Research on Adaptive Evolution of Existing Software

现有软件的自适应进化研究

• 批准号: 09245106
• 负责人: USHIJIMA Kazuo
• 金额: $ 26.24万
• 依托单位: Kyushu University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research on Priority Areas
• 财政年份: 1997
• 资助国家: 日本
• 起止时间: 1997 至 1999
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-09245106/
• 关键词: Understanding of software; Software reconstruction; Evaluation of adaptability; Object orientation; Adaptive evolution of operating system; Language processor for parallel computer system; Change of user environment; Change of requirements; 既存ソフトウ

Program slicing is one of the important means to understand existing software. Inoue et al. developed an efficient method to get program slice, and proposed accumulative versioning file. Based on his proposal of a systematic methodology to develop large-scale concurrent software, Cheng et al. designed and partially implemented a support environment for developing such software.In order to support reconstruction of existing software, Ushijima et al. proposed a method to make effective use of design patterns related to hot spots of the target software, and applied the method to redesign an application software and evaluated the effect of the method.Torii et al. constructed CAESE (Computer Aided Empirical Software Engineering) environment to support collection and analysis of data needed to quantitatively and qualitatively evaluate adaptability of reconstructed software.Tokuda et al. made a research and development of software architecture of next generation OS to be adaptable to parallel and distributed systems, multimedia systems, mobile and ubiquitous systems, etc. Fukuda et al. designed an automatic generation method of device drivers as an evolutionary mechanism for OS, and implemented them to experimentally evaluate the method using Free BSD and Linux.In order to provide a portable compiler which reflects individual architecture of every parallel computer, Yuasa et al. proposed an optimization mechanism based on special characteristics of individual architecture, and investigated how to implement the mechanism.In order to guarantee service execution time independent of hardware performance, Taniguchi devised a process scheduling method to regulate allocation of processors to the program, implemented and evaluated the method.

程序切片是理解现有软件的重要手段之一。井上等人。开发了一种有效的程序切片获取方法，并提出了累积版本控制文件。 Cheng 等人基于他提出的开发大规模并发软件的系统方法论。设计并部分实现了开发此类软件的支持环境。为了支持现有软件的重建，Ushijima 等人。 Torii等人提出了一种有效利用与目标软件热点相关的设计模式的方法，并应用该方法重新设计了一个应用软件并评估了该方法的效果。构建了CAESE（计算机辅助经验软件工程）环境，以支持定量和定性评估重构软件适应性所需数据的收集和分析。 Fukuda等人研究开发了适应并行分布式系统、多媒体系统、移动泛在系统等的下一代操作系统软件架构。设计了一种设备驱动程序的自动生成方法作为操作系统的进化机制，并使用 Free BSD 和 Linux 实现了该方法并通过实验评估该方法。为了保证服务执行时间与硬件性能无关，谷口设计了一种进程调度方法来调节处理器对程序的分配，并实现和评估了该方法。

项目成果

鳥居 宏次: "Ginger2:An environment for CAESE(Computer Aided Empirical Software Engineering)"IEEE Trans.on Software Engineering. 25,4. 474-492 (1999)

Koji Torii：“Ginger2：CAESE（计算机辅助经验软件工程）环境”IEEE Trans.on 软件工程 25,4 (1999)。

MASUDA Gou: "Applying Design Patterns to Decision Tree Learning System"Proc. ACM SIGSOFT 6th International Symposium on the Foundation of Software Engineering (FSE-6). 111-120 (1998)

MASUDA Gou：“将设计模式应用于决策树学习系统”Proc。

TORII Koji: "Ginger2 : An environment for CAESE(Computer Aided Empirical Software Engineering)"IEEE Trans. on Software Engineering. 25,4. 474-492 (1999)

TORII Koji：“Ginger2：CAESE（计算机辅助经验软件工程）环境”IEEE Trans。

谷口 秀夫: "プロセススケジュールの制御によるプログラムの実行速度制御法の評価"電子情報通信学会論文誌Dー1. J83D-1,1. 184-193 (2000)

Hideo Taniguchi：“使用进程调度控制的程序执行速度控制方法的评估”IEICE Transactions J83D-1,1 (2000)。

G.Masuda,N.Sakamoto,K.Ushijima: "Applying Design Patterns to Decision Tree Learning System" Pruc.ACMSIGSOFT 6th Imtern.Symp.on the Formdating Software Emgineering. 111-120 (1998)

G.Masuda、N.Sakamoto、K.Ushijima：“将设计模式应用于决策树学习系统” Pruc.ACMSIGSOFT 6th Imtern.Symp.on 形成软件工程。