Method for implementing sonar real-time signal processing based on Linux group

Abstract

The invention relates to a method for implementing sonar real-time signal processing based on a Linux group. In the method, two parts, namely a sonar base array signal real-time simulator and sonar real-time signal processing, are involved. By the method, a parallel processing design idea is introduced into the sonar signal processing, a signal processing task is divided into various independent sub tasks, allocation of hardware resources and data links are automatically realized by the Linux group, the various sub tasks are automatically allocated and run on various calculation cores, a datastream and a control stream are automatically realized by a network communication method, and human intervention is not required. The method has the advantages that: the Linux group is adopted in a signal processing hardware platform; as the Linux group is low in cost and easy in construction, the expandibility is higher and the calculation capability is strong. By adoption of a Linux operating system, a hardware platform and a software platform are separated and the signal processing programming is executed in a universal exploitation environment, so the difficulty of signal processing software is simplified. Due to adoption of a message passing interface (MPI) base for parallel program exploitation, the parallel calculation capability of the group is developed fully.

Description

technical field [0001] The invention belongs to the field of cluster parallel computers and the field of sonar signal processing, and mainly relates to a Linux cluster-based sonar real-time signal processing realization method. Background technique [0002] Sonar digital signal processing mainly completes digital signal processing functions such as beamforming, digital filtering, and spectrum analysis, and is an important part of modern digital sonar. At present, the sonar signal processing system is mainly implemented on a dedicated signal processor, which requires dedicated signal processing chips and data transmission hardware support. The topology of communication between boards and chips is fixed. As sonar signal processing becomes more and more Complicated, signal processors are often limited by bus and inter-board data transmission bottlenecks to improve computing power in the case of surges in computing and data scale. [0003] With the continuous development of son...

AI Technical Summary

Problems solved by technology

At present, the sonar signal processing system is mainly implemented on a dedicated signal processor, which requires dedicated signal processing chips and data transmission hardware support. The topology of communication between boards and chips is fixed. As sonar signal processing becomes more and more Complicated, signal processors are often restricted by bus and inter-board data transmission bottlenecks in the case of surges in computing and data scale.

Images