Phase II Amount
$1,499,622
Robust and reliable communications is the cornerstone that maintains data transport between sensors, shooters, and command and control systems. This no-fail control system is maintained by a complex combination of modern and legacy communications systems with wildly varying capabilities and limitations. To improve the resiliency of this network, MTI Systems will adapt the FlowAgility network intelligence controller, leveraging complex physics models, weather prediction, and motion propagation models to generate and enact optimized network plans for both the current and future state of the network. Predictive planning is critical not just reacting to network outages, but preemptively avoiding them, reducing or eliminating network downtime. The current FlowAgility controller has been developed for both commercial and government use cases. This tool was developed to support the control of network edge devices which can utilize multiple commercial satcom, terrestrial, and other networks to backhaul data for command and control. For AFNWC, this product will be adapted in three distinct areas during this Phase II engagement: physics modeling and predictive analysis, the FlowAgility optimization algorithm, and NC-3 hardware control interface. Physics models must be developed for the specialized communications protocols and waveforms utilized by NC-3. This includes beyond line of sight (BLOS), MILSATCOM, COMSATCOM, and other communication systems unique to the NC-3 use case. This will enable the FlowAgility controller to estimate the link quality and network performance of available network paths between shooters, sensors, and command and control. Predictive analysis is a unique feature of FlowAgility that sets this tool apart from other commercially available SDN solutions. This integrates predictive motion, weather, and other analyses that enable the FlowAgility controller to generate a network plan for both the current network state and the planned future state of the network, enabling preemptive link planning that can avoid foreseeable network impairments. The FlowAgility network optimization algorithm utilizes machine learning to quickly analyze the current and future state of the network and generates a time-varying network plan tuned towards customer-specific requirements. Here network performance can be tuned to maximize throughout, minimize delay, or to organize network flows into redundant network paths. Enabling system redundancy is key to fast or no-loss recovery from network failures, which is critical for no-fail networks such as NC-3. Finally, in Phase II we will focus on developing the NC-3 hardware interfaces that will enable the FlowAgility controller to configure network hardware and enact the developed network plan. At the end of Phase II, we will have a network control algorithm, capable of NC-3 edge device optimization, that will enable reliable, robust, optimized communications throughout the NC-3 communications network.
