Study On Cooperative Autonomous Mechanisms Of The Internet Of Things Based On Biological Intelligence

As a future development trend of the communication network technologies, the Internet of Things (IoT) is an integrated interdisciplinary research area which faces many difficulties and challenges on designing, researching and realization. There are a lot of basic scientific problems to be solved, e.g. complexity of the large-scale network, dynamic characteristics of network environments, efficient control of massive heterogeneous network elements, resource and energy limitations, etc.. With the rapid development of biology science, the biological intelligent theory, which is inspired by the life regulation and control mechanisms in nature, has become more developed, and the application and popularization have already come into a new stage. Many bio-inspired methods endow systems or networks with the heterogeneity, autonomy, autonomous, dynamics and scalability, which can be references and inspirations for designing new efficient network architectures and mechanisms. Taking the features and requirements of the IoT into account, this work focuses on the issues of data exchange and interaction among heterogeneous network elements, distributed self-organization schemes and intelligent service provision in the dynamic system environment, and tries to solve these problems by applying biological intelligence methods and theories, which can lay the foundation for the cooperative autonomous theory of the IoT. The main contribution of the thesis lies in:(1) A summarization of the features, challenges and relevant research work of the IoT is provided. The feasibility of the biological intelligence is analyzed in solving the problems in the communication networks like the IoT, especially the regulation principles and mechanism in maintaining the homeostasis of internal environment by the endocrine system and neural system. This thesis also provides the methods of the researches on the interactions among network elements, self-organization, maintenance of network’s stability and scalability, and service request response methods.(2) An information interaction mechanism among the IoT elements is designed inspired by the hormone transmission and regulation principles in the endocrine system. Based on this, an endocrine-based intelligent distributed cooperative algorithm (EIDCA) is also given for object tracking in the wireless sensor networks, and a numerical evaluating method is designed to provide an explicit measurement for comparison of different algorithms. Experiment results show that with the proposed EIDCA, the network elements can exchange information and interactive efficiently in a distributed way and the EIDCA accomplishes the tracking mission for the invading object more efficiently and persistently compared with other algorithms.(3) An intelligent self-organizing scheme (ISOS) for the IoT inspired by the endocrine regulating mechanisms of blood glucose regulation is proposed. For each node in the network, an autonomous area is established, where the node can effectively interact with its peers and perform self-control according to its own status and dynamic circumstance in a decentralized infrastructure. To act as the media for the information transmission and data sharing among the nodes, the concept of the hormone with different purposes is introduced and implemented numerically, and the nodes can therefore collaborate with each other and worked in a cooperative way. By adjusting the releasing procedure of the hormones, the ability of effectively detecting service requests randomly generated can also be guaranteed in the probabilistic partially-working IoT. A series of scenario-based simulations with results verifies the performance of the proposed mechanism entitles the IoT with the ability of maintaining its status in a globally stable status and effectively discovering the random service requests in a resource-critical configuration, which would be of great significance for the practical implementation of the IoT.(4) Inspired by the neuroendocrine comprehensive regulation mechanism of maintaining the homeostasis in human body, an IoT intelligent service model and dynamic service request detection algorithm (HITS) for realizing "always response service" in the dynamic environment are proposed. A dynamic information exchange scheme plays a critical part in the HITS to guarantee the rapidness of the network on responding the randomly generated service requests. Simulation results show that the proposed HITS algorithm can be regarded as a qualified instance of the distributed IoT infrastructure. The application of the HITS algorithm entitles the network with intelligent decision-making and cooperative working abilities, by which the autonomy of the nodes, the stability of the network and the effectiveness for providing service to the potential service-in-need targets can be guaranteed simultaneously.Finally a conclusion is made for the whole contents of this dissertation, together with the perspectives of this field for the next step.