The Multimedia Intelligent Network of Unattended Mobile Agents (Minuteman), a portable airborne network system, will provide local communications for the increasing array of unmanned air vehicles (UAVs) and unmanned ground vehicles (UGVs). This agile, dynamic “Internet in the sky” will support the demanding communications requirements of unmanned missions, according to Mario Gerla, UCLA professor of computer science at the Henry Samueli School of Engineering and Applied Science.
“Minuteman will enable the Navy to bring fully networked force to the battlefield,” Gerla said. “This will be the ‘glue’ that holds together supporting technologies such as mission planning, path planning, reasoning, decision making and distributed real-time computing and control.”
Gerla heads the $11 million, five-year project funded by the Office of Naval Research. Such a network could also aid emergency workers responding to natural or manmade disasters, Gerla said. The project is part of a program on Intelligent Autonomous Agent Systems. Allen Moshfegh oversees the program for the Office of Naval Research.
Designing such a network, however, presents a number of challenges. The first challenge is handling a vast array of agents, some moving at the roving speed of the UGVs, while airborne assets travel at speeds of several hundred miles per hour. “The traditional Mobile IP approach will not scale up to such a large number of mobile agents moving at such high speeds,” Gerla said.
To meet this challenge, UCLA researchers are developing an addressing system that will exploit the fact that agents typically move in groups. As part of the new addressing scheme, the address of each agent will includes the name of its group.
UAVs gathering intelligence at the forefront, however, must be able to transmit multimedia streams with bandwidth guarantees. This presents a second challenge. Traditional quality of service methods do not meet these needs. To accomplish this, researchers have replaced the concept of guaranteed quality of service with that of “adaptively renegotiable” quality of service.
For example, the video signal must be dynamically adjusted to the available bandwidth. “To this end, our team is developing innovative, flexible video encoding schemes,” Gerla said.
Additionally, Automatic Target Recognition data must be received with extremely tight time constraints. Scanning of the entire area may not be feasible because too much time would be required. To compensate for this, researchers are devising a distributed database that can provide global information about assets in the battlefield as well as video images captured during routine surveillance. The database will supplement the UAV and sensor-image data with stored information, reducing time to target detection and recognition.
In addition to their individual duties, some agents can support multiple functions. For example, a UAV can be used as a node on the network, a remote camera for gathering intelligence video as well as a weapon.
The network in the sky must also be distributed and fault tolerant in order to dynamically adjust to failures due to natural causes or enemy actions.
“If one plane goes down, we don’t want to lose the whole network,” Gerla said.
In addition, a demanding network like this requires a new family of advanced, programmable radios.
Because it is important to know how well these components operate together, Gerla said, the research team is developing a novel simulator platform on which to test them. These concepts will be integrated and demonstrated in a sophisticated test bed that will include sensors, robots and helicopters being assembled under the direction of John Villasenor, UCLA professor of electrical engineering.
Also involved in the project are UCLA researchers Babak Daneshrad, associate professor, electrical engineering; Izhak Rubin, professor, electrical engineering; Gregory J. Pottie, professor, electrical engineering; Mani Srivastava, associate professor, electrical engineering; and Rajive Bagrodia, professor, computer science.