keepcalculator

06 Numbers and Powers

IP Bans Explained: TCP & MTU Fingerprints

Card No. 06·81770 · Updated · Standard formula, shown below

IP bans block traffic from a specific address, but TCP and MTU fingerprints reveal what kind of device is sending that traffic, not who is behind it. This article explains what those fingerprints can and cannot tell a network administrator, and how they are used alongside IP bans.

Background: Why IP Bans Are Not Enough

An IP ban is a blunt tool. It blocks all traffic from a single address, but addresses are easy to change—a user can reboot a router, use a VPN, or switch to mobile data. That is why network administrators look for more stable identifiers. TCP and MTU fingerprints are two such identifiers. They describe the technical behavior of the client’s network stack, which is determined by the operating system, the TCP/IP stack implementation, and sometimes the network hardware. These fingerprints do not identify a person, but they can tell an administrator that the same device is connecting from a new IP address.

How TCP Fingerprinting Works

When a device connects to a server, it sends a TCP SYN packet. That packet contains a set of options and parameters that are set by the operating system’s TCP stack. The exact combination of these values is called a TCP fingerprint. Key fields include:

  • Window size – the receive window advertised by the client.
  • Maximum Segment Size (MSS) – the largest segment the client can receive.
  • TCP options – such as SACK permitted, timestamps, and window scale.
  • TTL (Time to Live) – the initial hop limit set by the OS.
  • Flags – the control bits set in the SYN packet.

For example, a Windows 10 machine typically sends a window size of 64240, an MSS of 1460, and a window scale of 8. A Linux machine (kernel 4.x) might send a window of 29200, MSS of 1460, and a window scale of 7. These differences arise because each OS implements TCP differently. Tools like p0f and Nmap’s OS detection use these signatures to guess the operating system.

What MTU Fingerprinting Adds

The Maximum Transmission Unit (MTU) is the largest packet size a network interface can send without fragmentation. Most Ethernet networks use an MTU of 1500 bytes, but VPNs, PPPoE connections, and some tunnels use lower values like 1492 or 1400. When a device sends a packet that is too large for a link, it gets fragmented or dropped, and the device must adjust. The way a device handles this—whether it uses Path MTU Discovery, how it reacts to ICMP “fragmentation needed” messages—can be observed and used as a fingerprint.

MTU fingerprinting is less common than TCP fingerprinting, but it can distinguish between, say, a device behind a VPN (MTU 1400) and one on a direct Ethernet connection (MTU 1500). It is often used in conjunction with TCP fingerprints to narrow down the device type.

What Fingerprints Can Reveal

Fingerprints can tell you the operating system family (Windows, Linux, macOS, iOS, Android), sometimes the specific version or kernel, and occasionally the type of device (a printer, a router, a smartphone). They can also reveal if the client is using a VPN or proxy, because those services often alter the TCP stack parameters. For example, a VPN client might use a different window scale or MSS than the underlying OS.

In the context of IP bans, fingerprints can help identify a repeat offender who changes IP addresses. If the same TCP fingerprint appears from a new IP, it is likely the same device. This is a technique used by some websites to enforce bans more effectively.

What Fingerprints Cannot Reveal

Fingerprints cannot tell you the identity, location, or physical device of a user. They are not a unique identifier like a MAC address or a serial number. Many devices share the same fingerprint because they run the same OS version. For example, all iPhones on iOS 17 have very similar TCP fingerprints, so you cannot distinguish one iPhone from another. Also, fingerprints can be spoofed. A determined user can modify their TCP stack to mimic another OS, or use tools that randomize the parameters. This is a common technique in penetration testing and privacy tools.

Moreover, fingerprints are not static. OS updates change the TCP stack, so a fingerprint may change after an update. Network conditions also affect some parameters: the TTL can decrease with each hop, and the window size can be scaled differently depending on the connection.

How IP Bans Use Fingerprints in Practice

When a server decides to ban an IP, it can record the TCP and MTU fingerprints of the offending connections. Later, if a new connection arrives from a different IP but with the same fingerprint, the server can flag it as a potential bypass. The server might then apply a more aggressive CAPTCHA, rate-limit the new IP, or ban it outright. This is often called “behavioral fingerprinting” or “dynamic banning.”

However, this approach has false positives. A corporate network with many Windows 10 machines will produce many identical fingerprints. Banning all of them because one user misbehaved would block innocent users. Therefore, fingerprint-based bans are usually used as a signal, not a definitive proof.

Limitations and Evasion

Fingerprints are not a security boundary. They are a heuristic. A user who knows how to modify their TCP stack can evade fingerprint-based bans. Tools like Scapy allow crafting custom packets, and some VPN services intentionally normalize their TCP fingerprints to be generic. Also, mobile networks often use carrier-grade NAT, meaning many users share the same public IP, and their fingerprints may be altered by the carrier’s middleboxes.

Another limitation is that fingerprinting only works on the TCP handshake. If the connection is encrypted (HTTPS), the server can still see the SYN packet, but the payload is encrypted. That is enough for fingerprinting, but it does not give any application-level information.

Frequently Asked Questions

Can a TCP fingerprint identify me personally?

No. A TCP fingerprint identifies the type of operating system and sometimes the device, but not the user. Many people use the same OS and have identical fingerprints. It is a class identifier, not a unique identifier.

Can I change my TCP fingerprint?

Yes, you can. On Linux, you can use tools like `iptables` to modify TCP options, or use a user-space TCP stack. On Windows, you can edit the registry to change window size and other parameters. However, it is not trivial, and any change may affect network performance.

Does using a VPN hide my TCP fingerprint?

It depends. A VPN encapsulates your traffic, so the server sees the VPN server’s TCP fingerprint, not yours. However, if the VPN client modifies your local TCP stack (e.g., for MTU adjustment), the fingerprint may change. Some VPNs intentionally normalize their servers’ fingerprints to be generic.

How accurate is OS detection via TCP fingerprints?

It is fairly accurate for unmodified, up-to-date systems. Accuracy drops when the OS is old, patched, or customized. Also, firewalls and load balancers may alter the packets, reducing accuracy. In practice, it is a good guess, not a certainty.

Can MTU fingerprinting be used alone to identify a device?

No. MTU is determined by the network path, not just the device. Two different devices on the same network will have the same MTU. It is only useful when combined with other fingerprints.

Is fingerprint-based banning legal?

Generally, yes, as it is a technical measure to protect a service. However, privacy regulations like GDPR may require you to inform users if you collect such data. Always check local laws and terms of service.

Related

Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.

More cards in this drawer

The whole Numbers and Powers drawer →