Intelligent Network Traffic Management

We offer to research, design and prototype portions of an intelligent messaging framework that provides a layered approach to the problem, and leverages and integrates with commercial SOA ESB technologies. The first layer applies content filtering on the producer or consumer side based on network conditions, filter probability and message distribution. The second layer provides state filtering which determines whether the content has changed enough from the last transmission based on consumer constraints. The third layer uses a commercial product, DDS, to apply history, prioritization and queuing QoS. The fourth layer applies transactional filtering which removes elements that have not changed from the last transaction. The fifth layer applies binary compression using the Efficient XML emerging standard. The fifth layer applies network topology aware routing, leveraging MANET developing techniques. In summary, messages are sent based on consumer content and state constraints, reliably delivered using DDS QoS, reduced in size by only sending what changed and compressing the message, and uses knowledge of the network topology to reduce redundant messaging across the same connection. The framework approach is to use JBoss, an Open Source ESB to manage the rules at each layer and the workflow.