Here’s an article on headless initial configuration on the Jetson Nano:
You will need to have a freshly flashed SD card running L4T 32.2+
Also, you will need to connect the Nano to your Linux, Macintosh or Windows host using a USB cable plugged into the Micro USB connector on the Nano. Make sure that the cable is data capable, a USB phone charger cable may not work. You will need to power your Jetson using the 5V 4A power jack, don’t forget the jumper on J48!
After powering on the Jetson, it takes ~ 30-45 seconds to boot with a default image. You should then be able to use a serial terminal application (such as screen on an Ubuntu or Macintosh, or puTTY on a Windows machine) and connect to the Jetson.
On a Linux machine, the Nano will present itself as /dev/ttyACM* (typically /dev/ttyACM0). On a Macintosh it is /dev/tty.usbmodem*. On Windows it is a COM port. The speed of the serial connection should be set as 115200. Connect to the Jetson (you may have to hit the ESC key a couple of times) and go through the configuration. The configuration is covered in the official NVIDIA L4T documentation: https://bit.ly/2ZmXJS1
There are several network paths you can select (Ethernet cable, wireless, USB and so on).
Once you have configured the Jetson, the Jetson will reboot and you will be able to SSH into it. You may find it useful after rebooting to serial app back into it to figure out what IP address is assigned to the Jetson Nano. You will be able to log into the account through the serial debugging port that you set up in the preceding steps.