An underwater acoustic data compression method based on compressed sensing

The use of underwater acoustic data has rapidly expanded with the application of multichannel, large-aperture underwater detection arrays. This study presents an underwater acoustic data compression method that is based on compressed sensing. Underwater acoustic signals are transformed into the sparse domain for data storage at a receiving terminal, and the improved orthogonal matching pursuit(IOMP) algorithm is used to reconstruct the original underwater acoustic signals at a data processing terminal. When an increase in sidelobe level occasionally causes a direction of arrival estimation error, the proposed compression method can achieve a 10 times stronger compression for narrowband signals and a 5 times stronger compression for wideband signals than the orthogonal matching pursuit(OMP) algorithm. The IOMP algorithm also reduces the computing time by about 20% more than the original OMP algorithm. The simulation and experimental results are discussed.

基于改进正交匹配追踪算法的振动信号修复方法

当振动信号数据成段连续丢失,实现压缩感知理论解决丢失信号数据修复时,在信号重构阶段,正交匹配追踪算法(OMP)的迭代误差会在下一次迭代循环中继续传导,导致该算法的稳定性变得更差。由此提出了一种改进的正交匹配追踪算法(IOMP),利用离散余弦变换得到完备字典矩阵,设计了一种适应于缺失振动信号的观测矩阵,最后分别利用OMP算法与IOMP算法完成缺失振动信号的重构。通过数据仿真和实测数据的实验对比,结果表明IOMP算法在振动信号的损失为10%和20%时,均能够减小重构信号与原始信号之间的绝对误差。对信号损失20%时进行定量数据分析,重构信号的统计特性更接近原始信号的统计特性。

DMS/EOAS数据中心设备监测方案的设计和应用

为实施对DMS/EOAS数据中心设备的监测,设计了一款基于CTC3.0的综合运维平台,并具备实时监测DMS/EOAS数据中心设备运行状态的能力。通过SNMP方式,获取DMS/EOAS数据中心系统设备及网络设备监测数据;系统间通过网闸隔离、开放IMM端口方式,获取服务器运行数据,实现DMS/EOAS数据中心与综合运维平台间的数据协同、互联互通。维护人员借助该综合运维平台,可以全面掌握DMS/EOAS数据中心设备、网络运行情况,及时发现问题,辅助分析排查故障。