We're planning to implement following scenario for a secure machine-to-machine (M2M) communication network. We have some gateways and a lot of secure sensors which should only be able to "speak" to each other if they can authenticate themselves. No DNS is available, thus everything is working IP-address based.
The gateway will have a gateway.crt and gateway.key most probably stored inside a TPM. The gateway.crt is signed by an internal CA. The corresponding CA certificate CA.crt is stored on the gateway. The gateway.crt is created once during the production process. All gateways will have the same CN="TrustedGateway".
Each sensor will have it's unique sensor.key and sensor.crt. Sensor.crt is signed by CA.crt during production of the sensor and CA.crt is stored on the sensor in it's list of trusted CAs. All sensors will have the same CN="TrustedSensor".
Communication will be done using SSL / TLS with both server and client authentication required. In my understanding, the communication and connection establishment should work as both device types (gateway, sensor) have the CA.crt in their list of trusted CAs. What I'm unsure about though, is that all sensors and gateways have the same CN="TrustedSensor" / CN="TrustedGateway". Is there any security issue with this approach of all sensors and gateways having the same CN?
