915 MHz MICA2 range report
Ken
I've posted a short report on my experiment on outdoor transmission
range of 915 MHz MICA2 motes. You can find it at http://cytheric.net/masters/915range.pdf.
I'm pretty happy with the methodology and the results. I wish the
networking/software handled more than one mote at a time though (read:
collisions). Maybe version 2. I also wish the lacrosse team would
start practice at my test site later than 8am on the weekends. ;-)
I also tested 433 MHz motes. At 240 ft the signal strength was -77,
-74, and -69 for 0 dBm, 5 dBm, and 10 dBm transmit power respectively.
That's when I ran out of space. I plan to retest using the full extent
of the field to find the limit. I'll update the report when I get
those results.
Ken
Gianni
your experiment is very interesting. I was just going to post the
results of a similar experiment I did using TelosB boards. Placing the
nodes at 1.3m (4.26ft) above the ground I measured a maximum
transmission range of about 140 meters (470 ft). It is quite
surprising how the MICA2 boards, which operate at much lower frequency
(915MHz vs. 2.4 GHz) and have higher transmission power (+5dBm vs
0dBm), have a max transmission range that is only one third of TelosB.
Anyway your measurements confirm my own experiences working with MICA2
and MICAZ motes: although I have never done accurate experiments, I've
always suspected that the max transmission range was much shorter than
what claimed in the datasheets.
I'm going to post my results soon - after that we can discuss more
about this.
Gianni
On Feb 20, 9:19 pm, "Ken" <r...@deru.com> wrote:
> Hey everyone,
>
> I've posted a short report on my experiment on outdoor transmission
> range of 915 MHz MICA2 motes. You can find it athttp://cytheric.net/masters/915range.pdf.
Ken
in your TelosB experiments. I'm planning to install a sensor network
at a botanical garden. The distances may be as much as 35m with
foliage and uneven ground, so I want to be sure I understand
propagation well enough to make good recommendations on the motes.
With 802.15.4, 2.4 GHz has become popular. So if I need to learn if
that would work for me. Have you written a paper about your
experiment, or was this test for another purpose?
I may try to replicate your results. The lab has a couple of TelosB
available for the experiment. I'll let you know how it turns out.
I'm also interested in your directional antennas. We actually met
about a year ago, and you were showing the copper cubes at the lab. I
guess this is what you used in the experiment. Are these antennas your
creation? Are they expensive to make? I've looked at small Yagis but
they're around $50 US, which is expensive relative to the cost of the
mote itself. So an inexpensive solution that gives me single hop and
saves transmission power is worth considering.
My background is computer applications, so RF has been a major
learning experience. It's a lot of fun though.
Ken
Gianni
yes, I used the "copper cubes" for the experiments. My background is
also mostly computer software... I wouldn't know how to build the
antenna myself, but from what I see in my lab they are very simple and
inexpensive to build (it's just a rectangular patch of copper on a
dielectric substrate). Try to look for "patch antenna", you should be
able to find something inexpensive.
Anyway, if you plan to use directional antennas on some nodes and
omnidirectional antennas on others, beware that not every combination
will work well. I have update the page I posted few days ago with a
new plot showing the effect of using antenna with different
polarization.
Let me know when you plan to do your experiments with TelosB. If you
are interested, I can give you the Java software I used for my
experiments. It has a GUI interface, so it should be easy to use. Also
I was interested to know if you measured some RSSI values without the
cardboard covered with aluminum foil under them. Let me know if you
have any data.
I am working on a chapter of my thesis that describes more in details
the RSSI experiments with the antenna and the TwoRay model. It should
be ready in a few days... I'll send it to you as soon as I complete it.
Gianni
Ken
G. Anastasi, A. Falchi, A. Passarella, M. Conti, E. Gregori. Sensor
networks: Performance measurements of
motes sensor networks. In Proceedings of the 7th ACM international
symposium on Modeling, analysis and simulation of wireless and mobile
systems, pages 174-181. ACM Press, 2004
They report 60-70 m for 915 MHz MICA2 using a ground plane outdoors,
with motes elevated 1 m off the ground. So somewhat greater range than
I found, but still no more than half what Crossbow states.
I should say that I *assume* they use 915 MHz motes because they don't
specifically say so. However, a photo shows the "900 MHz" label, and
they state max power is 5 dBm.
Ken
On Feb 20, 2:44 pm, "Gianni" <giann...@gmail.com> wrote:
Ken
original 1/4 wave antenna. The results are available at
http://cytheric.net/masters/rangetest.pdf. Go to page 4. I also tested
using a wire mesh screen as a primitive reflector.
Also you can learn more about my project itself by browsing to
http://cytheric.net/masters/project.shtml.
I've been reading up on other SNR, noise, performance, etc. reports.
I've linked to several of these at the bottom of
http://cytheric.net/masters/rangetime-test.shtml. People are trying to
figure out what performance to expect from motes and what factors
generate this performance.
I've included some of the older, highly cited papers like Cerpa 2003
and Zhao, Govindan 2003 that first identified the "gray area", where
packet reception rates vary widely. The emerging understanding of this
phenomenon is that it occurs at the edge of a mote's transmission
range. I really like Zuniga, Krishnamachari (2004) because it explains
these issues clearly.
I just came across Srinivasan, et al (2006), which includes a lot of
data on transmission characteristics of TelosB. There's a lot to
digest there. I also think you'll enjoy Raman, et al 2006, which
includes tests on Tmotes with various antennas. On the theoretical
side that paper pokes at Srinivasan, Levis 2006 by saying that RSSI
varies a lot more than they claim. I think it depends on the
environment. Raman's got a lot of trees and so on that will mess with
signal strength. I've certainly seen that at the botanical garden.
Regarding these controversies, I think sensor network people would be
well served to review the research by the cellphone people. They
looked at a lot of the same issues in the 90s. For example there are
some good papers about how wind in trees affects RSSI.
I'll look forward to reviewing your thesis chapter on RSSI.
Ken