Mifare cards: distinguish between 4-byte and 7-byte UIDs

uniqueidentifier rfid mifare smartcard-reader contactless-smartcard
bbaker picture bbaker · Jun 15, 2016 · Viewed 8.1k times · Source

I have a card reader that always report 64 bits, and can read cards with 4 or 7 byte UIDs.

As an example, I see it can report:

  • 04-18-c5-82-00-00-00-00 - a 4-byte UID in the form uid0-uid1-uid2-uid3-00-00-00-00
  • 04-18-c5-82-f1-3b-81-00 - a 7-byte UID in the form uid0-uid1-uid2-uid3-uid4-uid5-uid6-00

What prevents a 7-byte UID from having uid4, uid5 and uid6 set to zero? Is this covered in a spec? If so, which spec?

Answer

Michael Roland picture Michael Roland · Jun 24, 2016

What prevents a 7-byte UID from having uid4, uid5 and uid6 set to zero?

Nothing. The format of the UID (as used by MIFARE cards) is defined in ISO/IEC 14443-3. Specifically for MIFARE cards, NXP has (or at least had?) some further allocation logic for 4 byte UIDs, but that's not publicly available.

Is it possible to distinguish the two cases?

If the reader outputs the UIDs exactly in the form that you showed in your example, then the answer is no (at least not reliably). However, some readers output the UID on 8 bytes and include the cascade tag for 7-byte-UIDs. Thus, all 7-byte-UIDs start with 0x88 for those readers. This does not seem to be the case with your reader.

Are there possible strategies to distinguish the two cases?

Some strategies come to my mind to distinguish 4-byte-UIDs from 7-byte-UIDs.

  1. The first byte of a 7-byte-UID is the manufacturer code (as defined in ISO/IEC 7816-6 (see How to detect manufacturer from NFC tag using Android? on how to obtain the list). Thus, if you have a limited set of manufacturers (e.g. if you only use MIFARE cards with chips from NXP), you could interpret all UIDs that start with NXP's manufacturer code (0x04) as 7-byte-UIDs. Nevertheless, you should be aware that 4-byte-UIDs are allowed to start with 0x04 as well. Hence, this method is not 100% reliable and may fail for some cases.

  2. The first byte of 4-byte-UIDs must not contain any of the following values: 'x8' (with x != '0'), 'xF'. If you find the first byte to match any of those values, you can assume the UID to consist of 7 bytes.

Related questions

How to get the UID of RFID in Arduino?

I have a RFID-RC522 (MF-RC522) module and I'm using Arduino sketch program. To use this RFID, I downloaded the Arduino MFRC522 library. And I run the example code of library. Here is the code. /* * MFRC522 - Library to use ARDUINO …

Read More
Serials on NFC Tags - truly unique? cloneable?

So are NFC tags really UNIQUE from each other, at least in their SERIAL NUMBER? And can we rely on the fact that no 2 NFC tags can have the same serial number? I'm highly skeptical about this as there are (…

Read More
RFID algorithm to get card ID

I receive bytes from an RFID reader when presenting a card, but I'm unable to figure out how to derive the card ID from these bytes. For example, I have a card that has these numbers printed on it: 0007625328 116,23152. I …

Read More