ucsd/cse2008082549200808252008-09-30ucsd/cseDataset of comprehensive traces of wireless activity in the UCSD Computer Science building.To characterize the sources of delay in 802.11 production network,
we collected comprehensive traces of wireless activity in the UCSD
Computer Science building.the initial version2008-08-252007-01-112007-01-11cheng-cross-layercheng-cross-layer-journalcheng-jigsawData Set Website192http://sysnet.ucsd.edu/wirelesshttp://www.crawdad.org/wiki/pmwiki.php?n=Main.Dataset.ucsd-cse802.11802.11 frames802.11b802.11gRFMONpacket tracetcpdumpNetwork Diagnosis802.11 infrastructureTo characterize the sources of delay in 802.11 production network,
we collected comprehensive traces of wireless activity in the UCSD
Computer Science building. The traces was collected on Thursday,
January 11, 2007.The production 802.11 network consists of 40 Avaya AP-8 802.11 b/g
access points covering four floors and the basement. The APs are
identically confiured (except for their channel assignment) and
support both 802.11b and 802.11g without encryption.
Our CSE wireless network has 40 APs. Their locations are in [labels.txt]
and five .png files - [1st floor], [2nd floor], [3rd floor], [4th floor],
and [basement].We use the Jigsaw system described in [cheng-jigsaw] to collect
the traces. Jigsaw is a distributed wireless monitoring platform
that we have deployed in our department building to monitor the
production 802.11 network.
The hardware monitors consist of 192 radios interspersed between
the infrastructure APs. The radios passively monitor the
wireless network and report all wireless events across location,
channel, and time via a private wired network to a back-end storage
server. Jigsaw merges and time synchronizes these separate radio
traces into a single, global uni ed trace. Moreover, Jigsaw performs
this operation in real time; a single 2.2Ghz AMD Opteron server
can synchronize one minute of raw trace data in under 15 seconds.
We configure Jigsaw to capture the first 120 bytes of each wireless
frame. As a result, the aggregate monitor traffic from all radios
ranges from 2-10Mbps and is roughly five times the amount of production
wireless traffic.The last 3 octest of MAC addresses except 0:0:0 are anonymized but
the OUIs are preserved. All IP addresses in IP header or payload are
anonymized as well except the UCSD wireless subnet prefix
(128.54.42/16, 42.0.0.0/16) and private addresses. We do not preserv
any other IP prefixes. Everything beyond TCP, UDP, and DHCP header
is removed. Also we do not recompute the IP/TCP checksums.Please be careful that the wired packet and the wireles packets are not 1-to-1 match:
* Every wired packet may have many multiple 802.11 retransmissions.
* APs only forward 802.11 data frames. Management, control, NULL frames only exist in 802.11 network.
* The sniffers may pick up non-CSE AP signals. Similarly, the sniffers may miss CSE packets.
* The wired gateway forwards broadcast traffic among other two nearby bulidings wireless VLAN./download/ucsd/cse/1f.png/download/ucsd/cse/2f.png/download/ucsd/cse/3f.png/download/ucsd/cse/4f.png/download/ucsd/cse/bf.png/download/ucsd/cse/labels.txt78200808252008-09-30the initial version.ucsd/cse/jigsawJigsaw traces of wireless activity in the UCSD Computer Science building.We used Jigsaw - a tool for analyzing wireless traffic - to collect comprehensive
traces of wireless activity in the UCSD Computer Science building.2008-08-252007-01-112007-01-11Network Diagnosis1. Software
Jigsaw is a tool for analyzing wireless traffic across locations, channels,
time, and protocol layers. It takes traces from multiple sniffers at distinct
vantage points, identifies and synchronize the duplicate wireless frames
in the traces, rebuild link layer and transport layer conversasions. This version
also includes a madwifi driver patch that reduces the overhead of excessive
logging of PHY and CRC error events. Jigsaw is available under GPL licence.
2. Hardware
The guts of our wireless node/sensor is a Soekris net4801 or net4826 embedded
computer, which has a 266 Mhz 586 class CPU (Geode) single chip processor.
The 4801 board includes one Compact Flash slot, three 10/100 ethernet ports,
128 Megabytes of RAM, serial ports, MiniPCI/PCI slot. In addition, 4801 has
IDE port and two USB 1.1 ports. 4826 can be powered over Ethernet. Most of
our nodes are 4826 boxes.
The Compact Flash slot is loaded with a Compact Flash card (4801 has 256M,
4826 has 64M), used to store the moderately patched Pebble Linux image and
related files. In normal operation the card is mounted in read-only mode
to reduce wear and help ensure filesystem consistency in the face of power outages.
A small portion of the memory is mounted for RW file system access. Each node
is equipped with two Atheros-based 802.11 a/b/g wireless cards. Two NICs enable
a broader range of experiments. The radio is attached to a 5dBi omni-directional
attenna. We use heavily patched versions of the Atheros MadWiFi driver for these radios.
Originally, the 4801 has a 20 Gigabyte (minimum) IDE hard disk. But we found
hard disk failure is the major cause for crashes, so we removed them from 4801 boxes.
Otherwise, the boxes are pretty stable and seldom crashes beside our own
Kernel/drivers bugs. For our traffic monitoring project, all traces are directly
dumped over NFS to one RAID 0 2 TB storage server.
We have done several things to help us test new software and run experiments
more conviently. First we install/re-install the kernels and other software
through a master controller to keep all software synchronized and up-to-date
automatically. It usually takes 1-2 minutes to re-install everything for all boxes.
Since the kernel logs are gone after reboot because they are stored in memory
file systems, we have all kernel logs remotely logged into our master server.
This helps us to perform post-crash analysis or makes system management easier
in general. In cases when the kernel hangs/panics or for some reason we can not
login to perform a manual reboot, we can remotely reboot these boxes
(and instruct them to boot to a stable kernel) in a minute. In addition, we use
Geode CPU watch dog functions to make the boxes reboot themselves after certain
timeout. Thus we minimize manual intervention for software update, experiements,
and debugging.ucsd/cse252200808252008-09-30the initial versionucsd/cse/jigsaw/wirelessJigsaw traces of wireless activity in the UCSD Computer Science building.Jigsaw traces of wireless activity in the UCSD Computer Science building.false2008-08-252007-01-112007-01-11the (merged) jigsaw traces collected using 192 sniffers in UCSD CSE building.The file is a series of jcap_hdr ([jcap_hdr.h]) packets like the pcap_pkthdr packet format.
We created our own header simply to save spaces./download/ucsd/cse/jcap_hdr.h/download/ucsd/cse/wirelessucsd/cse/jigsaw79200808252008-09-30the initial version.ucsd/cse/tcpdumpTcpdump traces of wireless activity in the UCSD Computer Science building.We collected tcpdump traces of wireless activity in the UCSD Computer Science building.2008-08-252007-01-112007-01-11Network DiagnosisThe tcpdump trace was collected at the gateway router that interfaces
the campus giga-ether network and the CSE wireless VLAN.ucsd/cse253200808252008-09-30the initial versionucsd/cse/tcpdump/wiredTcpdump traces of wireless activity in the UCSD Computer Science building.Tcpdump traces of wireless activity in the UCSD Computer Science building.false2008-08-252007-01-112007-01-11the tcpdump trace at the gateway router that interfaces the campus giga-ether network and the CSE wireless VLAN.The format is gzipped tcpdump pcap./download/ucsd/cse/wireducsd/cse/tcpdump192ucsd/cseYu-Chung Chengycheng@cs.ucsd.eduGoogle Inc.http://sysnet.ucsd.edu/~ycheng/cheng-cross-layerYu-Chung ChengMikhail AfanasyevPatrick VerkaikPéter BenköJennifer ChiangAlex C. SnoerenStefan SavageGeoffrey M. VoelkerAutomating cross-layer diagnosis of enterprise wireless networkscrawdaducsd/jigsawmeasurementwirelessSIGCOMM '07: Proceedings of the 2007 conference on Applications, technologies, architectures, and protocols for computer communications2007978-1-59593-713-125-36Kyoto, Japanhttp://doi.acm.org/10.1145/1282380.1282384ACMucsd/cse20070001cheng-cross-layer-journalYu-Chung ChengMikhail AfanasyevPatrick VerkaikPéter BenköJennifer ChiangAlex C. SnoerenStefan SavageGeoffrey M. VoelkerAutomating cross-layer diagnosis of enterprise wireless networkscrawdaducsd/jigsawmeasurementwirelessSIGCOMM Comput. Commun. Rev.37420070146-483325-36http://doi.acm.org/10.1145/1282427.1282384http://doi.acm.org/10.1145/1282427.1282384ACMNew York, NY, USAucsd/cse20070001cheng-jigsawYu-Chung ChengJohn BellardoPéter BenköAlex C. SnoerenGeoffrey M. VoelkerStefan SavageJigsaw: solving the puzzle of enterprise 802.11 analysiscrawdaducsd/jigsawmeasurementwirelessSIGCOMM Comput. Commun. Rev.36420060146-483339-50http://doi.acm.org/10.1145/1151659.1159920http://doi.acm.org/10.1145/1151659.1159920ACMNew York, NY, USAucsd/cse20060001