An Autonomous Decentralized System Architecture Using a Software-Based Secure Data Field

Autonomous Decentralized System (ADS) has been proposed for providing on-line expansion and maintenance and also fault-tolerance. The associated architecture is composed by a Data Field (DF - on this element, the information is flowing among all the elements connected to it) and Autonomous Control Processors (ACP - part of an entity where the information coming from the DF is preprocessed and manages data consistency and fault tolerance) with their Application Software Module (ASM - part of an entity which processes the information belonging to its own tasks). We are developing an ADS Infrastructure that will permit us to integrate heterogeneous systems with the characteristics provided by this paradigm. We develop a Software-Based DF composed of Nodes. Each Node is a software application with a simple communication protocol and a basic management user interface. The nodes, all together, are responsible for forming a network communicating one with each others. The main task for each node is to identify any entity and redirect messages for delivery. In our infrastructure, the DF is the core of the architecture and since it is a software implementation there may exist several security risks such as intruders or apocryphal components which can pretend to belong to the system. In this paper we show the technology proposed for providing a secure DF which allows to identify and authenticate each component that pretends to connect to the DF.