An Optimal Scheme for Multiprocessor Task Scheduling: a Machine Learning Approach

We consider the problem of scheduling periodic task sets on identical multiprocessors. In this case deadlines are equal periods, although results described here can be applied to sporadic task sets with periods not equal to deadlines. We present an online scheduling policy that can be used to design scheduling algorithms. We came up with scheduling rules to minimize the number of preemptions and reduce the number of overheads in an online multiprocessor scheduling algorithm and will imply these rules by incorporating machine learning techniques.