Cognitive Airborne Networks for Defense Operations (CAN-DO)

The CAN-DO project will create an adaptive cognitive network whose components work together within the rapidly changing challenges and constraints of an airborne tactical wireless MANET to enable fast and reliable delivery of one-to one, one-to-many and many-to-many information. Such a network will be capable of modifying its own behavior in response to changing network conditions to satisfy its users requirements.  According to this vision, the resources for which the network is responsible include not only the traditional nodes, ports, packets, and interfaces, but also the content of the information it carries.  The central focus of our proposal is a cross-layer network architecture, the sharing of information between layers of network function that have traditionally hidden their information from one another, as a unifying principle for solving the problems of one-to-many and many-to-many data exchange over an unstable wireless infrastructure.  We utilize network coding principles and tools to improve network reliability and capacity for one-to-many and many-to-many information exchanges.  We also identify the key helper technologies that will act together to make the architecture real: Software-Defined Radio, the Click Modular Router software, a unifying framework to implement an adaptive network stack, and a Cognitive Engine for network state-sensing and decision-making. BENEFIT: The founding vision of mZeal is to make real the possibilities of autonomous mobile networking, and all our work is guided toward that goal. The CAN-DO architecture is an incremental step in that direction, following logically from our previous work with adaptive applications for tactical MANETs. CAN-DO will incorporate the technologies that we have determined to be most critical to accomplishing our corporate goals, including a flexible and extensible cognitive reasoning capability, a framework for gathering and reporting information using mobile intelligent agents, and a conceptual cross-layer model of the network that allows us to implement novel solutions to the novel problems posed by node mobility and self-government.  We see our near-term commercial opportunities in the places where the challenges to networking are most difficult - in tactical military operations on the ground, sea, and in the air; in search and rescue operations; in many aspects of emergency response in challenging environments, indoor and outdoor. We are excited by the opportunity to work with iRobot Corporation, because we are sure that the technical capabilities and the goals of the two companies are not simply compatible but truly synergistic, creating a whole greater than the sum of its parts. The immediate relevance of robotics in CAN-DO is to the participation of unmanned vehicles in the network, but outside the military arena, robotic vehicles have applications in the other areas we named as well. In our vision, anywhere that people cannot or will not go, robots can and will.  The purpose of any network is to disseminate useful information to people (and devices) who need it. This is why our first products were adaptive applications; the first problem to solve is to deliver information in usable form. CAN-DO will be measured by that standard, and our commercialization decisions will be driven by that understanding.  We see two avenues to commercialization of CAN-DO. One avenue is to package the components in preconfigured form for particular applications; we can imagine, for instance, emergency response vehicles with additional special intelligence about forest fires or hazmat spills. The other avenue is to offer system integration services to resellers or end users.