1 (edited by tyson.key 2012-01-16 23:02:01)

Topic: Anticollision, and Multiple Token Support

Hi,

I'm new to working with FeliCa, but I decided to purchase some tags (which I received today) from Japan, and begin researching this rather mysterious, obscure, and seemingly regionally-isolated technology - since I was getting bored with MiFare, and haven't much progress on interfacing with my Maxim ISO/IEC 14443-B sample cards.

I believe that FeliCa's anticollision mechanism is based upon a "timeslot" mechanism, and documented in various public PDFs.

With that mind, I've noticed that LibNFC and the ACS card readers (especially the ACR122U variants) will only report the NFCID2 of the first tag to respond to a poll at 212 kbps (regardless of how tags are stacked, or distance), when testing the nfc-list utility.

For example - with 2 tags on the reader, the following is reported:

tyson@UmBongo:~/nfc-tools-read-only/nfcd$ nfc-list -v
nfc-list uses libnfc 1.5.1 (r1236)
Connected to NFC device: ACS ACR122U 00 00 / ACR122U103 - PN532 v1.6 (0x07)
0 ISO14443A passive target(s) found.

1 Felica (212 kbps) passive target(s) found:
        ID (NFCID2): 01  27  00  5d  1a  05  88  cd  
    Parameter (PAD): 00  f0  00  00  02  06  03  00  
   System Code (SC): 88  b4  

1 Felica (424 kbps) passive target(s) found:
        ID (NFCID2): 01  27  00  5d  1a  05  88  cd  
    Parameter (PAD): 00  f0  00  00  02  06  03  00  
   System Code (SC): 88  b4  

With a single tag held above the reader, the following is reported:

tyson@UmBongo:~/nfc-tools-read-only/nfcd$ nfc-list -v
nfc-list uses libnfc 1.5.1 (r1236)
Connected to NFC device: ACS ACR122U 00 00 / ACR122U103 - PN532 v1.6 (0x07)
0 ISO14443A passive target(s) found.

0 Felica (212 kbps) passive target(s) found.

1 Felica (424 kbps) passive target(s) found:
        ID (NFCID2): 01  27  00  5d  1a  05  8a  cd  
    Parameter (PAD): 00  f0  00  00  02  06  03  00  
   System Code (SC): 88  b4  

Additionally, I've noticed that nfc-list will sometimes claim that no tags are present - when in fact, two FeliCa tags are; and sometimes, only the 212 kbps bitrate is reported for a tag. I've also been able to "drown out" FeliCa tags by stacking ISO/IEC 14443-A and 14443-B tags on top.

Out of curiosity - are these issues related to missing functionality, or bugs in LibNFC? Or are they related to limitations of the (implementation of the) chipset in this family of readers? If the former is true, are there any plans to implement support for this functionality? (Or is it just a case of "whoever's interested enough should implement it"?)

Thanks,

Tyson.