System for data link networking based on machine learning

A method, comprising:

• receiving, at a network interface, state data including orientation information associated with at least one of a first remote platform or a receive antenna of the first remote platform;
• executing, by a processing circuit, a communication performance function using at least one of the orientation information associated with at least one of the first remote platform or the receive antenna of the first remote platform to estimate a performance of a communication link that a transmit antenna uses to transmit an output signal to a receive antenna of the first remote platform;
• evaluating, by the processing circuit, a performance condition based on the performance and a threshold performance;
• responsive to the performance condition being satisfied, transmitting, using the transmit antenna, the output signal to the first remote platform;
• and
• responsive to the performance condition not being satisfied, at least one of (i) adjusting, by the processing circuit, at least one of a time parameter or a frequency parameter that the transmit antenna uses to transmit the output signal to the receive antenna, or (ii) providing, by the processing circuit, an instruction to a second remote platform to cause the second remote platform to transmit the output signal to the first remote platform.

A method, comprising:

• receiving, by a machine learning circuit, a plurality of performance data points regarding performance of a communication link between a first platform and a second platform during a maneuver performed by the first platform;
• receiving, by the machine learning circuit, a plurality of state data points regarding at least one of the first platform or the second platform during the maneuver performed by the first platform, the plurality of state data points including orientation information associated with at least one of the first platform or a receive antenna associated with the second platform;
• assigning, by the machine learning circuit, each performance data point of the plurality of performance data points to a corresponding state data point of the plurality of state data points;
• providing, by the machine learning circuit, the plurality of state data points as an input to a candidate performance function, the candidate performance function including at least one parameter that operates on the input to generate a candidate output;
• executing, by the machine learning circuit, the candidate performance function to generate the candidate output;
• and
• adjusting, by the machine learning circuit, the at least one parameter based on the plurality of performance data points and the candidate output to output an adjusted performance function.