SIGNetS: signal and information gathering for networked surveillance

Decentralized intelligence, surveillance, and reconnaissance (ISR) play a significant role for situational awareness and decision-making in a cluttered battlefield environment. However, data collected by decentralized or distributed sensors may be corrupted by noise, of large volume, in high dimensions, or unreliable due to failures of sensors or communication nodes. Signal and information processing is an enabling technology for dealing with heterogeneous sensory data, extracting actionable information from such data, and addressing the challenges associated with scalability, dimensionality, and uncertainties. This may be achieved with model-based approaches (physics-based or statistical models), or data-driven approaches (such as machine learning). In this talk, we will provide an overview of the related work in the SIGNetS project, a collaboration among Cambridge, Sheffield, and Surrey Universities. We will also provide a perspective of using machine learning approaches to improve the resilience and efficiency of distributed sensors operating in a cluttered (undersea) environment.