Analysis services scripting language (ASSL) is a specification language for autonomic systems. Its goal is to provide a way to specify self-protecting systems that are able to monitor and repair themselves.
This paper uses ASSL to provide a specification for a distributed modular audio recognition framework (DMARF). DMARF can be used for such things as determining the identity of a speaker in recorded audio. Outside a university laboratory environment, such a system needs to be protected from a number of threats, such as the potential intercalation of a malicious element into one of the pipelines. One approach to handling this problem is to require each node to prove its identity to the other nodes in the pipeline. This approach is similar to the way the domain name system (DNS) security extensions (DNSSEC) operate. Given this mechanism, self-protection requires one to specify how the components of the system manage the message authentication issues that arise in the case of spurious messages. This is where ASSL comes in. ASSL is based on viewing an autonomic system as having three tiers: the autonomic system (AS) tier that defines the service-level objectives, the AS interaction tier that defines the interaction protocols, and the autonomous element (AE) tier that defines the behavior of low-level units. Using ASSL, Mokhov and Vassev specify the way the system manages messages and their authentication.
The main body of the paper is supplemented with an appendix that provides the detailed ASSL specification for DMARF. Before reading this paper, the interested reader would do well to read one of the earlier papers on ASSL referenced by the authors, in order to gain a more in-depth understanding of it.
This paper uses ASSL to specify how self-protection can be incorporated into an existing system. It would be interesting to see a case where ASSL is used from the start.