Shamiri ühissalastusest inspireeritud turvalise ühisarvutuse protokollide komplekt
Files
Date
2014
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Tänapäeva maailm on täis saladusi. Mõnikord me sooviks teada midagi nende kohta ilma oma saladusi avalikustamata. Näiteks võib kedagi huvitada, kas ta on rikkam kui mõni tema sõber. Valitsusi võib huvitada, et nende satelliidid ei satuks kokkupõrkesse, kuid samas ei tahaks nad välismaailmale oma strateegiliste satelliitide liikumise kohta liiga palju infot jagada. Turvaline ühisarvutus võimaldab osapooltel ühiselt arvutada mingi funktsiooni väärtusi ilma sisendeid avalikustamata. Sharemind on üks praktiline raamistik turvaliste ühisarvutuste tegemiseks. Selles töös loomegi protokollide komplekti Sharemindi jaoks. Meie protokollid on inspireeritud Shamiri ühissalastusest, mis võimaldab saladusi osadeks jagada. Me anname algoritmid liitmise, korrutamise ja võrdlustehete jaoks koos vastavate turvatõestustega. Lisaks võrdleme realiseeritud protokollide jõudlust juba varasemalt olemasoleva protokollistikuga. Esialgsed keerukus- ja jõudlustulemused on lubavad, kuid on arenguruumi.
The world today is full of secrets. Sometimes, we would like to know something about them without revealing the secrets themselves. For example, whether I have more money than my friend or whether two satellites would collide without publishing their moving trajectories. Secure multi-party computation allows us to jointly compute some functions while keeping the privacy of our inputs. Sharemind is a practical framework for performing secure multi-party computations. In this work, we added a protocol suite to Sharemind. This protocol suite was inspired by Shamir's secret sharing scheme, which describes a way to divide a secret into pieces. We describe algorithms for addition, multiplication, equality-testing and less-than comparison. We also give correctness and security proofs for the protocols. The resulting implementations were compared to an existing protocol suite inspired by additive secret sharing. The initial complexities and benchmarking results are promising, but there is room for improvement.
The world today is full of secrets. Sometimes, we would like to know something about them without revealing the secrets themselves. For example, whether I have more money than my friend or whether two satellites would collide without publishing their moving trajectories. Secure multi-party computation allows us to jointly compute some functions while keeping the privacy of our inputs. Sharemind is a practical framework for performing secure multi-party computations. In this work, we added a protocol suite to Sharemind. This protocol suite was inspired by Shamir's secret sharing scheme, which describes a way to divide a secret into pieces. We describe algorithms for addition, multiplication, equality-testing and less-than comparison. We also give correctness and security proofs for the protocols. The resulting implementations were compared to an existing protocol suite inspired by additive secret sharing. The initial complexities and benchmarking results are promising, but there is room for improvement.