Berkeley Realtime-Application Network Demonstration

The BRAND program is a development and demonstration of two network applications that require the capacity and/or low latency of an open testbed communications network such as that provided by the Next Generation Internet (NGI) system program at DARPA. The resulting demonstrations created by this effort (sensor Web and networked MEMS CAD) will demonstrate the benefits of an open research network capability based on an optical transport system and associated high performance/high capacity networks and management systems that are ultimately necessary to enable these new stressing applications. As high performance networks continue to evolve, the difficulties of providing guaranteed performance, low-latency connectivity service have grown in importance on the networking research agenda. For military applications in particular, a growing emphasis on reachback makes sub-second latencies increasingly important in achieving high quality real-time interpretation of sensor feeds from various sources. With ever-increasing sensor sophistication and bandwidth requirements this requirement is expected to place increasingly taxing demands on existing network infrastructure.

Deployed ground sensors hold out the promise of affording analysts wide area monitoring and reconnaissance of personnel movements within specified critical areas.While the information flow from each individual sensor begins as a trickle, an immediate demand for deployment of tens of thousands of these sensors immediately leads to the need to distribute a composite real time operational picture with sizable network requirements. In addition, it is anticipated that next generation based sensors are likely to include a video or imaging component, so networks must be engineered to scale to meet these demands as well. This work entails a range of activities including routing algorithm development to bring data to an exfiltration node, engineering of communications network interfaces to preserve the low latency connections, network engineering to ensure appropriate resources are in place, and human interface development to ensure the human users are not misled by the vagaries of network transmissions.
Real time visualization on the microscale is also an important future application of high performance networks. This work addresses the potential use of high performance networks as a tool in the design and manufacture of microsystems. The rising investments required to build foundries that support successive advances in sophistication bring networks into play, with their ability to allow operators to remotely share scarce facilities. Our goal is to close the design loop by enabling design, simulation, fabrication, comparison of measurement with simulation or other data, diagnostics, and then redesign. The model is that a user would be able to use all the facilities (simulation, measurement, and data repositories) remotely at high speed. Speed is important because of the enormous measurement files produced and the ability to control and observe the devices being measured in real time. In addition to simulating and measuring devices from our own local users, we will identify and support outside users to make sure that the system supports their needs.