| dc.description.abstract | Secure multiparty computation (SMC) focuses on solving the problem of how two or more parties can jointly compute the result of a function without disclosing their private input. One approach to achieving SMC is a garbled circuit (GC), which transforms an arbitrary function into a secure version by representing it as an obfuscated logic circuit. GCs have been utilized to solve a wide range of practical applications, such as secure statistical tests and auctions. A number of secure computation frameworks (SCFs) have been developed to enable programmers with the ability to design and translate GCs into working implementations. Though the procedure of modeling the function as a circuit is independent, the implementation of the code is specialized to each SCF. This is problematic because SCFs evolve over time to incorporate novel strategies, which necessitates reprogramming of the GC. In this thesis, we introduce a strategy (VisualGC) to facilitate the transformation of GCs into their corresponding code of an SCF. Specifically, we design a generic approach to represent GCs through a domain specific modeling language (DSML), which is invoked to design meta-models in the Generic Modeling Environment (GME) platform. Built on meta-models, we provide users with a library that covers GCs with a wide range of computational functionality to help build their own custom-defined circuit, To demonstrate the potential for model-based GC design, we integrate our approach with the model interpreter of FastGC, a recently proposed SCF. During the model interpretation, GC models are automatically transformed into working codes. We validate our approach by demonstrating that the automatically generated code can achieve the expected computation functionality that their corresponding GCs represent. | |
