On polynomial solvability of two multiprocessor scheduling problems

Abstract.

We discuss a new approach for solving multiprocessor scheduling problems by using and improving results for guillotine pallet loading problem. We introduce a new class of schedules by analogy with the guillotine restriction for cutting stock problem and show that many well-known algorithms from classical scheduling theory construct schedules only from this class. We also consider two multiprocessor scheduling problems and prove that they can be solved in polynomial time.