No abstract

Abstract. Context-awareness is one of the fundamental requirements for achieving user-oriented ubiquity. In this paper, we present the design and approach to a middleware solution that expedites context-awareness in a ubiquitous computing environment. Context-Aware Middleware for Ubiquitous computing Systems (CAMUS)1 envisions a comprehensive middleware solution that not only focuses on providing context composition at the software level but also facilitates dynamic features retrieval at the hardware level by masking the inherent heterogeneity of environment sensors. Complexity is handled by providing 'separation of concerns' between environment features extraction, contextual data composition and context interpretation. Different reasoning mechanisms are incorporated in CAMUS as pluggable services. Ontology based formal context modeling using OWL is described. With a systematic approach, CAMUS is proved to be a flexible and reusable middleware framework.

No abstract

No abstract

In recent years, malware has grown extremely rapidly in complexity and rates of system infection. Current generation anti-virus and anti-malware software provides system protection through the use of locally installed monitoring agents, which are dependent upon vendor generated signature and heuristic based rules. However, because these monitoring agents are installed within the systems they are trying to protect, they themselves are potential targets of attack by malware. Pathogen overcomes this issue by using a real-time system monitoring and analysis framework that utilises Virtual Machine introspection (VMI) to allow the monitoring of a system without the need for any locally installed agents. One of the main research problems in VMI is how to parse and interpret the memory of an executing system from outside of that system. Pathogen's contribution is a lightweight introspection framework that bridges the semantic gap.