6.9.1. Network Delay-based Model
Benson, Dowsley and Shacham : 1 proposed a model for checking if distinct copies of data are kept at different geographic locations (BDS model"). The idea behind, is that a verifier must receive an answer within a certain time.
The server which does not posses its own copy and has to retrieve it first from other servers at different locations - which takes additional time.
Assumption 1
The locations of all data centers of the cloud provider are known.
Assumption 2
The cloud provider does not have any exclusive Internet connection between the data centers.
Assumption 3
For each datacenter \(s\), it is possible to have access to a machine that is
located very close to \(s\) (i.e. the network latency is small).
Client checks if server \(s_i\) has a copy of the file
- Let:
-
- \(s_i\), \(s_j\) be colluding malicious servers who only store one copy of the file at \(s_j\).
- \(T_i\) be the upper bound on the execution time an auditing protocol run at \(s_i\),
- \(t_i\) be the network delay between the client and \(s_i\),
- \(t_{i,j}\) be the network delay between data centers \(s_i\) and \(s_j\).
- \(max(t)\) to denote the upper bound on \(t\)
- \(min(t)\) to denote the lower bound on \(t\).
- Attack:
- if \(s_i\) is queried, it redirects the challenge to \(s_j\) , which answers directly to the client.
- Discovery:
- if (answer received in: \(T_i + 2\cdot max(t_i)\)) then \(s_i\) is honest else if (answer received in: \(min(t_i)+min(t_{i,j})+min(t_j)\) ) then attack is probable.
Benson, K., Dowsley, R., Shacham, H.: Do you know where your cloud files are?
In Cachin, C., Ristenpart, T., eds.: Proceedings of the 3rd ACM Cloud Computing Security Workshop, CCSW 2011, Chicago, IL, USA, October 21, 2011, ACM (2011) 73{82. http://doi.acm.org/10.1145/2046660.2046677
Chen, B., Curtmola, R: Towards self-repairing replication-based stor-age systems using untrusted clouds.
In Bertino, E., Sandhu, R.S., Bauer, L., Park, J., eds.: Third ACM Conference on Data and Ap-plication Security and Privacy, CODASPY'13, San Antonio, TX, USA, February 18-20, 2013, ACM (2013) 377-388. http://doi.acm.org/10.1145/2435349.2435402