Greetings,
I am trying to capture the MACs in the frames. I need to know the time of
arrival of the MACs.
To get to the point, I have wired three signals from the Intersil Prism chip
set to a FPGA and logic analyzer.
The signals are RXD, RXCLK and MD_RDY and I picked them off the HFA3861B
chip. With the analyzer I can
trigger storage on the rising edge of MD_RDY and store RXD. I can see
SERVICE, SIGNAL and LENGTH
data along with the CRC on the analyzer. The data I see in the PLCP header
makes sense and agrees with the way the system is set up.
Its when we get into the PSDU part of the frame things start to not adhere
to the 802.11 spec section 7 frame format definition.
I am trying to locate the MAC of the transmitter in the frame but when I
look at the captured data in the FPGA there is no data that
matches the MAC. I have looked at the bits reversed and tried matching on a
byte by byte. The logic analyzer data agrees with the
data in the FPGA so that validates my FPGA design.
The vendor of the card sometimes uses a protocol of 01 which is reserved
and various subtypes are also used that are reserved.
I haven't seen a type of 11 (reserved) in the frame control.
So my questions are these:
1. Is there some sort of bit interleaving of the mac address being done?
2. Is there some sort of mapping of the mac to some other value being
done?
3. If I restrict my data capture to association request and
authentication will the MAC show up in the frame?
The MAC I speak about is the MAC that is printed on the label on the
equipment and matches the one displayed in the monitor
program via the serial interface,
Thanks for any info.
Regards
Jer
gary - 30 Jan 2004 05:11 GMT
You don't say whether this is an infrastructure or ad-hoc network. I just
reviewed the 802.11 standards, and it looks like all control, management,
and data frames put a MAC address in at least one of the four address
fields. A couple of things come to mind:
1. I believe that address bits are transmitted in least-to-most significant
bit order, since the group and broadcast bits must be processed first to
properly interpret the remaining address bits. I don't know how your
analyzer is set up to display the bits. Do you just have a simple
bit-ordering problem, so that the octets displayed by the analyzer are
reversed compared to the hex values you are looking for? Byte order might
also be reversed. If so, why don't you also have a problem making sense of
the fields in the frame control field (or do you - are you really seeing a
reserved protocol value and weird subtypes, or just parsing the bits in the
wrong order)?
2. If it's an ad-hoc network, the BSSID is a randomly-generated
locally-administered MAC address, and won't match any MAC you're looking
for. However, in data frames address field 1 should contain either the DA
(ad-hoc network) or the BSSID (infrastructure - see table 4). The BSSID will
be the AP's MAC address in an infrastructure network, so presumably address
field 1 contains the MAC address of the station you're monitoring, whichever
kind of network it is.
Could you enclose a couple of frame dumps, and the MAC address you expect to
find?
> Greetings,
> I am trying to capture the MACs in the frames. I need to know the time of
[quoted text clipped - 34 lines]
> Regards
> Jer
gary - 30 Jan 2004 05:23 GMT
Sorry, just reread your post and I realize that you already checked bit
ordering, so I assume you tried the combo of reverse-byte/reverse-bit. It
would still be informative to see a couple of example frame dumps. Does the
data in address field 1 look like any kind of MAC address? Is it possible
that the MAC is being reassigned by a driver or has been reset in an eprom
to something other than what is on the equipment label?
> Greetings,
> I am trying to capture the MACs in the frames. I need to know the time of
[quoted text clipped - 34 lines]
> Regards
> Jer
Jerry - 31 Jan 2004 01:12 GMT
My bad, I was dropping bits every now and then while shifting them into the
FPGA.
After fixing that I now see the MACs of the units on the bit dtream out of
the demod receiver chip.
All is well.
Tks Gary for the response.
> Greetings,
> I am trying to capture the MACs in the frames. I need to know the time of
[quoted text clipped - 34 lines]
> Regards
> Jer
gary - 31 Jan 2004 05:16 GMT
Aha! Bit-dropping can be hazardous to your health.
> My bad, I was dropping bits every now and then while shifting them into the
> FPGA.
[quoted text clipped - 46 lines]
> > Regards
> > Jer
James Knott - 31 Jan 2004 13:03 GMT
> My bad, I was dropping bits every now and then while shifting them into
> the FPGA.
Did you catch them with a bit bucket? ;-)

Signature
Fundamentalism is fundamentally wrong.
To reply to this message, replace everything to the left of "@" with
james.knott.