论文部分内容阅读
为了提高通信系统的保密性,降低制造成本,需要进行专用处理器的设计。基于正弦激励线性预测(SELP)算法模型,设计了一款多速率语音专用处理器。芯片使用可重构体系结构和超长指令字(VLIW),优化了高复杂度函数。仿真结果表明:该处理器对0.6kb/s速率SELP算法的执行效率明显优于通用数字信号处理器(DSP)。处理器内部程序数据外部不可见,指令并行度显著提高,常用函数可被修改,从而达到高保密性、低复杂度、易开发性。
In order to improve the confidentiality of the communication system and reduce the manufacturing cost, a dedicated processor design is required. Based on the sinusoidal excitation linear prediction (SELP) algorithm model, a multi-rate speech dedicated processor is designed. The chip uses a reconfigurable architecture and Very Long Instruction Word (VLIW) to optimize high-complexity functions. The simulation results show that the processor performs significantly better than the general digital signal processor (DSP) for the 0.6kb / s SELP algorithm. Processor internal program data is not visible outside the external command line significantly increased the degree of parallelism, commonly used functions can be modified to achieve high confidentiality, low complexity, easy to develop.