Online scheduling algorithm for parallel batch processing machines

This paper studies two online scheduling problems on parallel machines that process computing jobs arriving stochastically in a batch pattern. With the objective function as minimizing total completion time of weighted jobs or minimizing total latency of job completion, we propose two solutions for above scheduling problems, respectively. With the assumption that the processing time of a batch is a constant, in the first solution the jobs with higher weights get processed first. In the second problem, the job with a greater weight is assigned higher priority and we assume that the latency of any job doesn't exceed a given range. Considering such two impact factors as delay penalty parameter and job processing time, we adjust the job's waiting time to eventually minimize the value of objective function.