On the re-allocation of processing resources for digital audio signal processing

When implemented digitally, large-scale audio processing systems require the realisation of highly parallel architectures. Task scheduling strategies are necessary to distribute the processing algorithms required-termed the taskforce-across such architectures. The results obtained show that good-quality solutions can be expected from simple heuristic methods with minimum scheduling overhead. More advanced heuristic policies demonstrate particular promise for fast near-optimal re-allocation of processing resources in real-time audio processing applications