Multiprocessor system models at present are very important and widely used in modelling transaction processing systems, communication networks, mobile networks, and flexible machine shops with groups of machines. Heterogeneous processors system with one faster main server and several identical servers are studied. In this paper reconfiguration and rebooting delays are considered to study the performance measures for both bounded and unbounded system. Numerical results are presented for various performability parameters.

Abstract—Many problems in computer vision and image processing present potential for parallel implementations through one of the three major paradigms of geometric parallelism, algorithmic parallelism and processor farming. Static process scheduling techniques are used successfully to exploit geometric and algorithmic parallelism, while dynamic process scheduling is better suited to dealing with the independent processes inherent in the process farming paradigm. This paper considers the application of parallel or multi-computers to a class of problems exhibiting spatial data characteristic of the geometric paradigm. However, by using processor farming paradigm, a dynamic scheduling technique is developed to suit the MIMD structure of the multi-computers. A hybrid scheme of scheduling is also developed and compared with the other schemes. The specific problem chosen for the investigation is the Hough transform for line detection. Keywords—Hough transforms, parallel computer, parallel paradigms, scheduling.