Fault Tolerant Hierarchical Con guration of Mesh Networks

In an e ort to improve the overall performance of distributed mesh systems a hierar chical approach of dividing the processing units into logical levels has been proposed The scheme has been tested and shown to achieve substantial performance improvement in many situations This hierarchical con guration divides the mesh into progressively smaller clusters thus allowing for computation to occur in small local groups at the lower levels After local operations the results are passed to higher logical levels Un der certain assumptions this method has been shown to produce an overall minimal total communication cost for the entire system As the distributed mesh systems attempt to achieve optimal performance with this hierarchical con guration questions arise concerning the system s ability to handle node failure When using a hierarchical con guration certain elements in the mesh become more important to the overall system than others It is important that the hierarchical system have a re organizing mechanism in the case of node s failing to function in such a way that the performance gain from hierarchical con guration is salvaged as much as possible The fault tolerance problems considered in this project are concerned with minimiz ing the loss of performance in the system hierarchy due to the presence of failing nodes We will propose schemes algorithms for that purpose The performance results will be compared to that of an ideal fault free system The rst phase of the project will be to design strategies to reconstruct the hierarchy accommodating to the situation that some nodes in the original hierarchy are not functioning anymore To that end new local heads may be selected and local nodes regrouped In the second phase of the project simulation experiments will be developed to examine the e ectiveness of the proposed schemes Examples of both faulty and fault free hierarchical mesh systems will be tested to quantify how good the proposed schemes are Index Terms Distributed Processing Fault tolerance Hierarchical control Inter connection networks Mesh