You can set up an intrusion detection/prevention system, that logs/blocks certain traffic. If you do have public services running, you could block access based on location, lists of known bad actors etc. I guess you could argue that this is beyond the scope of a traditional firewall.