Research On Macroscopic Topology And Characteristic Of Statistical Time For Embedded Internet

The Internet has been growing rapidly in recent years, more and more embedded devices have been accessed Internet, which make man’s perception ability get unprecedented expanding. But real-time application of embedded devices is limited because of all kinds of uncertainty factors in the Internet. Facing the mass and complex Internet at present and timeliness requirements of embedded devices, the real-time interaction of embedded devices and Internet is confronted with greatness challenge. According to the finiteness principle of resources, the resources of time and space of embedded devices and Internet are limited. The guarantee of real-time has to take cost of spatial resources so that the solution of real-time question of embedded Internet (El) is the key to defining the type of cost. So research on macroscopic topology (as space) and characteristic of statistical time (as time) for El is essential and significant to the development of real-time application of Internet and embedded technology.Analyzed by application behavior of Internet, It is showed that El is a interconnection of the smallest physical network (as basic information) and the biggest logical network (as Internet) in the world. Lots of basic information is stacked when the scale of El is expanded gradually. In order to satisfy with the demand of real-time application behavior of El in this paper, the theory basis of real-time guarantee of embedded Internet device (EID) is proved to be fractal iteration of El macroscopic topology and power-rate convergence of network delay based on the research on mathematics theory of basic information of El. According to the above conclusion, with the aid in mathematical methods about fractal geometry, chaos theory, graph theory, topology, information entropy, stochastic process as well as statistical physics and so on, combining the results of statistical analysis of Internet macroscopic topology and network delay of Internet and EID, thus the research on macroscopic topology and characteristics of statistical time for El is carried out. Then according to the results, a theory and optimization algorithm of flow control is proposed in order to enhance real-time of El.Becaues El macroscopic topology reflects the whole performance of network, the study of El macroscopic topology is helpful in the design related algorithm of real-time. Based on the mathematics theory of basic information in the paper, further by means of fractal and chaos theory, through studying evolution mechanism of El macroscopic topology and statistical analysis of Internet macroscopic topology, El macroscopic topology is proved to have not only the same characteristic with Internet which is scale free, self-duplication and self-similar property, but also the different characteristic which is power-law fractal and chaos property. Then according to findings, with the aid in topology entropy principle, analyzed by stability of El macroscopic topology, power-rate transform mechanism and the best fusion algorithm of time and space as for the real-time interactive condition of EID accessing Internet is given.AS the characteristic of statistical time of El is an interpretation of real-time of EID, its study is helpful in the design related algorithm of real-time. Also EID usually must be transmitted the single direction control information, it must be used the clock synchronization to guarantee the definitiveness of one trip time (OTT) of EID. Therefore, analyzed by theory of NTP, it is showed that NTP can not achieve 100% success ratio and millisecond level precision of synchronization, a theory of end-to-end clock synchronization (ETS) is proposed hereby. Then according to statistical test and comparative analysis of OTT of Internet and EID, the feasibility and effectiveness of ETS is verified. Finally, a statistical property of self-similar and Long-range dependence of network delay of El is demonstrated.After analyzed by resources finiteness principle and network availability, it is obtained that flow distribution is an expression form availability of resources of network time and space. Then combining analysis of El Macroscopic Topology and Characteristic of Statistical Time, a self-similar and long-range dependence property of network flow is demonstrated. On the basis of this analysis, only if the question of flow distribution imbalance in the basic information is solved based on power-rate topology reliability of basic information, the behavior of real-time application of El can be satisfied. Then by means of maximum-flow minimum-cut theorem of graph theory in the paper, a MFMC flow control model is established, and the effectiveness of model is verified by example analysis. On the basis of establishing of optimization goal on this model, an optimization algorithm of flow control based on dynamic discrete particle swarm optimization is proposed. Analyzed by the convergence of DDPSO, this algorithm is verified to have a good convergence in the domain of definition. Finally, according to the experiment simulation of network POS machine, DDPSO is verified to be applicable in some scale basic information and can be used to solve the imbalance problem of flow distribution.