The main focus of Software Separation is to identify and assess hazards through auditing the actions of users for all applications and cross domains to prevent unauthorized access to data/information in an Information System. A Software Separation architecture must employ various assurances that each classification level of data/information is kept separate and do not intermingle. These hazards may be:
· Internal
- An unauthorized user attempts to gain access to higher classification levels of date for which he/she is not authorized.
- An authorized user attempts to acquire information for the purpose of unauthorized use.
· External
- A vehicle is captured intact and the encryption/decryption keys have not been zeroized, thereby allowing the captor to use the system operationally or access the data/information within the Information System.
There are several levels of hazard severity. By assessing credible actions that unauthorized users might take to gain access, risk mitigations can be determined to ensure unauthorized users cannot accidentally or covertly access data/information due to architectural deficiencies. The risk of loss is unlikely depending on the number of security layers implemented. Through the use of Role Based Access Control and Auditing risk mitigators can keep risk low.
· Employ strong authentication for all users.
· Successful login requires user authentication.
· Decrypted private keys are held within a key-holding agent or hard token.
· Login is replaced by login Proxy in each application domain.