Opinion Model Based Security Reputation Enabling Cloud Broker Architecture - Cloud Computing - page 110

Where W AT EU1,EU2…,EUq is consensus opinion for CSP-1 given by enterprise user EU1
to EUq based on the authentication. Similarly W AR EU1,EU2…,EUq , W IA EU1,EU2…,EUq ,
W MT EU1,EU2…,EUq , W DL EU1,EU2…,EUq and W MP EU1,EU2…,EUq are the consensus opinion for
CSP-1 by EU1 to EUq based on authorization, inside attacks, multi-tenent attack,
data leakage and malware propogation respectively.

4.3 Trust of Cloud Service Provider

The trustworthiness(T) of the cloud service provider is given by calculating the
expectation of the opinions W M and W EUR given by Incidence monitoring and the
Enterprise User respectively. The trustworthiness(T) can be represented as:

T = Expectation (W M Ʌ W EUR ) = Expectation (W M Ʌ EUR )

(9)

Where and the expectation of the

W M Ʌ EUR = (b M Ʌ EUR , d M Ʌ EUR , u M Ʌ EUR , a M Ʌ EUR )
W M Ʌ EUR opinion is given as :

(10)

E(W M Ʌ EUR ) = b M Ʌ EUR + (a M Ʌ EUR )( u M Ʌ EUR )

5

Applicability of This Architecture

The cloud broker architecture proposed in this paper is a very generic and not limited
to any specific environment. However, a practical, environment specific
implementation of the proposed architecture is being used in the OPTIMIS [3][11]
project. OPTIMIS toolkit is a set of software components for simplified management
of cloud services and infrastructures that assists the cloud service providers to provide
optimized services based on the TREC (Trust, Risk, Eco-efficiency and Cost).
TREC components are part of the basic toolkit. The trustworthiness of an IP
(Infrastructure Provider) enables the SPs (Service Provider) to identify and select the
IP having proven capabilities to provide the required service. The risk assessment
performed provides the SP with the risk involved in the construction, deployment and
operation of a service. The eco-efficiency aids in selecting a cloud service provider
based on the energy consumption. Along with the trust, risk and eco-efficiency factor,
cost forms the trade-off factor in providing of the optimized service.
The broker architecture [11] in the OPTIMIS project already have a support of the
TREC toolkit, SLA agreement and the monitoring infrastructure which can be
enabled to build the security reputation of the IP using the proposed reputation model
[13] described in section 4 and the security related events captured in section 3.
Figure 3 shows the high level sequence diagram for broker implementation in
OPTIMIS project. Following are the sequence of steps: a) The SP uses the IDE
(Integrated Development Environment) to create a service which is described in a
service manifest b) The IDE passes the service manifest and the optimization
objective to the SD (Service Deployer) for deployment of the service c) The SD uses
the cloud broker interface to submit the service manifest and the optimization
objective d) The cloud broker has Registry where all SPs and IPs register before using
the cloud broker services e) The broker after receiving a request for deployment of a