Electrical Engineering and Computer Science

CSE Technical Reports Sorted by Technical Report Number

TR Number Title Authors Date Pages

CSE-TR-600-16 InvisiMem: Smart Memory for Trusted Computing Shaizeen Aga and Satish Narayanasamy November, 2016 13
A practically feasible low-overhead secure hardware designthat provides strong defenses against memory bus side chan-nels remains elusive. This paper observes that smart memory,memory with compute capability and a packetized interface,can dramatically simplify this problem. InvisiMem expandsthe trust base to include the logic layer in the smart memory toimplement cryptographic primitives, which allows the securehost processor to send encrypted addresses over the untrustedmemory bus. This eliminates the need for expensive addressobfuscation techniques based on Oblivious RAM (ORAM).In addition, smart memory enables simple and efficient solu-tions for ensuring freshness using authenticated encryption,and for mitigating memory bus timing channel using constantheart-beat packets. We demonstrate that InvisiMem designsare one to two orders of magnitude lower in performance,space, energy, and memory bandwidth overhead, compared toORAM-based solutions.

CSE-TR-601-16 Spreadsheet Property Detection With Rule-assisted Active Learning Zhe Chen, Xin Rong, Sasha Dadiomov, Richard Wesley, Gang Xiao, Daniel Cory, Michael Cafarella, Jock Mackinlay 11, 2016 9
Spreadsheets are a critical and widely-used data management tool. Converting spreadsheet data into relational tables would bring benefits to a number of fields, including public policy, public health, and economics. Research to date has focused on designing domain-specific languages to describe transformation processes or automatically converting a specific type of spreadsheets. To handle a larger variety of spreadsheets, we have to identify various spreadsheet properties, which correspond to a series of transformation programs that contribute towards a general framework that converts spreadsheets to relational tables. In this paper, we focus on the problem of spreadsheet property detection, identifying when a corresponding transformation program should be applied. We propose a hybrid approach of building a variety of spreadsheet property detectors to reduce the amount of required human labeling effort. Our approach integrates an active learning framework with crude, easy-to-write, user-provided rules. Our experiments show that when compared to a standard active learning approach, we reduced the training data needed to reach the performance plateau by 34--44% when a human provides relatively high-quality rules, and by a comparable amount with low-quality rules. A study on a large-scale web-crawled spreadsheet dataset demonstrates that it is crucial to detect a variety of spreadsheet properties in order to transform a large portion of the spreadsheets into a relational form.

Technical Reports Page