U.S. military and intelligence agencies are looking for ways to increase data security in wildlife communications.
Among the proposed approaches are emerging intelligent radio platforms that can provide classified information. One of the efforts is the Intelligence Advanced Research Projects Activity (IARPA), which seeks industry proposals for a program called Shared Information Protection with Smart Radio Systems or SCISRS.
The program seeks to fill security gaps between sensitive government facilities where data is transmitted, received and stored. For example, transmissions in the US Navy’s stretched base can be disrupted by unexpected radio frequency transmissions.
“The goal of the SCISRS program is to develop intelligent radio techniques to automatically detect and characterize these suspicious signals and other RF anomalies in complex RF environments,” explained Paul Kolb, IARPA Program Manager. Anomalous signals are often used to mask exploits designed to disrupt RF communications.
Examples of exploits range from unexpected transmissions to a specific location, “or signals that don’t otherwise belong,” Kolb said. “A signal that mimics a cell tower, for example, is not a good sign. This means that something may be trying to communicate with your device. “
Among the possible security solutions is the use of “advanced intelligent radio techniques to automatically detect and characterize RF anomalies,” he added.
Potential data security solutions include RF machine learning, software-defined radios, security platforms that combine GPUs and CPUs, and “classic” signal analysis, Kolb said recently.Candidates’ Day”Meeting held by IARPA.
The agency said emerging software radios and published datasets have improved the state of intelligent radio technology. “What once took expensive, specialized equipment can now be done with an SDR and a laptop,” Kolb said. Add to that the fact that large data sets opened up [RF machine learning] study for researchers without RF laboratory. “
The intelligent radio effort is also building on earlier work on radio frequency machine learning at the Agency for Advanced Defense Research Projects, Kolb added.
In selecting providers of smart radios, the agency said it was considering a combination of machine learning and signal processing techniques. They range from unobserved training and observation and a combination of neural networks and deep training to recovery signals and filtering techniques.
The deadline for proposals for the industry was November 13. The agency expects to award contracts for the development of intelligent radio early next year, with initial project reviews 18 months from now. Program officials said IARPA was looking for “complementary opportunities” as it sought industry teams to develop intelligent radio infrastructure.