ARCHIVES

In view of the fast evolution of the field of designing intelligent embedded systems, there is a necessity for having a technique to handle data effectively and schedule tasks in a heterogeneous multiprocessor system. The traditional techniques of static handling of data and task scheduling have failed to provide sufficient flexibility resulting in several problems like heavy communication overhead, poor resource utilization and inefficiency of overall system performance. In this paper we present a new approach to dynamic handling of data and intelligent task allocation using adaptive artificial intelligence algorithms in order to improve the performance of the high-performance embedded multiprocessor systems. The proposed approach consists of using machine learning algorithms in conjunction with dynamic data allocation and intelligent task allocation to optimize processor utilization and reduce communication and memory access delays. The proposed approach also involves using a sequential parallel processing approach to ensure dependencies among data and maximize concurrency and even load balancing among processors.