Wasif, How is it going? I have just seen you mail. First, I am wonder whether you are working with the CI(Constructive Inference) based synchronization. Please introduce more about your project in detail. Second. as Thomas said, you had better make a good understanding of conventional synchronization method such as that in FTSP. #by the way, writing TXFIFO buffer while transmiting if not safe for program that has intense tasks.
About your method of configuring: TX_MODE = 3, is it feasible? Janricliu -----Original Message----- From: Thomas Schmid [mailto:thomas.sch...@gmail.com] Sent: Saturday, February 09, 2013 12:15 AM To: wasif masood Cc: janricliutinyos; tinyos forum Subject: Re: [Tinyos-help] Time to transmit in CC2420! Wasif, Not sure how you are synchronizing, but have a look at how packet timestamps are implemented. You get a timestamp at transmit (SFD rise on the transmitter) that gets added to the packet that is about to be sent out. The CC2420 actually lets you fill the TXFIFO while it is already in transmit, as long as you are fast enough. The receiver then timestamps the SFD, and voila, you have an almost perfect "signal" between the two nodes. The time between SFD rise on the transmit to SFD rise on the receive is ~3.1us with a confidence interval of ~200ns. - Thomas On Fri, Feb 8, 2013 at 1:43 AM, wasif masood <rwmas...@gmail.com> wrote: > > Thanks fellows for your useful inputs. Well apparently, I did dig > deeper into the CC2420 datasheet and what I have decided to do is to > use the CC2420 transmitter test modes i.e. TX_MODE = 3 (as given in > CC2420 datasheet page 65). The idea to use a modulated signal and just > send out the Synchronization header, I think that would be enough to > raise the SDF interrupt at the receiver end. > > > Jonas, > > I am working on a time synchronization protocol according to which, on > reception of an incoming pulse, the receiver has to reply in the same > manner via a broadcast message. Now according to the default stack, as > I mentioned above and you all know very well, that from the point of > initiating a send signal to the point when it should be received, 10's > of ms of delay is already experienced, now what I want to achieve is > to reduce this delay between the send and receive, because according > to my approach the smaller the time between send/rec the better the synchronization would be. > > Thanks! > > BR, > Wasif > > On Fri, Feb 8, 2013 at 8:02 AM, janricliutinyos > <tinyos.janric...@gmail.com> > wrote: >> >> 1. By default, the CSMA is used. Therefore the transmission costs >> more time due to the backoffs. >> >> 2. The on air transmission time as well as the cc2420 buffer writing >> time varies according to the packet size. A small size packet >> requires a shorter period for transmission. >> >> >> >> You should read the CC2420TransmitP file, and do some test based on >> the program given in the ./apps/ >> >> >> >> Janric Liu >> >> >> >> From: tinyos-help-boun...@millennium.berkeley.edu >> [mailto:tinyos-help-boun...@millennium.berkeley.edu] On Behalf Of >> wasif masood >> Sent: Thursday, February 07, 2013 11:03 PM >> To: tinyos forum >> Subject: [Tinyos-help] Time to transmit in CC2420! >> >> >> >> >> >> Hi All, >> >> >> >> I know this is probably an old topic but I still couldn't find any >> solution to that. Well, I am using CC2420Transmit.Send interface >> which is connected to CC2420TransmitC. Even at this lowest level of >> wiring the time elapsed between the Send and SendDone signal is in >> 10's of milliseconds (I have tested it for both Z1 and TelosB). Is >> there any way I can reduce this time? Please share your expertise! >> >> >> >> Regards, >> Wasif Masood > > > > > -- > Wasif Masood > > _______________________________________________ > Tinyos-help mailing list > Tinyos-help@millennium.berkeley.edu > https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help _______________________________________________ Tinyos-help mailing list Tinyos-help@millennium.berkeley.edu https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
