La capacidad humana para procesar información de EW del campo de combate está hoy en día sobrepasada. El Ejército de EEUU quiere utilizar la IA y técnicas de aprendizaje mecánico para monitorear, evaluar y contrarrestar las amenazas en ese campo. La detección y análisis de emisiones de radiofrecuencia, los patrones tipos de emisor y las estructuras de señales están dentro del alcance de la capacidad que se busca, así como la coordinación de las operaciones defensivas y ofensivas de EW…
Facing an electromagnetic jungle of electronic warfare threats, the Army wants smarter computers for EW and jam-proof navigation.
The Army has put out a request for information for machine learning technology — computers capable of learning on their own how to track electronic warfare threats and manage responses.
The problem is that soldiers are overloaded with trying to keep track of threats. «Compounding the issue is the source and location information of, potentially, infinite number of electronic signals through various sensing means, coordinating defensive and offensive operations to defend against electronic attack, and blunt the electronic means of the threat,» said the RFI, which was issued by the Assistant Secretary of the Army for Acquisition Logistics and Technology Rapid Capabilities Office.
One solution is artificial intelligence, or AI, that can quickly analyze, adapt and respond to threats. The Army wants «solutions capable of using machine learning techniques on monitoring and assessing threat Radio Frequency emissions, to establish normal and abnormal patterns of life, and to characterize emitter types and signal structures are within scope of this capability. Solutions capable of integrating with and analyzing Army data, both tactical network and Intelligence, Surveillance, and Reconnaissance (ISR), to include electro-optical sensors and Electronic Support Radio Frequency Electronic Warfare sensors, are also within the scope of this capability.»
The Army is interested in two forms of machine learning: supervised learning, which involves data already labeled, and unsupervised learning, in which the software must determine structure for uncategorized data.
In addition, the Army hopes that smarter machine intelligence will help create alternate means of positioning, navigation and timing, or PNT, in situations when GPS is jammed, or the signal can’t get through in urban or subterranean environments.
The RFI includes 13 questions regarding the ability of potential contractors to respond to EW and PNT requirements. Among the questions are:
* Is there analysis, testing, demonstrations, or operational usage available showing the utility of potential EW capability/concepts in relevant scenarios?
* Are there capabilities available, including machine learning and artificial intelligence, to assist in understanding local Electromagnetic spectrum usage and performing Command and Control (C2) of available EW assets?
* Are there capabilities available to provide ground units, Army aviation, or munitions with alternative means of establishing PNT information?
* Does your solution provide full PNT or just a sub-component (ie. just positioning, just navigation, just timing, or any combination)?
* How robust and resilient is the [PNT] capability, e.g., radio frequency interference, weather changes, environmental effects, etc.?
Fuente: https://defensesystems.com
