25-08-2017, 09:32 PM
Ranked, Efficient and Secure Keyword search over encrypted cloud data
Efficient and Secure Keyword.pptx (Size: 69.77 KB / Downloads: 11)
Existing Problem
Private and Government, IT and non-IT companies require to store their big and confidential data for easy and efficient access.
2 options:
a) Owning a data center - Too difficult and cost heavy but secure.
b) Cloud services (SaaS) - Too easy and cheap but basically
insecure.
For cloud services, encryption can overcome insecurity but is difficult to efficiently and securely search for relevant data among hundreds of thousands of files.
We need a mechanism to search and retrieve the data in the encrypted domain from the server.
System Design
With an MVC architectural pattern in mind, the system is divided into 8 different modules as follows:
Client Authentication
Data Acquisition
Key Generation and Management
Data Processing
DTP
Storage Manager
TrapdoorGen
Search Index
Data Acquisition
This module solely deals with the acquirement of data from the user, along with its verification.
It is of 2 types:
File acquisition
Keyword acquisition
- For file acquisition, it checks the type of file being uploaded as it supports only plain text files as of now. Other files are rejected.
- For keyword acquisition, it matches against the whitelist of characters to prevent malicious character entry.
Search Index
This module is also a part of the data retrieval process which unlike the previous module, sits on the server.
When a trapdoor of the keyword(s) is received by the server, this module searches the stored keywords index to find the relevant documents.
Search index
Retrieve the relevant files.
Sort the files based on its score.
Check for user authorization on each file.
After this, the DTP is used to transfer the files back to the client.