The existence of error when compressing nonlinear functions into the coefficients of the characteristic model is known to be a key issue in existing characteristic modeling approaches,which is solved in this work by a...The existence of error when compressing nonlinear functions into the coefficients of the characteristic model is known to be a key issue in existing characteristic modeling approaches,which is solved in this work by an error-free compression method.We first define a key concept of relevant states with corresponding compressing methods into their coefficients,where the coefficients are continuous and bounded and the compression is error-free.Then,we give the conditions for decoupling characteristic modeling for MIMO systems,and sequentially,we establish characteristic models for nonlinear systems with minimum phase and relative order two as well as the flexible spacecrafts,realizing the equivalence in the characteristic model theory.Finally,we explicitly explain the reasons for normalization in the characteristic model theory.展开更多
A single channel with a 160-Gb/s optical time-division-multiplexing （OTDM） transmission over 100 km is fabricated. With the help of 500-GHz optical sampling oscilloscopes, the fiber length is adjusted to the order o...A single channel with a 160-Gb/s optical time-division-multiplexing （OTDM） transmission over 100 km is fabricated. With the help of 500-GHz optical sampling oscilloscopes, the fiber length is adjusted to the order of 10 m, which corresponds to the accuracy of 0.4 ps for the dispersion compensation. The dispersion map is optimized for the 100-km transmission link. A completely error-free transmission with the power penalty of 3.6 dB is achieved for 2 h without using forward error correction.展开更多
RISC-V作为一种新兴的开源精简指令集架构,是后摩尔时代处理器技术发展与创新的关键之一.浮点求和与点积运算是数值运算的基础组成部分,在众多领域应用广泛.目前RISC-V架构尚未适配兼具高精度和高效率的求和与点积运算算法,这是因为现...RISC-V作为一种新兴的开源精简指令集架构,是后摩尔时代处理器技术发展与创新的关键之一.浮点求和与点积运算是数值运算的基础组成部分,在众多领域应用广泛.目前RISC-V架构尚未适配兼具高精度和高效率的求和与点积运算算法,这是因为现有优化方案难以良好地平衡运算精度和效率,要么侧重于低精度算法效率,要么通过牺牲效率实现高精度运算.本文利用RVV(RISC-V Vector instruction set extension,RVV)矢量扩展指令,设计并实现了一种基于无误差变换技术的高效、高精度求和与点积算法.首先避免使用规约指令以防止运算精度降低,实现并优化两类运算基于RVV的向量化算法;其次根据算法中的数据依赖关系,对寄存器配置参数进行优化.最后针对算法核心步骤进行汇编优化,增加指令级并行度,提高流水线利用率.实验结果表明,与两类运算操作的原始算法相比,优化后的算法运算效率分别提高了4.4和4.2倍.优化后的算法与多精度库MPFR中的四精度算法有相同精度,但其运算效率明显优于后者,其计算速度与OpenBLAS的双精度计算速度相当.展开更多
基金This work was supported by the National Key R&D Program of China(Grant Nos.2018YFA0703800 and 2018AAA0100800)the Science and Technology on Space Intelligent Control Laboratory Foundation of China(Grant No.ZDSYS-2018-04)the National Natural Science Foundation of China(Grant Nos.U20B2054 and 51805025).
文摘The existence of error when compressing nonlinear functions into the coefficients of the characteristic model is known to be a key issue in existing characteristic modeling approaches,which is solved in this work by an error-free compression method.We first define a key concept of relevant states with corresponding compressing methods into their coefficients,where the coefficients are continuous and bounded and the compression is error-free.Then,we give the conditions for decoupling characteristic modeling for MIMO systems,and sequentially,we establish characteristic models for nonlinear systems with minimum phase and relative order two as well as the flexible spacecrafts,realizing the equivalence in the characteristic model theory.Finally,we explicitly explain the reasons for normalization in the characteristic model theory.
基金supported by the Natural Science Foundation of Beijing(No.4062027)the National"863"Project of China(Nos.2007AA01Z258 and 2008AA01Z15)the National Natural Science Foundation of China(Nos.60877042 and 60837003)
文摘A single channel with a 160-Gb/s optical time-division-multiplexing （OTDM） transmission over 100 km is fabricated. With the help of 500-GHz optical sampling oscilloscopes, the fiber length is adjusted to the order of 10 m, which corresponds to the accuracy of 0.4 ps for the dispersion compensation. The dispersion map is optimized for the 100-km transmission link. A completely error-free transmission with the power penalty of 3.6 dB is achieved for 2 h without using forward error correction.
文摘RISC-V作为一种新兴的开源精简指令集架构,是后摩尔时代处理器技术发展与创新的关键之一.浮点求和与点积运算是数值运算的基础组成部分,在众多领域应用广泛.目前RISC-V架构尚未适配兼具高精度和高效率的求和与点积运算算法,这是因为现有优化方案难以良好地平衡运算精度和效率,要么侧重于低精度算法效率,要么通过牺牲效率实现高精度运算.本文利用RVV(RISC-V Vector instruction set extension,RVV)矢量扩展指令,设计并实现了一种基于无误差变换技术的高效、高精度求和与点积算法.首先避免使用规约指令以防止运算精度降低,实现并优化两类运算基于RVV的向量化算法;其次根据算法中的数据依赖关系,对寄存器配置参数进行优化.最后针对算法核心步骤进行汇编优化,增加指令级并行度,提高流水线利用率.实验结果表明,与两类运算操作的原始算法相比,优化后的算法运算效率分别提高了4.4和4.2倍.优化后的算法与多精度库MPFR中的四精度算法有相同精度,但其运算效率明显优于后者,其计算速度与OpenBLAS的双精度计算速度相当.
