|
| |
|
|
Cloud computing
Bräuer, Jonáš ; Dzurenda, Petr (oponent) ; Ricci, Sara (vedoucí práce)
Security and privacy preservation of data stored on public cloud platforms is a widely addressed topic nowadays. In this thesis we explore the possible use of data splitting technique as opposed to more traditional encryption of sensitive data. Goal of this thesis is to implement and compare two already proposed protocols for matrix product computation, that use two different data splitting approaches. In addition, from these two protocols, working in non-colluding scenario, an original variant was proposed, adding colluding scenario compliance. All three protocols have been implemented using Java platform and real public cloud providers, and both performance and storage requirements have been measured.
|
|
|
Cloud computing
Bräuer, Jonáš ; Dzurenda, Petr (oponent) ; Ricci, Sara (vedoucí práce)
Security and privacy preservation of data stored on public cloud platforms is a widely addressed topic nowadays. In this thesis we explore the possible use of data splitting technique as opposed to more traditional encryption of sensitive data. Goal of this thesis is to implement and compare two already proposed protocols for matrix product computation, that use two different data splitting approaches. In addition, from these two protocols, working in non-colluding scenario, an original variant was proposed, adding colluding scenario compliance. All three protocols have been implemented using Java platform and real public cloud providers, and both performance and storage requirements have been measured.
|
|
| |
host ::
RSS.