Re: Inteligencia artificial, redes neuronales, machine learning
Publicado: Mié Mar 23, 2022 6:55 pmSobre el programa AIRE ("Artificial Intelligence Radar Enhancement"): https://sam.gov/opp/aac2e6bcc5d14d4dad1 ... ac0aa/view
...
BACKGROUND: Current Army radar systems utilize predetermined waveform schedules and a fixed number of waveforms for search and track functionality. These settings are typically configured prior to deployment and as a result are difficult to change in the field in order to adapt to a rapidly changing threat environment. This leaves the radar system vulnerable to Electronic Warfare (EW) attacks and evolving threat target tactics. The objective of the Artificial Intelligence Radar Enhancement (AIRE) is to overcome these limitations through Artificial Intelligence (AI) and Machine Learning (ML).
DESCRIPTION: The AIRE will be anAI based radar scheduler and waveform selector that autonomously manages the search and tracking aspects of a radar system. Solutions can consist of hardware and/or software, but must support integration into existing radar systems. The AIRE should be developed using advanced AI and/or ML techniques, i.e. reinforcement learning. Any candidate solution should address the following capabilities:
· Maximize the use of radar resources and optimize tracking waveform choice based on the operational environment or engagement constraints.
· Provide an adaptive search volume and obey a threat priority hierarchy provided by the user.
· Automatically detect and recognize EW countermeasures that are a threat to the radar system.
· Alert the radar operator of any detected EW threats and provide information on the type of threat, location of the threat, and the effects on the radar system.
· Dynamically change the radar’s waveforms and beam scheduling in order to adapt to any EW threats while maintaining search and track performance to the greatest extent possible.
· Allow the customer to deploy antagonistic scenarios and training sets for continued AI learning.
...
