Super-resolution radar for autonomous vehicles

1. An autonomous driving system, comprising:

a sensor fusion module;

a system controller;

a memory;

a radar controller adapted for dynamic control of steerable beams focusing on one or multiple portions of vehicle field of view and is adapted to detect and identify targets, wherein:

the radar controller is adapted to interact with the sensor fusion module,

the radar controller is configured to implement a super-resolution network configured to receive radar data with a first resolution and produce radar data with a second resolution different from the first resolution using one or more first neural networks trained on a dataset comprising radar training data and ground truth densified data time-synchronized with the radar training data, and

the radar controller is configured to dynamically control one or more steerable beams based on detection and identification of targets with the radar data with the second resolution.

2. The system of claim 1 , further comprising an infrastructure sensor adapted to read a smart road configuration.

3. The system of claim 1 , further comprising a noise detection module to identify multiple radar signals.

4. The system of claim 3 , wherein noise detection is used to adjust a beam size of a transmitted signal.

5. The system of claim 1 , further comprising an environmental sensor configured to detect weather conditions.

6. The system of claim 1 , further comprising means for turning off sensors.

7. The system of claim 1 , further comprising means for reducing a sampling rate of at least one sensor.

8. The system of claim 1 , wherein the communication module is adapted to communicate within the vehicle and with other vehicles and store information in the memory.

9. The system of claim 1 , wherein the targets comprise traffic lines on a road.

10. An autonomous driving method, comprising:

determining a drive control based on sensor data wherein drive control includes vehicle steering, acceleration and braking;

determining environmental conditions of the vehicle, including weather conditions;

communicating with at least one other module in the vehicle;

receiving sensor information and providing to a sensor fusion module;

processing, with a super-resolution network, radar data with a first resolution and produce radar data with a second resolution different from the first resolution using one or more first neural networks trained on a dataset comprising radar training data and ground truth densified data time-synchronized with the radar training data; and

controlling a radar module for dynamic control of steerable beams focusing on one or multiple portions of vehicle field of view and is adapted to detect and identify targets, wherein the radar controller is adapted to interact with the sensor fusion module, and the radar controller is configured to dynamically control one or more steerable beams based on detection and identification of targets with the radar data with the second resolution.

11. The method of claim 10 , further comprising application of advanced decision making algorithms for interaction with sensors through the sensor fusion module.

12. The method of claim 11 , wherein data is processed by the sensor fusion module to assist in training and perceptual inference performance.

13. The method of claim 12 , further comprising providing vehicle to vehicle communications.

14. The method of claim 13 , wherein the vehicle to vehicle communications include data received from lidar systems.

15. The method of claim 10 , further comprising operating a perception module having a super resolution network and a target identification and decision module.

16. The method of claim 10 , wherein the super resolution network is trained to identify targets.

17. The method of claim 16 , wherein information of targets identified outside the vehicle is used to adjust operation of the vehicle.