************************
...
963600.713: residual capacity = 5.50754e+07
device 0 (physical layer) consumed 4645 mW-s at
activity 0: 0 mWs and 0sec
activity 1: 0 mWs and 0sec
activity 2: 4645 mWs and 121.98sec
activity 3: 0 mWs and 0sec
activity 4: 0 mWs and 0sec
WARNING: device physical layer total time 121.98 != sim time 963600.713
(may not matter)
963600.713: residual capacity = 5.50737e+07
device 0 (physical layer) consumed 6303.93 mW-s at
activity 0: 0 mWs and 0sec
activity 1: 0 mWs and 0sec
activity 2: 6303.93 mWs and 121.98sec
activity 3: 0 mWs and 0sec
activity 4: 0 mWs and 0sec
WARNING: device physical layer total time 121.98 != sim time 963600.713
(may not matter)
963600.713: residual capacity = 5.50765e+07
device 0 (physical layer) consumed 3525.22 mW-s at
activity 0: 0 mWs and 0sec
activity 1: 0 mWs and 0sec
activity 2: 3525.22 mWs and 121.98sec
activity 3: 0 mWs and 0sec
activity 4: 0 mWs and 0sec
WARNING: device physical layer total time 121.98 != sim time 963600.713
(may not matter)
***********************************************
All of the nodes has different consumption like expected and only for
transmission (i just want to see how long they can send) the power is
drawn.
Regards,
Marius
Hi Jes,
I'm not sure "too big" is exactly the right description. The values are
intended to model real batteries -- if you think there's an error in the
values, please report it!
Using realistic battery capacities means that if it would take a year
for the application to drain the battery, then you need to run a year of
simulated time to see the battery drain. Obviously that's not
practical, so you need to think about what your simulation experiment is
trying to measure.
If you want to show that protocol A is more energy efficient than
protocol B, then you care more about the energy consumed when running
protocol A vs running protocol B than the absolute value of the battery
capacity - set it to a large value and don't care.
On the other hand, if you are interested in protocols that respond to
differences in capacity (e.g. load balancing), then you need to think
about how to model what is a likely distribution of relative capacities
at various points in the evolution of the systems. That is the reason
for supporting both capacity and nominal capacity in the interface: You
may decide that manufacturing variance is modeled as uniformly
distributed around +/- 0% of the nominal capacty, for example. Or you
might model energy capacity after some time as normally distributed. Or
you might want to model the effect of a particularly heavily loaded node
(possibly due to exogenous effect such as an attacker) on the network.
Similarly, if you want to investigate the lifetime of a deployed system
(i.e. will this system run for two years?), then you probably need to
use a different kind of model.
Regards.
Laura
However, the energy consumption values below suggest that there is some
issue how you are recording the energy consumption. First, only one
activity is being recorded - usually a radio interface has at least two,
send and receive. Second, the duty cycle seems implausibly low - the
radio seems to have been active for ~100s out of ~1M seconds of
simulation time or a duty cycle of .01%, which seems very strange.
Third, the energy consumed seems plausible for the reported values (35 -
50 mW imples a current of 10-20 mA @ 3V), but it's unusual that each
node reports using a different current for the same activity over the
same time -- definitely not impossible, but you should perhaps check
that this is what you intended.
Regards,
Laura
Hello everyone, I implemented a routing protocol for WSN using Mixim and I want to know the energy consumption of each node, but after simulation i foud that all the nodes consume the same way. A node witch participate on the routing and the other no, a cluster Head node and a simple node also even the sink it consumes the same value as other nodes , Does any one know what's the problem?I joind a screenshot of my results.Thanks in advance