雷达数据实时处理软件多线程技术的改进
|
摘要
针对雷达数据实时处理软件按任务静态配置线程的传统方式带来的功能模块粒度较大、线程利用率低和并行性能不佳等问题,提出了基于线程池技术的软件改进模型,使用标准C++线程库设计并实现了跨平台通用线程池,能够通过灵活的数据分解合理划分和均衡分配计算任务。测试结果表明使用线程池的程序充分发挥了多核平台本身的并行计算能力,其并行性能优于采用传统技术的程序,在处理短时、密集型任务时优势更加明显。线程池为各种实时处理软件提供统一、可重用的并行处理功能,可以提高软件的开发质量和效率。
Since the traditional way that the real-time processing software for radar data configures threads fixedly by tasks has the problems of large function module granularity, low thread utilization and poor parallel performance, a solution based on thread pool scheme was put forward. This solution used standard thread library of C + + to design and implement a crossplatform, general-purpose thread pool, by which the computing tasks could be rationally divided and evenly distributed through flexible data decomposition. The test results show that the program with thread pool scheme can give full play to parallel calculating performance of multi-core platform, which is superior to that of traditional technology, especially when dealing with short-time and intensive tasks. The thread pool can provide a unified and reusable parallel processing capability for all kinds of real-time processing softwares, improving the quality and efficiency of software development.
Since the traditional way that the real-time processing software for radar data configures threads fixedly by tasks has the problems of large function module granularity, low thread utilization and poor parallel performance, a solution based on thread pool scheme was put forward. This solution used standard thread library of C + + to design and implement a crossplatform, general-purpose thread pool, by which the computing tasks could be rationally divided and evenly distributed through flexible data decomposition. The test results show that the program with thread pool scheme can give full play to parallel calculating performance of multi-core platform, which is superior to that of traditional technology, especially when dealing with short-time and intensive tasks. The thread pool can provide a unified and reusable parallel processing capability for all kinds of real-time processing softwares, improving the quality and efficiency of software development.
引文
[1]成求青,李波,余浩章,等.导弹测控系统总体设计原理与方法[M].北京:清华大学出版社,2014:2-31.
[2]李润平,许聚常.测控中心实时应用软件系统采用线程机制的探讨[J].飞行器测控学报,2004,23(3):81-84.
[3]赵源,姜小峰.基于多线程技术的自动测试系统优化设计[J].计算机应用,2014,34(7):2124-2128.
[4]眭俊华,刘慧娜,王建鑫,等.多核多线程技术综述[J].计算机应用,2013,33(S1):239-242,261.
[5] GOVE D.多核应用编程实战[M].郭晴霞,译.北京:人民邮电出版社,2013:24-112.
[6]朱海涛,倪世道.并行处理技术在雷达数据处理中的应用研究[J].计算技术与自动化,2013,32(2):109-114.
[7]王剑,谭守林,雷刚,等.基于多线程的导弹与空间碎片碰撞预警方法[J].弹箭与制导学报,2012,32(5):107-110.
[8]张培辉,黄一敏.多线程和UDP在等效飞行控制软件中的应用[J].兵工自动化,2012,31(2):85-92.
[9]王科特,王力生.信号实时采集系统的最佳并行线程数的研究[J].计算机应用,2011,31(10):2593-2596.
[10] ALIAGA J I, BOLLHOFER M, MARTIN A F, et al. Exploiting thread-level parallelism in the iterative solution of sparse linear systems[J]. Parallel Computing, 2011, 37(3):183-202.
[11]邱长泉,路辉.飞行数据管理仿真系统软件设计及实现[J].测试技术学报,2011,25(1):40-46.
[12]高德宏,张新家,陈春雷,等.发布订阅模式数据交换中间件设计与实现[J].计算机应用,2010,30(4):1110-1113.
[13] KANG D H, HAN S, YOO S H, et al. Prediction-based dynamic thread pool scheme for efficient resource usage[C]//Proceedings of the 2008 IEEE 8th International Conference on Computer and Information Technology Workshops. Washington, DC:IEEE Computer Society, 2008:159-164.
[14] LING Y, MULLEN T, LIN X. Analysis of optimal thread pool size[J]. ACM SIGOPS Operating Systems Review, 2000, 34(2):42-55.
[15] GREGOIRE M. C++高级编程[M].张永强,译. 3版.北京:清华大学出版社,2015:267-675.
[2]李润平,许聚常.测控中心实时应用软件系统采用线程机制的探讨[J].飞行器测控学报,2004,23(3):81-84.
[3]赵源,姜小峰.基于多线程技术的自动测试系统优化设计[J].计算机应用,2014,34(7):2124-2128.
[4]眭俊华,刘慧娜,王建鑫,等.多核多线程技术综述[J].计算机应用,2013,33(S1):239-242,261.
[5] GOVE D.多核应用编程实战[M].郭晴霞,译.北京:人民邮电出版社,2013:24-112.
[6]朱海涛,倪世道.并行处理技术在雷达数据处理中的应用研究[J].计算技术与自动化,2013,32(2):109-114.
[7]王剑,谭守林,雷刚,等.基于多线程的导弹与空间碎片碰撞预警方法[J].弹箭与制导学报,2012,32(5):107-110.
[8]张培辉,黄一敏.多线程和UDP在等效飞行控制软件中的应用[J].兵工自动化,2012,31(2):85-92.
[9]王科特,王力生.信号实时采集系统的最佳并行线程数的研究[J].计算机应用,2011,31(10):2593-2596.
[10] ALIAGA J I, BOLLHOFER M, MARTIN A F, et al. Exploiting thread-level parallelism in the iterative solution of sparse linear systems[J]. Parallel Computing, 2011, 37(3):183-202.
[11]邱长泉,路辉.飞行数据管理仿真系统软件设计及实现[J].测试技术学报,2011,25(1):40-46.
[12]高德宏,张新家,陈春雷,等.发布订阅模式数据交换中间件设计与实现[J].计算机应用,2010,30(4):1110-1113.
[13] KANG D H, HAN S, YOO S H, et al. Prediction-based dynamic thread pool scheme for efficient resource usage[C]//Proceedings of the 2008 IEEE 8th International Conference on Computer and Information Technology Workshops. Washington, DC:IEEE Computer Society, 2008:159-164.
[14] LING Y, MULLEN T, LIN X. Analysis of optimal thread pool size[J]. ACM SIGOPS Operating Systems Review, 2000, 34(2):42-55.
[15] GREGOIRE M. C++高级编程[M].张永强,译. 3版.北京:清华大学出版社,2015:267-675.