Based on a thorough analysis and comparison of several business activity monitoring(BAM) system models, noting that a BAM system is open, integrative, and real-time, and using the SOA software engineering techniques, this paper presents a service-oriented dual-bus BAM system model and analyzes in detail the functions of its service components and communications among them. Much more open than prior models, this BAM system model has a two-level bus architecture, which enables flexible and extensible implementations of practical systems, as well as provides a foundation for software reuse.

While existing solutions can provide authentication services, they are insufficient for a new generation of network coding-based reprogramming protocols in wireless sensor networks. We present a security approach that is able to defend pollution attack against reprogramming protocols based on network coding. It employs a homomorphic hashing function and an identity-based aggregate signature to allow sensor nodes to check packets on-the-fly before they accept incoming encoded packets, and introduces an efficient mechanism to reduce the computation overhead at each node and to eliminate bad packets quickly. Castalia simulations show that when the 5% of the nodes in a network of 100 nodes are rogue, using our approach, the efficiency of the secure reprogramming protocol based on network coding improves almost ten-fold for a checking probability of 2%.

Reprogramming protocols provide a convenient way to update program images via wireless communication.
In hostile environments where there may be malicious attacks against wireless sensor networks, the process of reprogramming faces threats from potentially compromised nodes. While existing solutions can provide authentication services, they are insufficient for a new generation of network coding-based reprogramming protocols in wireless sensor networks. We present a security approach that is able to defend pollution attack against reprogramming protocols based on network coding. It employs a homomorphic hashing function and an identity-based aggregate signature to allow sensor nodes to check packets on-the-fly before they accept incoming encoded packets, and introduces an efficient mechanism to reduce the computation overhead at each node and to eliminate bad packets quickly. Castalia simulations show that when the 5% of the nodes in a network of 100 nodes are rogue, using our approach, the efficiency of the secure reprogramming protocol based on network coding improves almost ten-fold for a checking probability of 2%.