Premium
MMLT: A mutual multilevel trust framework based on trusted third parties in multicloud environments
Author(s) -
Aghaee Ghazvini Golnaz,
Mohsenzadeh Mehran,
Nasiri Ramin,
Masoud Rahmani Amir
Publication year - 2020
Publication title -
software: practice and experience
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.437
H-Index - 70
eISSN - 1097-024X
pISSN - 0038-0644
DOI - 10.1002/spe.2798
Subject(s) - cloud computing , computer science , trusted network connect , trust management (information system) , trusted computing , trusted third party , fuzzy logic , ranking (information retrieval) , trustworthiness , direct anonymous attestation , service (business) , set (abstract data type) , rank (graph theory) , service level , computer security , data mining , information retrieval , artificial intelligence , business , mathematics , marketing , combinatorics , programming language , operating system
Summary In this article, a mutual multilevel trust framework is proposed, which involves managing trust from the perspective of cloud users (CUs) and cloud service providers (CSPs) in a multicloud environment based on a set of trusted third parties (TTPs). These independent agents are trusted by CUs and CSPs and distributed on different clouds. The TTPs evaluate the CUs' trustworthiness based on the accuracy of feedback ratings and assess the CSPs' trustworthiness based on the quality of service monitoring information. They are connected themselves through the trusted release network, which enables a TTP to obtain trust information about CSPs and CUs from other clouds. With the objective of developing an effective trust management framework, a new approach has been provided to improve trust‐based interactions, that is, able to rank the trusted cloud services (CSs) based on CU's priorities via fuzzy logic. Fuzzy logic is applied to manage the different priorities of CUs, all the CUs do not have the same priorities to use trusted CSs. Customizing service ranking allows CUs to apply trusted CSs based on their priorities. Experiments on real datasets well matched the analytical results, indicating that our proposed approach is effective and outperforms the existing approaches.