https://hgpu.org/?p=9891
Scalable Techniques for Scheduling and Mapping DSP Applications onto Embedded Multiprocessor Platforms