Hello Aytekin,
If I don't get you wrong, you're connecting to a computer in the LAN over the Internet with RDP. the internal computer is communicating to the internet with a public ip address on the firrewall. this is also called NAT. Within NATed traffic, the destination ip address is changed to your computer's ip address and the source ip address is changed to your firewall local ip address. the real public ip address knows only the firewall, your computer inside does not know. when replying to it, the source ip address to which it sends packets back is the firewall local ip address. this is the nature of the NAT process.
Install wireshark on the computer you are connected with RDP and start capturing traffic, then re-apply the scenario. Stop capturing after connecting with RDP. review wireshark logs; As you can see, the source ip address field will be 192.168.1.1 (firewall local address or default gateway), and the destination ip address will be the computer you are connecting to as 192.168.1.5. you will find another field called originate source ip address in the same capture package. this will be the public ip address you connect to from the outside. This last field does not appear in Windows log records. Because windows knows that the source that speaks to itself is the default gateway (firewall). this information appears in the pcap log, because pcap records low level traffic.
I suggest you install Suricata on your local network. mirror all internal traffic of the firewall to the Suricata via the switch. Integrate Suricata with Wazuh. You can also take pcap logs that are also created in Suricata as log on Wazuh. after that you can create an alarm by following the rules on Wazuh. I did this in my own environment.
Zufikar CAGLAR