Title :
An application-specific low power speech processor for cochlear implants
Author :
Mai, Songping ; Zhang, Chun ; Wang, ZhiHua
Author_Institution :
Inf. Sci. & Technol. Div., Tsinghua Univ., Shenzhen, China
Abstract :
Fully implantable cochlear prostheses are considered as next generation products, in which low power speech processors are very crucial. This paper reports a digital speech processor dedicated to cochlear prostheses. It is a programmable platform and able to carry out all kinds of speech processing algorithms. Several application-specific low power techniques including algorithm optimization, wait mode, operand isolation, clock gating, loop cache and memory partitioning are introduced in the processor design. Implemented using 0.18 ¿m CMOS technology, the processor prototype consumes about 2.3 mW when executing the continuous-interleaved-sampling (CIS) algorithm at a clock frequency of 10 MHz.
Keywords :
CMOS integrated circuits; application specific integrated circuits; biomedical electronics; cochlear implants; digital signal processing chips; low-power electronics; speech processing; CMOS technology; algorithm optimisation; application specific low power speech processor; application specific low power techniques; clock gating; cochlear implants; continuous interleaved sampling algorithm; digital speech processor; implantable cochlear prostheses; loop cache partitioning; memory partitioning; next generation products; operand isolation; speech processing algorithms; wait mode; Algorithm design and analysis; CMOS process; CMOS technology; Clocks; Cochlear implants; Design optimization; Partitioning algorithms; Process design; Prosthetics; Speech processing;
Conference_Titel :
Biomedical Circuits and Systems Conference, 2009. BioCAS 2009. IEEE
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-4917-0
Electronic_ISBN :
978-1-4244-4918-7
DOI :
10.1109/BIOCAS.2009.5372054