individuals of e-commerce network.Many problems in mobile ad-hoc networks and Internet are closely related with trust such as route, security distribution, and access control. Their goals are to build a network environment with high trust degree. Until now, their trust is not guaranteed completely yet due to the openness and anonymity of popular Internet and rising MANET. Generally, the protocols of link layer and network layer assume that every node is trusted, however, this assumption does not hold in many situations where attackers are easy to destroy network through disturbing protocol regulations. The lack of a kind of satisfied trust has heavily blocked the further development of Internet and mobile ad-hoc network, and has affected the recognition extent of applications. This work concentrates on trust problems including routing and multi-secret sharing in MANET and access control in Internet.The main contributions and novelties of this work are summarized as follows:(i) In the propagation aspect, Internet: a congestion function and the concept of"delay time"are introduced. And a new evaluation function of network efficiency is given. Based on Statnet and Webgraph topologies the effects of network structure and size, delay time, processing ability and traffic generation speed on congestion propagation are investigated. Also congestion propagation process composed of three phases and some factors affecting transition phenomenon are uncovered. Software network: a topological model in which directional edges and the corresponding weight values are used to represent calling relationships and close degree among functions is proposed. By introducing the function fault-tolerant capability and the software fault intensity, a software fault propagation model is designed to explore interactions between fault propagation dynamics and software parameters. Vehicular ad-hoc network: the theoretic spreading results under urban scenario with equilibrium traffic are analyzed through modeling mobility pattern, communication channel, medium access control, worm propagation, and others. The extensive Monte Carlo simulations uncovered the impacts of the transmission range, the minimum velocity and the maximum velocity, and the vehicle density on epidemic spreading of such worms in vehicular ad-hoc networks. Further, the wireless worm propagation in the same scenario with dynamic traffic is simulated by using network simulation software, the correlation between theoretic results and practical outcome is discussed, the reasons resulting in local differences are also analyzed, and finally, the interconnection of affected rate and other network parameters is pointed out.(ii) In evolution aspect, Internet: the concept of"attraction degree"is introduced. And the idea that the growth of Internet topology is attributed to interaction between internal factors and external factors of node is pointed out. A new modeling algorithm for Internet router-level topology is proposed by considering the influence of both node property evolution and geography limit. E-commence network: by introducing the concept of attribute similarity, an attribute matching mechanism is designed. Based on attribute matching mechanism a feasible bring-and-buy scheme in E-Commerce is proposed, which is expected to give guidelines about how the aid's decision-making process is designed to improve users'perceptions of usefulness and trustworthiness of on-line decision aids.(iii) In trust aspect, MANET: on one hand, different from most the past routing algorithms only considering the shortest path first, or evaluating the trust level solely on the forwarding behavior, or depending on recommendation-based trust mechanism, we propose a novel subjective trust routing model under consideration to both communication reliability and path length. Inspired by the establishment pattern of trust relations in human life, we introduce the concept of attribute similarity to inherently bind neighbor selection, trust evaluation, packet forwarding, and other routing processes. Based on the similarity degree, we put forward a new forwarding behavior without needing handshake process and decrease the transmission delay, so a subjective trust routing is achieved. We give a recommended method to calculate the similarity degree between attributes. On the other hand, a new verifiable multi-secret sharing model based on Shamir secret sharing is proposed and realized. Internet: a framework of hierarchy trusted access and remediation model based on the grade division were put forward, and the partition rules of trusted attributes, the methods of grade division as well as the work flow of remediation model were explained in detail. Moreover, the potential applications and the access procedure of terminals were discussed.