Hi,
My name is David and I'm a Spanish student. Now I try to evaluate a wireless routing protocol to ad-hoc networks (concretely OLSR). I have a real scheme to evaluate some parameters, one is the bandwidth. My scheme have two end-to-end nodes (portables pc) and two intermediate nodes that are two PDA's Sharp Zaurus SL5500S. Like this:
PDA1
/ \
/ \
PC1 PC2
\ /
\ /
PDA2
I use Iperf v1.7.0 to get the bandwidth of communication end-to-end (PC1-to-PC2) and I see a lot of strange messages in captures of ethereal. This messages are MDS and FC (Fibre Channel) and her interval of send are very short (0.1 ms) and they occupy the media access. Seems that this messages do a direction MAC tracking, I don't now why... This messages only appear in the scheme that the useful info that I generate with Iperf (UDP or TCP data) cross the intermediate node PDA (PC1 --> PDAx --> PC2). If I try to use Iperf directly (PC1 --> PC2) this strange messages no appear. I try the same test with other traffic generator and never appear this strange messages. I need to measure the bandwidth of communication and this strange messages use the media access and I have a lot of collisions.
The messages that I see in my wireless interface:
Frame 3085 (1520 bytes on wire, 1520 bytes captured)
Arrival Time: Jul 15, 2004 10:48:30.046797000
Time delta from previous packet: 0.006827000 seconds
Time since reference or first frame: 29.504213000 seconds
Frame Number: 3085
Packet Length: 1520 bytes
Capture Length: 1520 bytes
Ethernet II, Src: 00:80:c8:2d:07:e5, Dst: 00:07:eb:30:fd:b6
Destination: 00:07:eb:30:fd:b6 (Cisco_30:fd:b6)
Source: 00:80:c8:2d:07:e5 (D-Link_2d:07:e5)
Type: Unknown (0x0000)
Trailer: 38393031323334353637383930313233...
Frame check sequence: 0x36373839 (incorrect, should be 0x3bea7501)
MDS Header(SOFi4/Unknown(50))
MDS Header
...0 0011 0000 0000 = Packet Len: 768
.... 0000 0000 10.. = Dst Index: 0x0002
.... ..00 0000 0000 = Src Index: 0x0000
.... 0011 0100 0000 = VSAN: 832
MDS Trailer
EOF: Unknown (50)
CRC: 0x34353637
Fibre Channel
R_CTL: 0x3f(FC-4 Link_Data/0xf)
Dest Addr: 11.e4.68
CS_CTL: 0xc0
Src Addr: a8.01.04
Type: Unknown (0xc0)
F_CTL: 0xa80102 Exchange Responder, Seq Initiator, Exchg First, Seq Last, CS_CTL, Last Data Frame - No Info, ABTS - Abort/MS,
1... .... .... .... .... .... = ExgRpd: Exchange Responder
.0.. .... .... .... .... .... = SeqRec: Seq Initiator
..1. .... .... .... .... .... = ExgFst: Exchg First
...0 .... .... .... .... .... = ExgLst: NOT exchg last
.... 1... .... .... .... .... = SeqLst: Seq Last
.... ..0. .... .... .... .... = Pri: CS_CTL
.... ...0 .... .... .... .... = TSI: NOT transfer seq initiative
.... .... 00.. .... .... .... = LDF: Last Data Frame - No Info (0x000000)
.... .... ..00 .... .... .... = A01: no ack required (0x000000)
.... .... .... ..0. .... .... = RetSeq: NOT retransmitted sequence
.... .... .... .... ..00 .... = AA: ABTS - Cont (0x000000)
.... .... .... .... .... 0... = RelOff: rel offset NOT set
SEQ_ID: 0x80
DF_CTL: 0x01
SEQ_CNT: 5001
OX_ID: 0x05c6
RX_ID: 0x77b8
Parameter: 0x00000884
Data (744 bytes)
How I can eliminate these messages to get a correct measurement? Can you help me?
Best regards, and thanks a lot for your help.
David.