Packing-based branch-and-bound for discrete malleable task scheduling

Abstract This paper addresses the problem of scheduling chain-like structures tasks on a single multiprocessor resource. In fact, sub-tasks unit-time length and predefined size are aggregated to composite that have be scheduled without preemption, but subject flexibility concerning resource allocation. setting most closely resembles malleable task scheduling, with being smallest atomic unit The specific type malleability is realized using precedence constraints minimum maximum time lags. A bin packing model established for this corresponding, dedicated branch-and-bound algorithm devised, alongside problem-specific bound tightening, symmetry breaking dominance concepts. efficacy solution approach demonstrated based extensive computational experiments, including randomized instances, adapted benchmark instances from literature, small real-world data set. comparison mixed-integer constraint programming formulations, new method able achieve considerably higher percentage optimal solutions at computation times up orders magnitude smaller.