I. Introduction
With the advent of hybrid and multi-tier distributed architectures, such as hybrid clouds and fog-cloud environments, where the physical location and characteristics of each resource may vary, meeting the security requirements of data processing has become a very complex challenge. Some data may be sensitive and thus facing higher security risks, such as data interception and unauthorized access. Consequently, they should be processed on secure resources that provide a higher level of security. On the other hand, non-sensitive data typically have a lower security risk and can be processed on resources with only a basic level of security. Hereafter, for simplicity purposes, resources providing only a basic level of security will be referred to as non-secure resources.