> If i access the timestamp at receiver i will have only one point of > reference to calculate the time take for the packet to travel from sender > to receiver...wat can i do to get another reference time?? if you run FTSP, you can get both A) global time of the message transmission (at sender) and B) global time of the message reception (at receiver) using TimeSyncAMSend and TimeSyncPacket interfaces. however, these times will most likely be the same, as the radio waves travel at the speed of light...
for more information, have a look at http://www.tinyos.net/tinyos-2.x/doc/html/tep133.html in particular, section 4.1, approach #1 > So i am accessing the local times at the application level at both the > sender and receiver and converting it to global times to take the > difference. is this wrong?? you'll see large, non-deterministic errors due to delays between application layer timestamping and the times when the packet is actually transmitted/received. mostly due to the MAC layers back-offs. have a look at any of the recent timesync papers - they explain in detail why application level timestamping is not good. > > wat do you suggest that i do if i want the end to end delay from the > sender to receiver??? > > Pls help... > > >> you should not read localtime directly in sendDone() or receive() events. >> instead, use the PacketTimeStamp interface to access the timestamp >> directly. >> see apps/tests/TestFTSP for an example how to wire the interface. >> >> the frequency of sync msgs is controlled by TIMESYNC_RATE environmental >> variable. it is already set to 3 seconds in the test application, which >> should synchronize all nodes after 9 seconds... >> >> brano >> >> On 11/05/10 5:02 PM, "AKHILA S RAO" <akhila...@ece.iisc.ernet.in> wrote: >> >>> Okay.. this means wat i am doing is rite.. >>> I am sending pkts from a node every 100 ms and collecting it at the >>> basestation. I am accessing the localtime and converting it to global >>> and >>> this value is sent in the packet. this is done at both sender and >>> receiver. >>> >>> So if the nodes are in sync then i should be able to use the difference >>> in >>> the time values of sender and receiver for some calculation... >>> >>> However, there seems to be some problem and the nodes are not >>> synchronised...Do u know how frequently the sync msgs are exchanged and >>> where it can be modified??? >>> >>> >>> >>>> no, you don't need to start ftsp explicitly. you do, however, need to >>>> wire >>>> the proper SoftwareInit and Boot interfaces in the main application >>>> file. >>>> you can find an example in apps/tests/TestFtsp/Ftsp/TestFtspAppC.nc >>>> >>>> brano >>>> >>>> >>>> On 11/05/10 2:04 PM, "Deeksha Rao Sahib" <eced...@ece.iisc.ernet.in> >>>> wrote: >>>> >>>>> Hi, >>>>> >>>>> I had a similar question... I have already included the directory and >>>>> also >>>>> the TimeSyncC component. Does this ensure that ftsp is running in the >>>>> background of my application code? I do not have to explicitly begin >>>>> or >>>>> activate ftsp rite?? >>>>> >>>>>> Hi Wafa, >>>>>> >>>>>> You should use the ftsp library from the tos/lib/ftsp directory of >>>>>> the >>>>>> tinyos-2.x tree. Look at the TestFTSP application to see how to use >>>>>> it. >>>>>> >>>>>> Best, >>>>>> Miklos >>>>>> >>>>>> On Mon, May 10, 2010 at 7:18 PM, wafa jaballah >>>>>> <wafa.jabal...@gmail.com> >>>>>> wrote: >>>>>>> Hi all, >>>>>>> >>>>>>> My application needs synchronised nodes in order to have correct >>>>>>> results. >>>>>>> To do this, I have to synchronise nodes with a Time Synchronization >>>>>>> Protocol >>>>>>> like FTSP. >>>>>>> How can I integrate FTSP in my application or on any application in >>>>>>> WSN.? Is >>>>>>> there a general method to do it? >>>>>>> >>>>>>> Best Regards, >>>>>>> >>>>>>> -- >>>>>>> Wafa Ben Jaballah >>>>>>> >>>>>>> >>>>>> _______________________________________________ >>>>>> Tinyos-help mailing list >>>>>> Tinyos-help@millennium.berkeley.edu >>>>>> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help >>>>>> >>>>>> -- >>>>>> This message has been scanned for viruses and >>>>>> dangerous content by MailScanner, and is >>>>>> believed to be clean. >>>>>> >>>>> >>>>> >>>>> --Deeksha >>>> >>>> >>>> _______________________________________________ >>>> Tinyos-help mailing list >>>> Tinyos-help@millennium.berkeley.edu >>>> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help >>>> >>>> -- >>>> This message has been scanned for viruses and >>>> dangerous content by MailScanner, and is >>>> believed to be clean. >>>> >>> >> >> >> >> -- >> This message has been scanned for viruses and >> dangerous content by MailScanner, and is >> believed to be clean. >> > > > --Deeksha _______________________________________________ Tinyos-help mailing list Tinyos-help@millennium.berkeley.edu https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
