SCIENTIA SINICA Informationis, Volume 48 , Issue 3 : 315-328(2018) https://doi.org/10.1360/N112017-00204

New generation software-defined architecture

More info
  • ReceivedOct 24, 2017
  • AcceptedFeb 8, 2018
  • PublishedMar 16, 2018


Architecture plays an important role in information systems. It not only determines the function and performance of a system, but the system efficiency and security as well. Therefore, the architecture directly determines the advanced nature of an information system. On the basis of introducing the concept of architecture, this paper summarizes the development of architecture, and points out that the rigidity of architecture is the essential reason behind why current information systems cannot tradeoff between flexibility and high efficiency. A new generation of software-defined architecture, which is characterized by software-defined interconnections and software-defined nodes, is proposed. Three typical systems, Web service, password recovery and image recognition, are realized based on a software-defined architecture. Compared with the traditional general systems, the comparison tests show that the performance of the software-defined architecture system increased by 29.4 to 344.5 times,łinebreak and the efficiency increased by 13.7 to 315.4 times, which demonstrates the high flexibility and high efficiency of software-defined architecture.

Funded by





[1] Amdahl G M. The structure of SYSTEM/360, Part III: Processing unit design considerations. IBM Syst J, 1964, 3: 144-164 CrossRef Google Scholar

[2] Dong Z. Discussion on the structure of science system. Sci Sci Manag ST, 1992, 11: 5--8. Google Scholar

[3] Luo A M, Huang L, Luo X S. Study on architechtural description and design of C4ISR. Fire Control Command Control, 2005, 30: 25--28. Google Scholar

[4] Xu J Y. Analysis of major developments in microelectronics. J Chinese Acad Electron Sci, 2006, 1: 215--218. Google Scholar

[5] Liu B W, Chen S M, Wang D. Survey on advance microprocessor architecture and its development trends. Appl Res Comput, 2007, 24: 16--20. Google Scholar

[6] Wang J. Architecture analysis of computer cache memory. Aeronaut Comput Technique, 2006, 36: 29--33. Google Scholar

[7] Zhao Z S. The bus-plugin styled architecture method research. Dissertation for Master's Degree. Chongqing: Chongqing University, 2001. Google Scholar

[8] Azodolmolky S. Software Defined Networking. Beijing: China Machine Press, 2014. Google Scholar

[9] Ye Y R. Software Defined Storage. Beijing: China Machine Press, 2016. Google Scholar

[10] Zhang X L, Zhang D, Cao L L, et al. Cloud computing virtualization platform performance research. Softw Guide, 2013, 12: 1--3. Google Scholar

[11] Chen X, Ricky S. Software Defined Data Center. Beijing: China Machine Press, 2015. Google Scholar

[12] Wang X, Chen M, Chao H U, et al. SDICN: a software defined deployable framework of information centric networking. China Commun, 2016, 13: 53--65. Google Scholar

[13] Huber N, Quast M V, Hauck M, et al. Evaluating and modeling virtualization performance overhead for cloud environments. In: Proceedings of the International Conference on Cloud Computing and Services Science, Noordwijkerhout, 2011. 563--573. Google Scholar

[14] Wu J X, Luo X G, Cao W, et al. A reconfigurable computing array and construction method. CN 102799563 B. 2015. Google Scholar

[15] Wu J X, Luo X G, Si X M. The “stochastic strain" mimic computing village. Sci Technol Overview, 2014, 5: 54--58. Google Scholar

Copyright 2020 Science China Press Co., Ltd. 《中国科学》杂志社有限责任公司 版权所有

京ICP备17057255号       京公网安备11010102003388号