It would be ideal find the root cause first. So, Log in to your VM and check the logs at that partcular time. Also, verify the host logs where the VM is located.
It could be a host disconnect for few secs etc. See if there are any alarms generated for host. Since you are saying this is the only VM having the problem, I would suspect VM related issue.
If you want to generate a custom alarm for this VM, you can do as the following.
Select VM -> Alarms tab -> New alarm -> Actions tab -> pick "Run a Command" -> Point to the script that can reboot the VM if it continuosly loose pings for 30 secs etc.
But, I would definitely find the root cause before implementing the alarm.