作者: Paolo Meloni , Giuseppe Tuveri , Danilo Pani , Luigi Raffo , Francesca Palumbo
DOI: 10.1109/DASIP.2015.7367243
关键词:
摘要: The design of embedded systems for neuroprosthetic applications represents an important challenge to be faced in electronic bioengineering. One the key research problems is decoding information encoded neural signals extract patient's motion intention. How implement a highly-portable and reliable integrated solution still open issue. In this paper, we investigate possibility adopting MPSoC paradigm application domain, presenting space exploration that evaluates different custom architectures, implementing on-line signal algorithm. evaluated points feature mappings parallel software tasks onto customized ASIP processing cores. Experimental results, obtained by FPGA-based prototyping, assess performance hardware-related costs considered configurations. clock frequency needed respect real-time constraints was reduced 22 MHz, making step further towards exploitation heterogeneous MPSoCs ultra-low power biomedical processing.