The main objective of the proposed research is to develop novel Software-Implemented Fault Tolerance (SIFT) techniques for increasing the dependability of the distributed and networked environments based on COTS. The issues of dependability will be addressed by the design and implementation of (SIFT), dynamic reconfiguration in distributed systems and high-speed computer networks. Novel fast dynamic reconfiguration techniques for various irregular network topologies will be developed and analyzed. An analytical method is proposed for evaluating the reliability improvement by using SIFT for any size of distributed systems . The goal of this research is to identify and develop new key building blocks for reliable distributed systems built with inexpensive off-the shelf components. The platform consists of computing nodes connected via multiple interfaces to networks configured in fault-tolerant topologies. Thus, the research community and a large number of users will have the opportunity to build reliable distributed systems based on COTS in a user transparent way.The goal will be to achieve a high reconfiguration rate and a small recomputation overhead in the presence of faults in the distributed systems and high-speed computer networks in a way transparent to the user, which will provide a high dependability. This research project will have a positive impact on a large community implementing applications in distributed and network environments such as : e-commerce, telepresence, telemedicine, Internet data centers, distributed applications running concurrently across
