CRAWDAD sunysb/multi_channel

Data set consisting of measurements from two different wireless mesh network testbeds (802.11g and 802.11a).

We conduct measurement using two mesh network testbeds in two different frequency bands - 802.11g in 2.4GHz band and 802.11a in 5GHz band.

date/time of measurement start: 2008-04-26 

date/time of measurement end: 2008-05-06 

collection environment: Multi-channel multi-radio architectures have been widely studied for 802.11-based wireless mesh networks to address the capacity problem due to wireless interference. They all utilize channel assignment algorithms that assume all channels and radio interfaces to be homogeneous. However, in practice, different channels exhibit different link qualities depending on the propagation environment for the same link. Different interfaces on the same node also exhibit link quality variations due to hardware differences and required antenna separations. To study these variations, we conduct measurement using two mesh network testbeds in two different frequency bands 802.11g in 2.4GHz band and 802.11a in 5GHz band. 

network configuration: The 802.11a testbed consists of 13 nodes each of which is a Soekris net4801 single board computer (SBC). The PCI-slot in the SBC is expanded into 4 miniPCI slots using a PCI-to-miniPCI adapter. Four 802.11a/b/g miniPCI wireless cards based on Atheros chipset with external antennas are used in each mesh node. The transmit powers are fixed to 15 dBm and data rate to 6 Mbps. The 802.11g testbed uses 10 Dell latitude D510 laptops each with one Atheros chipset based D-link DWL AG660 PCMCIA 802.11a/b/g card with an internal antenna. The transmit powers are fixed to 15 dBm and data rate to 11 Mbps. 

data collection methodology: Measurements from the 802.11g testbed were collected on 54 different links on three orthogonal channels 1, 6, 11 (2412, 2437 and 2462 MHz respectively) in the 802.11g band. Measurements from the 802.11a testbed were collected on 78 different links in 13 orthogonal channels (between 5180-5825 Mhz) in the 802.11a band. We used standard linux tools such as iperf to send UDP packets on the sender node for each link measured and tcpdump on the receiver node running on a raw monitoring interface to capture the packets. 

We conduct measurement using two mesh network testbeds in two different frequency bands at 802.11g in 2.4GHz band and 802.11a in 5GHz band.

Traceset

sunysb/multi_channel/link

Trace set consisting of measurements from two different wireless mesh network testbeds (802.11g and 802.11a).

• file: multi_channel_data_set.tar.gz
• description: We conduct measurement using two mesh network testbeds in two differentfrequency bands 802.11g in 2.4GHz band and 802.11a in 5GHz band.
• measurement purpose: Network Performance Analysis
• methodology: The measurements are from two different wireless mesh network testbeds (802.11g and 802.11a) set up in our departmental building as described below. 

All nodes in both the testbeds run Linux (kernel 2.6.22 in laptops and kernel 

2.4.29 in the Soekris boxes) and the widely used madwifi device driver (version 

v0.9.4) for the 802.11 interfaces. 

We used standard linux tools such as iperf to send UDP packets on the sender node 

for each link measured and tcpdump on the receiver node running on a raw monitoring 

interface to capture the packets. 

This gives us the additional prism monitoring header information such as the 

received signal strength (RSS), noise, channel and data rate for every received 

packet. 

1. 802.11g

The 802.11g testbed uses 10 Dell latitude D510 laptops each with one Atheros 

chipset based D-link DWL AG660 PCMCIA 802.11a/b/g card with an internal antenna. 

The transmit powers are fixed to 15 dBm and data rate to 11 Mbps. Measurements 

from this testbed were collected on 40 different links on three orthogonal 

channels 1, 6, 11 (2412, 2437 and 2462 MHz respectively) in the 802.11g band. 

2. 802.11a

The 802.11a testbed consists of 13 nodes each of which is a Soekris net4801 

single board computer (SBC). The PCI-slot in the SBC is expanded into 4 miniPCI 

slots using a PCI-to-miniPCI adapter. Four 802.11a/b/g miniPCI wireless cards 

based on Atheros chipset with external antennas are used in each mesh node. 

In order to overcome radio leakage problems, we physically separated the external 

antennas at a distance of about 0.5 meters based on measurements. Otherwise, 

there was a perceptible interference even among orthogonal channels across interfaces 

on the same node. 

Even with this setup, we could use only a subset of orthogonal channels without 

interference. These are 7 channels (channels 36, 44, 52, 60, 149, 157, 165) 

out of possible 13 orthogonal channels. 

The transmit powers are fixed to 15 dBm and data rate to 6 Mbps. Measurements 

from this testbed were collected on 78 different links in 13 orthogonal channels 

(between 5180-5825 Mhz) in the 802.11a band. 

Note that the 802.11a testbed is relatively free from external interference 

as there are no other networks operating in this band in the building. However, 

there are indeed several 802.11g networks in our building. Their influence is 

impossible to eliminate. We, however, did our experiments in this network 

during late night and early morning when other active 802.11g clients are unlikely. 

3. interface

For a given link between two multi-radio nodes, the choice of actual 

radio interfaces to use for this link could impact the link performance. 

To understand the variations caused by interface selection, we study 20 links 

(a subset of the 78 links studied before) in our 802.11a testbed using 16 possible 

interface pairs for each link. We select the same channel (channel 64, one of the 

good performing channels) for this measurement on all links in order to isolate 

the effect of interface selection.

sunysb/multi_channel/link Traces

• 802.11a: Traces of measurements from an 802.11a wireless mesh network testbed.
• configuration: This directory consist of measurements on 78 links each on 13 orthogonal channels
  in 802.11a band. In total, there are 1014 files. 
• format: DATAFORMAT of all files:
  All the data files are in csv format. The records in each file are of the following format

MAC_TIME_STAMP      PACKET_TYPE     CHANNEL     SNR     RSS     NOISE     DATA_RATE     MAC_SEQUENCE_NUMBERS

• 802.11g: Traces of measurements from an 802.11g wireless mesh network testbed.
• configuration: This directory consist of measurements on 54 links each on 3 orthogonal channels in 802.11g band. In total, there are 162 files.
• format: DATAFORMAT of all files:
  All the data files are in csv format. The records in each file are of the following format

MAC_TIME_STAMP      PACKET_TYPE     CHANNEL     SNR     RSS     NOISE     DATA_RATE     MAC_SEQUENCE_NUMBERS

• interface: Traces of measurements from 20 links of interface pairs between multi-radio nodes an 802.11a wireless mesh network testbed.
• configuration: This directory consist of measurements on 20 links working on channel 64 in the 802.11 band. For each link, measurements were done in 16 possible interface pairs. Note that our multi-radio node has 4 radios. So for any link there are 16 possible interface pairs. In total, there are 320 files.
• format: DATAFORMAT of all files:
  All the data files are in csv format. The records in each file are of the following format

MAC_TIME_STAMP      PACKET_TYPE     CHANNEL     SNR     RSS     NOISE     DATA_RATE     MAC_SEQUENCE_NUMBERS