[Dataset] isti/rural (v. 2008-05-06)

top

Five traces (1-marine, 2-marine, 3-navpol, 4-navzin, and 5-navpol) have been added to the traceset isti/rural/navacchio.

The changed components are as follows:

@MISC{isti-rural-2008-05-06,
  author = {Paolo Barsocchi and Gabriele Oligeri and Francesco Potortì},
  title = {{CRAWDAD} data set isti/rural (v. 2008-05-06)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/isti/rural},
  month = may,  
  year = 2008
}
					

We conducted a series of measurements for relating transmission distance 
and packet loss on a Wi-Fi network in rural areas to propose a model 
that relates distance with packet loss probability.

Commonly used frame loss models for simulations over Wi-Fi channels 
assume a simple double regression model with threshold. This model 
is widely accepted, but few measurements are available in the literature 
that try to validate it. As far as we know, none of them is based 
on field trials at the frame level. We conducted a series of 
measurements for relating transmission distance and packet loss 
on a Wi-Fi network in rural areas to propose a model that relates 
distance with packet loss probability.

We performed our outdoor rural measurement campaign using two IBM 
Thinkpad R40e laptops (Celeron 2 GHz with 256 MB ram running Debian 
Linux with a 2.6.8 kernel), equipped with CNet CNWLC-811 IEEE 802.11b 
PCMCIA wireless cards and standard drivers. The cards were put in ad
hoc mode, so that it was not necessary to depend on an access point, 
and no management overhead was present except for the periodic beacon.

The rural environment was a wide uncultivated field with an unobstructed 
line of sight, far from buildings, cell phone antennas and power lines.

We wrote Vbrsr, a pair of programs for sending and receiving frames 
with the aim of collecting statistics about frame errors and power levels, 
which is released with a free software copyright license and is available 
for download at http://wnlab.isti.cnr.it/paolo/measurements/Software.html.

[Traceset] isti/rural/packet_loss (v. 2008-05-06)

top

Five traces (1-marine, 2-marine, 3-navpol, 4-navzin, and 5-navpol) have been added to the traceset isti/rural/navacchio.

The changed components are as follows:

@MISC{isti-rural-packet_loss-2008-05-06,
  author = {Paolo Barsocchi and Gabriele Oligeri and Francesco Potortì},
  title = {{CRAWDAD} trace set isti/rural/packet_loss (v. 2008-05-06)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/isti/rural/packet_loss},
  month = may,  
  year = 2008
}
					

We conducted a series of measurements for relating transmission distance 
and packet loss on a Wi-Fi network in rural areas.

We disabled fragmentation, RTS/CTS, retransmissions (ARQ) and 
dynamic rate switching. We used different fixed speeds of 1, 2, 5.5 and 
11 Mb/s, with three fixed frame lengths (500, 1000, 1500 bytes), 
for different transmitter-receiver distances.

By disabling ARQ, the MAC layer transmits each packet only once, 
rather than trying to retransmit a frame up to 8 times after a loss. 
This means that we sampled the channel at a constant rate of 
200 frames per second, thus accurately measuring the frame error 
process in the time domain, using 200,000 frames for each measure. 
Notice that the procedure described makes the measurement process 
independent of the MAC protocol, and dependent only on the channel 
and the used hardware. 

The rural environment was a wide uncultivated field with an unobstructed 
line of sight, far from buildings, cell phone antennas and power lines.

We wrote Vbrsr, a pair of programs for sending and receiving frames 
with the aim of collecting statistics about frame errors and power levels, 
which is released with a free software copyright license and is available 
for download at http://wnlab.isti.cnr.it/paolo/measurements/Software.html.

[Trace] isti/rural/packet_loss/navacchio (v. 2007-12-19)

top

the initial version

@MISC{isti-rural-packet_loss-navacchio-2007-12-19,
  author = {Paolo Barsocchi and Gabriele Oligeri and Francesco Potortì},
  title = {{CRAWDAD} trace isti/rural/packet_loss/navacchio (v. 2007-12-19)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/isti/rural/packet_loss/navacchio},
  month = dec,  
  year = 2007
}
					

These traces are the data measured in Navacchio (Pisa) in April, 2006 
using two laptops in 802.11 ad hoc mode at different distance on a wide
uncultivated field.

These are the data measured in Navacchio (Pisa) in April, 2006 using two
laptops in 802.11 ad hoc mode at different distance on a wide
uncultivated field.  The laptops are two IBM Thinkpad R40e laptops
(Celeron 2 GHz with 256 MB ram running Debian Linux with a 2.6.8
kernel), equipped with CNet CNWLC-811 wireless cards and standard
drivers.

Retransmission, RTS/CTS and fragmentation are all disabled, long
preambles are used.  Frames are sent to the network card with a 5ms
intertransmission time for 11, 5.5 and 2Mb/s speeds, 10ms for the 1Mb/s
speed.  Notice that frames with data length 1500B are longer than 5ms at
2Mb/s and longer than 10ms at 1MB/s: in these cases the network card
transmits the packets back-to-back.

Each file name is built like this: rcv_SPEED_DISTANCE_LENGTH

where SPEED is the transmission speed (1, 2, 5.5, 11 Mb/s),
DISTANCE is the distance between the transmitter and the receiver
and LENGTH is the frame length (500, 1000, 1500 bytes).

The fields marked with an asterisk (*) below are always 0, because the
network card used for the measurements in this directory cannot give
these statistics.

Each line contains these fields:
 - receive time, or 0 when packet is not received at all
 - lenght of the frame in bytes to data: to obtain the whole frame
   lenght one must add an LLC/SNAP header, CRC and MAC header. There are
   three data lengths: 500, 1000 and 1500 bytes.  The respective frame
   durations can be computed as:

		+-------+-------+-------+-------+-------+
		|len[ms]| 1 Mb/s| 2 Mb/s|5.5Mb/s|11 Mb/s|
		+-------+-------+-------+-------+-------+
		| 500 B | 5.154 | 2.954 | 1.555 | 1.154 |
		+-------+-------+-------+-------+-------+
		|1000 B | 9.154 | 4.954 | 2.282 | 1.518 |
		+-------+-------+-------+-------+-------+
		|1500 B |13.154 | 6.954 | 3.009 | 1.881 |
		+-------+-------+-------+-------+-------+

 - sequence number starting from 0
*- quality level read from the network card
 - signal level read from the network card: multiply by 0.6 to get dB
*- noise level read from the network card
 - status: 1 for lost frames, 0 for received frames
*- number of corrupted bits in frames received with wrong CRC

[Trace] isti/rural/packet_loss/aggregated (v. 2007-12-19)

top

the initial version

@MISC{isti-rural-packet_loss-aggregated-2007-12-19,
  author = {Paolo Barsocchi and Gabriele Oligeri and Francesco Potortì},
  title = {{CRAWDAD} trace isti/rural/packet_loss/aggregated (v. 2007-12-19)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/isti/rural/packet_loss/aggregated},
  month = dec,  
  year = 2007
}
					

These traces are aggregated statistics on data measured in Navacchio at
the beginning of 2006 using two laptops in 802.11 ad hoc mode at
different distances on a wide uncultivated field.

These data are aggregated statistics on data measured in Navacchio at
the beginning of 2006 using two laptops in 802.11 ad hoc mode at
different distances on a wide uncultivated field.

Each file contains data relative to a single transmission speed: 1, 2,
5.5, 11 Mb/s.  Retransmission, RTS/CTS and fragmentation are all
disabled, long preambles are used.

A line in a file is relative to 1000 consecutive frames sent with 
a 5ms intertransmission time for the three top speeds, 10ms for the lowest, 
so each line is relative to a 5s or 10s measurement, depending on speed.  

Each line contains these fields:
 - sequence number starting from 0 relative to a ~15min measurement
 - lenght of the frame in bytes: to obtain the whole frame lenght one
   must add an LLC/SNAP header, CRC and MAC header.  There are three
   lengths: 500, 1000 and 1500 bytes.  The frame lengths in ms are:

		+-------+-------+-------+-------+-------+
		|len[ms]| 1 Mb/s| 2 Mb/s|5.5Mb/s|11 Mb/s|
		+-------+-------+-------+-------+-------+
		| 500 B | 5.154 | 2.954 | 1.555 | 1.154 |
		+-------+-------+-------+-------+-------+
		|1000 B | 9.154 | 4.954 | 2.282 | 1.518 |
		+-------+-------+-------+-------+-------+
		|1500 B |13.154 | 6.954 | 3.009 | 1.881 |
		+-------+-------+-------+-------+-------+

   notice that the cases of lenght 1500 B exceed 5ms at 2Mb/s and exceed
   10ms at 1MB/s: in these two cases the packets have been sent with the
   smallest possible intertranmission interval.
 - signal level presented by the network card: multiply by 0.6 to get dB
 - frame error rate

[Trace] isti/rural/packet_loss/1-marine (v. 2008-05-06)

top

the initial version

@MISC{isti-rural-packet_loss-1-marine-2008-05-06,
  author = {Paolo Barsocchi and Gabriele Oligeri and Francesco Potortì},
  title = {{CRAWDAD} trace isti/rural/packet_loss/1-marine (v. 2008-05-06)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/isti/rural/packet_loss/1-marine},
  month = may,  
  year = 2008
}
					

These traces are the data measured in Marinella, Navacchio (Pisa) in April, 2006 
using two laptops in 802.11 ad hoc mode at different distance on a wide
uncultivated field.

These are the data measured in Marinella, Navacchio (Pisa) in April, 2006 
using two laptops in 802.11 ad hoc mode at different distance on a wide
uncultivated field.  

The laptops are two IBM Thinkpad R40e laptops (Celeron 2 GHz with 
256 MB ram running Debian Linux with a 2.6.8 kernel), equipped with 
CNet CNWLC-811 wireless cards and standard drivers.

Retransmission, RTS/CTS and fragmentation are all disabled, long
preambles are used.  Frames are sent to the network card with a
intertransmission time varying from 5 to 20 ms for 11, 5.5, 2 and 1 Mb/s
speeds (be sure to check the transmission times).  The beacon is sent
every 102.4ms.

Each file name is built like this: rcv_SPEED_DISTANCE_LENGTH

where SPEED is the transmission speed (1, 2, 5.5, 11 Mb/s),
DISTANCE is the distance between the transmitter and the receiver
and LENGTH is the frame length (500, 1000, 1500 bytes).

The fields marked with an asterisk (*) below are always 0, because the
network card used for the measurements in this directory cannot give
these statistics.

Each line contains these fields:
 - receive time, or 0 when packet is not received at all
 - lenght of the frame in bytes to data: to obtain the whole frame
   lenght one must add an LLC/SNAP header, CRC and MAC header. There are
   three data lengths: 500, 1000 and 1500 bytes.  The respective frame
   durations can be computed as:

		+-------+-------+-------+-------+-------+
		|len[ms]| 1 Mb/s| 2 Mb/s|5.5Mb/s|11 Mb/s|
		+-------+-------+-------+-------+-------+
		| 500 B | 5.154 | 2.954 | 1.555 | 1.154 |
		+-------+-------+-------+-------+-------+
		|1000 B | 9.154 | 4.954 | 2.282 | 1.518 |
		+-------+-------+-------+-------+-------+
		|1500 B |13.154 | 6.954 | 3.009 | 1.881 |
		+-------+-------+-------+-------+-------+
		
 - sequence number starting from 0
*- quality level read from the network card
 - signal level read from the network card: multiply by 0.6 to get dB
*- noise level read from the network card
 - status: 1 for lost frames, 0 for received frames
*- number of corrupted bits in frames received with wrong CRC

[Trace] isti/rural/packet_loss/2-marine (v. 2008-05-06)

top

the initial version

@MISC{isti-rural-packet_loss-2-marine-2008-05-06,
  author = {Paolo Barsocchi and Gabriele Oligeri and Francesco Potortì},
  title = {{CRAWDAD} trace isti/rural/packet_loss/2-marine (v. 2008-05-06)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/isti/rural/packet_loss/2-marine},
  month = may,  
  year = 2008
}
					

These traces are the data measured in Marinella, Navacchio (Pisa) in April, 2006 
using two laptops in 802.11 ad hoc mode at different distance on a wide
uncultivated field.

These are the data measured in Marinella, Navacchio (Pisa) in April, 2006 
using two laptops in 802.11 ad hoc mode at different distance on a wide
uncultivated field.  

The laptops are two IBM Thinkpad R40e laptops (Celeron 2 GHz with 
256 MB ram running Debian Linux with a 2.6.8 kernel), equipped with 
CNet CNWLC-811 wireless cards and standard drivers.

Retransmission, RTS/CTS and fragmentation are all disabled, long
preambles are used.  Frames are sent to the network card with a
intertransmission time varying from 5 to 20 ms for 11, 5.5, 2 and 1 Mb/s
speeds (be sure to check the transmission times).  The beacon is sent
every 102.4ms.

Each file name is built like this: rcv_SPEED_DISTANCE_LENGTH

where SPEED is the transmission speed (1, 2, 5.5, 11 Mb/s),
DISTANCE is the distance between the transmitter and the receiver
and LENGTH is the frame length (500, 1000, 1500 bytes).

The fields marked with an asterisk (*) below are always 0, because the
network card used for the measurements in this directory cannot give
these statistics.

Each line contains these fields:
 - receive time, or 0 when packet is not received at all
 - lenght of the frame in bytes to data: to obtain the whole frame
   lenght one must add an LLC/SNAP header, CRC and MAC header. There are
   three data lengths: 500, 1000 and 1500 bytes.  The respective frame
   durations can be computed as:

		+-------+-------+-------+-------+-------+
		|len[ms]| 1 Mb/s| 2 Mb/s|5.5Mb/s|11 Mb/s|
		+-------+-------+-------+-------+-------+
		| 500 B | 5.154 | 2.954 | 1.555 | 1.154 |
		+-------+-------+-------+-------+-------+
		|1000 B | 9.154 | 4.954 | 2.282 | 1.518 |
		+-------+-------+-------+-------+-------+
		|1500 B |13.154 | 6.954 | 3.009 | 1.881 |
		+-------+-------+-------+-------+-------+
		
 - sequence number starting from 0
*- quality level read from the network card
 - signal level read from the network card: multiply by 0.6 to get dB
*- noise level read from the network card
 - status: 1 for lost frames, 0 for received frames
*- number of corrupted bits in frames received with wrong CRC

[Trace] isti/rural/packet_loss/3-navpol (v. 2008-05-06)

top

the initial version

@MISC{isti-rural-packet_loss-3-navpol-2008-05-06,
  author = {Paolo Barsocchi and Gabriele Oligeri and Francesco Potortì},
  title = {{CRAWDAD} trace isti/rural/packet_loss/3-navpol (v. 2008-05-06)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/isti/rural/packet_loss/3-navpol},
  month = may,  
  year = 2008
}
					

These traces are the data measured in poligono di Navacchio, Navacchio (Pisa) in April, 2006 
using two laptops in 802.11 ad hoc mode at different distance on a wide
uncultivated field.

These are the data measured in poligono di Navacchio (poligono means 
shooting range), Navacchio (Pisa) in April, 2006 using two laptops 
in 802.11 ad hoc mode at different distance on a wide uncultivated field.  

The laptops are two IBM Thinkpad R40e laptops (Celeron 2 GHz with 
256 MB ram running Debian Linux with a 2.6.8 kernel), equipped with 
CNet CNWLC-811 wireless cards and standard drivers.

Retransmission, RTS/CTS and fragmentation are all disabled, long
preambles are used.  Frames are sent to the network card with a
intertransmission time varying from 5 to 20 ms for 11, 5.5, 2 and 1 Mb/s
speeds (be sure to check the transmission times).  The beacon is sent
every 102.4ms.

Each file name is built like this: rcv_SPEED_DISTANCE_LENGTH

where SPEED is the transmission speed (1, 2, 5.5, 11 Mb/s),
DISTANCE is the distance between the transmitter and the receiver
and LENGTH is the frame length (500, 1000, 1500 bytes).

The fields marked with an asterisk (*) below are always 0, because the
network card used for the measurements in this directory cannot give
these statistics.

Each line contains these fields:
 - receive time, or 0 when packet is not received at all
 - lenght of the frame in bytes to data: to obtain the whole frame
   lenght one must add an LLC/SNAP header, CRC and MAC header. There are
   three data lengths: 500, 1000 and 1500 bytes.  The respective frame
   durations can be computed as:

		+-------+-------+-------+-------+-------+
		|len[ms]| 1 Mb/s| 2 Mb/s|5.5Mb/s|11 Mb/s|
		+-------+-------+-------+-------+-------+
		| 500 B | 5.154 | 2.954 | 1.555 | 1.154 |
		+-------+-------+-------+-------+-------+
		|1000 B | 9.154 | 4.954 | 2.282 | 1.518 |
		+-------+-------+-------+-------+-------+
		|1500 B |13.154 | 6.954 | 3.009 | 1.881 |
		+-------+-------+-------+-------+-------+
		
 - sequence number starting from 0
*- quality level read from the network card
 - signal level read from the network card: multiply by 0.6 to get dB
*- noise level read from the network card
 - status: 1 for lost frames, 0 for received frames
*- number of corrupted bits in frames received with wrong CRC

[Trace] isti/rural/packet_loss/4-navzin (v. 2008-05-06)

top

the initial version

@MISC{isti-rural-packet_loss-4-navzin-2008-05-06,
  author = {Paolo Barsocchi and Gabriele Oligeri and Francesco Potortì},
  title = {{CRAWDAD} trace isti/rural/packet_loss/4-navzin (v. 2008-05-06)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/isti/rural/packet_loss/4-navzin},
  month = may,  
  year = 2008
}
					

These traces are the data measured in campo zingari di Navacchio, 
Navacchio (Pisa) in April, 2006 using two laptops in 802.11 ad hoc 
mode at different distance on a wide uncultivated field.

These are the data measured in campo zingari di Navacchio (campo
zingari means gipsy field), Navacchio (Pisa) in April, 2006 using two laptops 
in 802.11 ad hoc mode at different distance on a wide uncultivated field.  

The names
come form the name of the locations: Marinella, poligono di Navacchio
(poligono means shooting range), campo zingari di Navacchio (campo
zingari means gipsy field).

The laptops are two IBM Thinkpad R40e laptops (Celeron 2 GHz with 
256 MB ram running Debian Linux with a 2.6.8 kernel), equipped with 
CNet CNWLC-811 wireless cards and standard drivers.

Retransmission, RTS/CTS and fragmentation are all disabled, long
preambles are used.  Frames are sent to the network card with a
intertransmission time varying from 5 to 20 ms for 11, 5.5, 2 and 1 Mb/s
speeds (be sure to check the transmission times).  The beacon is sent
every 102.4ms.

Each file name is built like this: rcv_SPEED_DISTANCE_LENGTH

where SPEED is the transmission speed (1, 2, 5.5, 11 Mb/s),
DISTANCE is the distance between the transmitter and the receiver
and LENGTH is the frame length (500, 1000, 1500 bytes).

The fields marked with an asterisk (*) below are always 0, because the
network card used for the measurements in this directory cannot give
these statistics.

Each line contains these fields:
 - receive time, or 0 when packet is not received at all
 - lenght of the frame in bytes to data: to obtain the whole frame
   lenght one must add an LLC/SNAP header, CRC and MAC header. There are
   three data lengths: 500, 1000 and 1500 bytes.  The respective frame
   durations can be computed as:

		+-------+-------+-------+-------+-------+
		|len[ms]| 1 Mb/s| 2 Mb/s|5.5Mb/s|11 Mb/s|
		+-------+-------+-------+-------+-------+
		| 500 B | 5.154 | 2.954 | 1.555 | 1.154 |
		+-------+-------+-------+-------+-------+
		|1000 B | 9.154 | 4.954 | 2.282 | 1.518 |
		+-------+-------+-------+-------+-------+
		|1500 B |13.154 | 6.954 | 3.009 | 1.881 |
		+-------+-------+-------+-------+-------+
		
 - sequence number starting from 0
*- quality level read from the network card
 - signal level read from the network card: multiply by 0.6 to get dB
*- noise level read from the network card
 - status: 1 for lost frames, 0 for received frames
*- number of corrupted bits in frames received with wrong CRC

[Trace] isti/rural/packet_loss/5-navpol (v. 2008-05-06)

top

the initial version

@MISC{isti-rural-packet_loss-5-navpol-2008-05-06,
  author = {Paolo Barsocchi and Gabriele Oligeri and Francesco Potortì},
  title = {{CRAWDAD} trace isti/rural/packet_loss/5-navpol (v. 2008-05-06)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/isti/rural/packet_loss/5-navpol},
  month = may,  
  year = 2008
}
					

These traces are the data measured in poligono di Navacchio, Navacchio (Pisa) 
in April, 2006 using two laptops in 802.11 ad hoc mode at different 
distance on a wide uncultivated field.

These are the data measured in poligono di Navacchio (poligono means 
shooting range), Navacchio (Pisa) in April, 2006 using two laptops 
in 802.11 ad hoc mode at different distance on a wide uncultivated field.  

The laptops are two IBM Thinkpad R40e laptops (Celeron 2 GHz with 
256 MB ram running Debian Linux with a 2.6.8 kernel), equipped with 
CNet CNWLC-811 wireless cards and standard drivers.

Retransmission, RTS/CTS and fragmentation are all disabled, long
preambles are used.  Frames are sent to the network card with a
intertransmission time varying from 5 to 20 ms for 11, 5.5, 2 and 1 Mb/s
speeds (be sure to check the transmission times).  The beacon is sent
every 102.4ms.

Each file name is built like this: rcv_SPEED_DISTANCE_LENGTH

where SPEED is the transmission speed (1, 2, 5.5, 11 Mb/s),
DISTANCE is the distance between the transmitter and the receiver
and LENGTH is the frame length (500, 1000, 1500 bytes).

The fields marked with an asterisk (*) below are always 0, because the
network card used for the measurements in this directory cannot give
these statistics.

Each line contains these fields:
 - receive time, or 0 when packet is not received at all
 - lenght of the frame in bytes to data: to obtain the whole frame
   lenght one must add an LLC/SNAP header, CRC and MAC header. There are
   three data lengths: 500, 1000 and 1500 bytes.  The respective frame
   durations can be computed as:

		+-------+-------+-------+-------+-------+
		|len[ms]| 1 Mb/s| 2 Mb/s|5.5Mb/s|11 Mb/s|
		+-------+-------+-------+-------+-------+
		| 500 B | 5.154 | 2.954 | 1.555 | 1.154 |
		+-------+-------+-------+-------+-------+
		|1000 B | 9.154 | 4.954 | 2.282 | 1.518 |
		+-------+-------+-------+-------+-------+
		|1500 B |13.154 | 6.954 | 3.009 | 1.881 |
		+-------+-------+-------+-------+-------+
		
 - sequence number starting from 0
*- quality level read from the network card
 - signal level read from the network card: multiply by 0.6 to get dB
*- noise level read from the network card
 - status: 1 for lost frames, 0 for received frames
*- number of corrupted bits in frames received with wrong CRC

[Author] Paolo Barsocchi

top

[Author] Gabriele Oligeri

top

[Author] Francesco Potortì

top

[Paper] barsocchi-rural_wifi

top

Commonly used frame loss models for simulations over Wi-Fi channels assume a 
simple double regression model with threshold. This model is widely accepted, 
but few measurements are available in the literature that try to validate it. 
As far as we know, none of them is based on field trials at the frame level. We 
present a series of measurements for relating transmission distance and packet 
loss on a Wi-Fi network in rural areas and propose a model that relates 
distance with packet loss probability. We show that a simple double regression 
propagation model like the one used in the ns-2 simulator can miss important 
transmission impairments that are apparent even at short transmitter-receiver 
distances. Measurements also show that packet loss at the frame level is a 
Bernoullian process for time spans of few seconds. We relate the packet loss 
probability to the received signal level using standard models for additive 
white Gaussian noise channels. The resulting model is much more similar to the 
measured channels than the simple models where all packets are received when 
the distance is below a given threshold and all are lost when the threshold is 
exceeded.