Mutual relationship between the neural network model and linear complexity for pseudorandom binary number sequence

Yuki Taketa, Yuta Kodera, Shogo Tanida, Takuya Kusaka, Yasuyuki Nogami, Norikazu Takahashi, Satoshi Uehara

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Machine learning (ML) technology has been getting popular in many applications where ML purposes to analyze or classify data, or predicting the phenomenon follows from the previous conditions, for example. However, the spread of ML technologies allows an attacker to use them into the attack for the sake of sniffing secret information. Since the randomness has been used for an inseparable part of the cryptographic applications to ensure the security, the resistance of a random sequence against analysis based on the ML technologies have to be required. The authors anticipate having the mutual relationship between the classical properties of the randomness, linear complexity (LC) in particular, and the structure of a neural network (NN), which is a class of ML. In this research, the authors find that the strength of each connection between nodes in the NN is relevant to the linear recurrence relation of the target sequence by observing parameters after complete learning. In other words, the difficulty of predicting the next bits from a given sequence would be discussed based on the LC of a sequence in most cases. The experimental results are introduced to clarify the black box in this research.

Original languageEnglish
Title of host publicationProceedings - 2019 7th International Symposium on Computing and Networking Workshops, CANDARW 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages394-400
Number of pages7
ISBN (Electronic)9781728152684
DOIs
Publication statusPublished - Nov 2019
Event7th International Symposium on Computing and Networking Workshops, CANDARW 2019 - Nagasaki, Japan
Duration: Nov 26 2019Nov 29 2019

Publication series

NameProceedings - 2019 7th International Symposium on Computing and Networking Workshops, CANDARW 2019

Conference

Conference7th International Symposium on Computing and Networking Workshops, CANDARW 2019
CountryJapan
CityNagasaki
Period11/26/1911/29/19

Keywords

  • Linear complexity
  • M-sequence
  • Machine learning
  • Prediction attack
  • Pseudorandom number sequence

ASJC Scopus subject areas

  • Hardware and Architecture
  • Information Systems
  • Artificial Intelligence
  • Computer Networks and Communications

Fingerprint Dive into the research topics of 'Mutual relationship between the neural network model and linear complexity for pseudorandom binary number sequence'. Together they form a unique fingerprint.

  • Cite this

    Taketa, Y., Kodera, Y., Tanida, S., Kusaka, T., Nogami, Y., Takahashi, N., & Uehara, S. (2019). Mutual relationship between the neural network model and linear complexity for pseudorandom binary number sequence. In Proceedings - 2019 7th International Symposium on Computing and Networking Workshops, CANDARW 2019 (pp. 394-400). [8951746] (Proceedings - 2019 7th International Symposium on Computing and Networking Workshops, CANDARW 2019). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CANDARW.2019.00074