Delocalization of a Dynamic System with Preservation or Self

We analyzed the problem of a dynamic system delocalization due to changes in the system environment - universe and system architecture. We developed a Delocalization of Dynamic Cores model to analyze the migration of functional properties in open informati

DELOCALIZATION OF A DYNAMIC SYSTEM WITH PRESERVATION OR SELF-RESURECTION OF INFORMATIONAL FUNCTIONALITY

Ghost machines or can a system survive its destruction Vadim Astakhov, Tamara Astakhova UCSD , vadim_astakhov@http://doc.guandang.net Abstract

Keywords: dynamic core, informational geometry, architecture, communication, holography, dynamical system, resurrection INTRODUCTION In the modern science and engineering, open information and dynamic systems such as various computer networks, biological neural network, gene network, social network and many others provide infrastructure for global communication and information sharing within a specified domain. These systems strongly depend on topological properties of underlying networks as well as on dynamic properties represented by hierarchy of communication pathways. These networks are not uniform sets of communicating units. They usually have hierarchical structures in which some nodes produce much more complex dynamic behavior compared to others. The situation becomes even more complicated at the network to network communication layer for brain neural networks, interactions among social groups or enterprise software systems where various nets run distributed applications such as distributed relational databases, grid, and J2EE/EJB and .NET platforms while others run thin client applications like browsers. Current mathematical formalism to analyze these communication systems initially evolve with an implicit assumption that we can call “point-to-point” communication paradigm. This implies a simple use-case in which one system element sends an information stream to another one through the network. That paradigm is fine and lead to great achievements in the world of software engineering such as client-server architecture in which most applications operate as either a client or a server. The underlying global network architecture and protocol dynamics imply many assumptions for the design of such systems. We argue that even if this paradigm applies for many applications it is not successful in some large-scale systems. We demonstrate that global properties of the system and surrounding environment might lead to conservation of the system functionality through the process that we call “Delocalization Of Dynamic Cores”. We will demonstrate several examples of peer-to-peer communication systems as well as examples from physics and neural network dynamics.

1 We analyzed the problem of a dynamic system delocalization due to changes in the system environment - universe and system architecture. We developed a Delocalization of Dynamic Cores model to analyze the migration of functional properties in open information and dynamic systems undergoing architecture transition and modifications. Information geometry and topological formalisms are proposed to analyze informational dynamic systems. Different physical and holographic models are proposed to construct systems able to conserve their functional properties under delocalization transition. We found several constraints for the system environment - universe which conserve the dynamic core system functionality under transition from localized explicit implementation of functions to the implicit distributed implementation.

We analyzed the problem of a dynamic system delocalization due to changes in the system environment - universe and system architecture. We developed a Delocalization of Dynamic Cores model to analyze the migration of functional properties in open informati

DELOCALIZATION OF DYNAMIC CORES With growing interest to ultra-large-scale distributed systems such as an internet, biological and social networks the global properties and architecture of the environment should be reviewed to support the software challenges of the future. That will lead to various implications for underlying network topology and design of network communication protocols. To address those challenges, we develop a formal mathematical model that will help us analyze the complex dynamics of information processes. We extended and modify some methods from the theory of dynamical systems to make it applicable for analysis of ultra-large-scale systems. A dynamical system is a mathematics concept in which a fixed rule describes the time dependence of a point in a geometric space. We employ the concept of “informational geometry” [1] as the geometric space and introduced an informational manifold and system state as a point on the manifold to describe the dynamics of information processes. A system state is determined by a collection of numbers that can be measured. Small changes in the state of the system correspond to small changes in the numbers. The numbers are also the coordinates of a geometric space—a manifold. The evolution rule of the dynamical system is a rule that describes what and how future states follow from the current state. The first problem we are interested in is to analyze conservation of information properties of a dynamic system and survival or self-resurrection of those properties under the transition of the underlying physical infrastructure from one medium to another. Tornado gives us a simplest physical example of the dynamic system with changing elements but conserved functionality such as a rotor of particle velocity. The more complex examples of the function transition in computer science are migration network from a wired to optical network or from the IPv4 to IPv6 protocol. These migrations usually encapsulate the details of the underlying medium architecture and do not effect higher level applications. Such an approach usually requires partitioning of layers a …… 此处隐藏：42739字，全部文档内容请下载后查看。喜欢就下载吧 ……