Heteregeneous Intelligent Filtering in tactical wireless mobile networks

In this paper, we propose a novel approach, which is termed heterogeneous intelligent filtering (HIF), for intelligent control and application data filtering in multi-domain heterogeneous networks. HIF creates intelligent gateways to rapidly and autonomously adapt the flow of information content to the changing mission needs and network characteristics. HIF employs a MANET management protocol (termed MMP) which enables a novel light-weight publish/subscribe mechanism. Interoperability is provided with upper echelon networks running legacy directory and session announcement protocols. MANET users can discover multicast information sourced in the backbone network and subscribe to only subset of the information that they need. HIF enables the MANET nodes to avoid being swamped by too much data - extraneous information not needed by the end-user is filtered by the HIF agents. Adaptation is triggered by QoS messages derived from ongoing application performance and network monitoring. We have implemented HIF in a laboratory testbed using the MANET emulation system, and successfully demonstrated it's publish/subscribe and filtering capabilities for real-time multimedia and tactical situational awareness applications. The experiments confirmed our conjecture that if adaptive and intelligent content filtering is performed, the end-user performance (and packet loss and latency) becomes significantly and consistently better during the periods when MANET nodes are subjected to overload conditions.