Author Description

Parallel processing has proven its significant role in various large scale scientific and engineering areas including weather forecasting, air traffic control, petroleum exploration, bio-medical, nuclear energy and defence. Very fast computations of a massive volume of data are an urgent need to support all such real time and real life applications. The computations involved in such applications usually follow some numeric and non-numeric algorithms which are inherently slow to run in a sequential machine. Designing efficient parallel algorithms for such computations have thus been an important area of research in which an interconnection network plays a central role. In this book, we study various interconnection networks and develop efficient parallel algorithms for several problems that include balanced ring formation for the fault tolerance, parallel prefix computation, shortest path routing and matrix multiplication. We present all these algorithms on popular interconnection networks namely, 2D-mesh, OTIS-mesh, OTIS-torus and multi-mesh of trees network. The parallel algorithms are analyzed by their time complexity and cost.