接觸點電阻式記憶元件應用於新型隨機亂數產生器之研究

隨著網際網路越來越發達，可攜式的資料傳輸裝置也越來越普及，因此加密系統變得更為重要。為了建立一個可靠的保密系統，產生穩定且非週期性的隨機亂數便是其中一個很重要的部分。本研究提出透過接觸點電阻式記憶元件所產生的隨機電報雜訊，來產生隨機亂數，藉此基礎來設計能產生真正亂數的新型隨機亂數產生器；跟傳統的隨機亂數產生器相比，新型隨機亂數產生器可以有效的降低電路面積及功率損耗。本論文使用的雜訊源為隨機電報雜訊，取代傳統在設計隨機亂數產生器上常使用的熱雜訊。因為隨機電報雜訊擁有較大的振幅(大於通道電流的百分之一)，可節省後端電路設計上的複雜度，但相對於熱雜訊，較難以在金屬氧化物半導體場效應電晶體上量測到，因此使用接觸點電阻式記憶元件來做為研究的雜訊源裝置。相較於金屬氧化物半導體場效應電晶體，接觸點電阻式記憶元件有較小的面積，且因其傳輸方式是屬於閘極電流的形式，因此隨機電報雜訊的影響會更明顯且更易被後序放大處理。更重要一點，接觸點電阻式記憶元件與CMOS邏輯製程完全相容，不需額外的光罩或製程步驟。在電路最佳化設計上，因隨機電報雜訊之高位準及低位準比值，不一定為一比一，因此需要設計機率修正電路。並使用NIST提供的測試套件，可更進一步驗證隨機電報雜訊的亂數性，來證實新型隨機亂數產生器的可靠性，進而確定應用在數位傳輸加密上的可能性。

關鍵字

隨機電報雜訊；電阻式記憶體；隨機亂數產生器

並列摘要

With the popularization of the Internet, data transmission has become more convenient and highly mobile. In order to prevent data theft, data encryption system is essential to modern communication chips. A stable and True Random Number Generator (TRNG) is a key element to ensure the security of transmitted data. TRNG based on the trapping and de-trapping process in the oxide defects has been proposed in recent years. Generating random numbers using electron trapping has several advantages. First, as the source of this random process occurs naturally, hence the resulting random number is aperiodic. In addition, the output is generally insensitive to temperature. Random number generator circuits depend on random latching of inverter pair can subject to mismatch effects. In a MOSFET, random telegraph noise (RTN) refers to the drain current fluctuation (ID-RTN) as a result of the change in the amount of carriers flowing from source to drain. These RTN signals with discrete drain current levels is generally very small and can lead to misleading results after amplifications. In this paper, we proposed a TRNG based on the RTN signals found in the contact resistive random access memory (CRRAM) device. As reported in previous studies, the resistance levels of both the high and low resistance states are subject to high RTN, leading to large read current noises as a results of electron trapping in the conductive pathway. These RTN signals from the embedded CRRAM cells is much larger than ID-RTN, which allows CRRAM to be more appropriate source for generating true random number for encryption circuitry.

並列關鍵字

無資料

參考文獻

[1] Z.Gutterman et al, “Analysis of the Linux RNG”, IEEE Symposium on Security and Privacy, pp. 371 - 385, May 2006

[2] Yuan Heng Tseng, Chia-En Huang, C. -H. Kuo, Y. -D. Chih, Chrong Jung Lin, “High Density and Ultra Small Cell Size of Contact ReRAM (CR-RAM) in 90nm CMOS Logic Technology and Circuits”, IEDM, 2009

[3] Yuan Heng Tseng, Wen Chao Shen, Chia-En Huang, Chrong Jung Lin, Ya-Chin King, “Electron Trapping Effect on the Switching Behavior of Contact RRAM Devices through Random Telegraph Noise Analysis”, IEDM, 2010

[5] Christian Monzio et al, "Statistical Model for Random Telegraph Noise in Flash Memories", IEEE Transactions on Electron Devices, Vol. 55, No. 1, January 2008

[6] Andrea Chimenton, Cristian Zambelli, Piero Olivo, "A New Methodology for Two-Level Random-Telegraph-Noise Identification and Statistical Analysis", IEEE Electron Device Letters, Vol. 31, No. 6, June 2010

國際替代計量

接觸點電阻式記憶元件應用於新型隨機亂數產生器之研究

全文下載

主題瀏覽