Research On Self-Aware Network Oriented Test Key Technologies

Along with the ever-increasing agility of the network, more and more researches on awareness technologies have been carried out. Awareness technologies of networks play a fundamental role in the highly smart ad hoc network. Through the awareness of the properties of the networks and the associated services carried by the networks, the automation of the operation, maintenance and administration of the networks is achieved without manual intervention anymore. Modeling of the aware networks, the standardization of the parameters that need to be aware and the corresponding test technologies are not only the hot topics in current researches on aware networks but also the key to the implementation of the self-aware networks.In this thesis, the model of the self-aware networks is built and the associated aware technologies are studied in different layers, including the physical layer, the network layer and the application layer. In terms of software engineering, parameters of the self-aware networks in separate layers and the corresponding detection and test techniques are studied. With the advantages of the software defined networks, innovation of the test techniques is achieved. The main research work is divided into the following several parts(1) Research on the model of self-aware networks. According to the diagram content and the queuing theory, the mathematical model of the content aware networks and their applications are analyzed and built. The application of detection techniques with the depth package inspection (DPI) and the inspired behavior inspection combined, can effectively solve the problems of the priority identification of the blocking services in conventional networks. Meanwhile, the priority of services in high priority levels and the fire use in resources is guaranteed.(2) Research on the main performance indexes and the associated test technologies in the physical layer. Practical test tool on the bit error ratio (BER) is designed and developed, which can emulate different BER levels. Moreover, the tool can be used not only as a test instrument solely, but also probe in content aware optical networks, and the verification platform of the optical performance monitoring (OPM) techniques. The test solution of chromatic dispersion in single mode optical fiber using auto-correlation function is proposed and validated experimentally.(3) Research on the test technologies of the performance indexes in the network layer of the communication networks. Based on the existence of a lot of mixed flow in the practical environment, in order to evaluate the real processing ability of the nodes of the network, test model of the performance indexes is built and the evaluation methods of the main parameters of the network nodes is given, including throughput, transmit ability of the mixed packages and rate of new connection.(4) Research on the test technologies in the application layer. On the one hand, detection technique of exceptional flow is studied and detection algorithm of the exceptional flow mainly based on principal component analysis is proposed; on the other hand, based on software defined network (SDN), test tool of the application layer of the network is developed. This test tool is aimed at the scenarios of various communication networks, OPM, simulation of the applied flow and is applicable in the monitoring and test system of the next generation of optical network.(5) Innovation of test technologies based on the software engineering. In terms of software engineering, the threats faced by the self-aware networks and the corresponding elimination measure are analyzed. Aiming at the safety problems faced by the aware networks, researches on the safety test work of every stages are carried out from the test technique of the software engineering, which is in combination with the design of the test use cases based on the threat modeling and the methods of the safety test. In conclusion, this thesis has focused on the modeling of the aware networks and the researches on the test technologies of the crucial parameters of the self-aware networks and its layers, which are from the test of the special index to the level of the software engineering. Combined with the practical engineering applications, aiming at the practical problems to be solved, the corresponding mathematical models are built. Moreover, the development of detailed test tools and the researches on the test of the detailed indexes are carried out. On the basis, several innovative research ideas and solutions are proposed and validated by experiments, which will provide technical support for the next generation agile optical networks.