Hi Chun-Yeow, Thanks for your response. I got it now. I have another question, I'll explain it in the following example and also please tell me if my understanding is right.
Assume that at t = 0, source (S) and Destination (D) does not know each other. At t=1 S needs to send data to D. S triggers the path selection procedure which then broadcasts PREQs. Eventually S will get a PREP back from D. And S creates a path from S to D and will then forward data packet on that path at t=1 seconds. It keeps sending data packets until t=4 seconds and stops. Then would the path expire at t = 1 + 6 = 7 seconds or at t = 4 + 5 = 9 seconds? (Considering the timeout to be 5 seconds). When the path did expire at some future point in time t = 7 or 9, and data packets has to be sent by S to D at t = 10 seconds, then it does broadcast PREQs again, from your previous reply, but not unicast PREQ to D? Also assume that there is an intermediate node A between S and D. nodes S and A are moving around. So, sometimes path S-A-D has a better metric than S-D. When I plot both the metrics against time with the number of hops from S->D, I see that not always the path from source to destination has the least metric value. My question is: does HWMP protocol guarantee best metric (low metric value) path all the time? Furthermore, how often is metric value updated or rather how often does HWMP protocol reads the metric value? I appreciate your time. Regards, Siva. On Sun, Jun 9, 2013 at 8:00 AM, <[email protected]> wrote: > Send Devel mailing list submissions to > [email protected] > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.open80211s.org/cgi-bin/mailman/listinfo/devel > or, via email, send a message with subject or body 'help' to > [email protected] > > You can reach the person managing the list at > [email protected] > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Devel digest..." > > > Today's Topics: > > 1. HWMP active path timeout (Sivateja Patibandla) > 2. Re: HWMP active path timeout (Yeoh Chun-Yeow) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Sun, 9 Jun 2013 02:36:52 -0400 > From: Sivateja Patibandla <[email protected]> > To: [email protected] > Subject: HWMP active path timeout > Message-ID: > < > ca+lehheig7xcpbjyshg1bhazs9caj5_nr7r-gzjyo3ivsd5...@mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > Hi, > > I want to know the purpose of the active path time out in the HWMP > protocol. I got different information from different sources and I was > confused with the operation. Here is the information I have: > In the open80211s implementation, after a path is being created, a timer > gets initialized and the path remains valid until *timeout* seconds. After > the timeout, if a new data packet needs to be sent, then a PREQ is > unicasted to the destination but not broadcasted. But in the ns-3 > simulator, PREQ's are broadcasted. From the IEEE 802.11-2012 standard, this > is what I got: > "This attribute specifies the time (in TUs) for which mesh STAs receiving a > PREQ to individual target(s) shall consider the forwarding information to > be valid." > Frankly, I could not understand what the standard document said. So, if > anyone could give me more information about the HWMP active path timeout > parameter, It would be really helpful for my research project. Thank you. > > Regards, > Siva. > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: < > http://lists.open80211s.org/pipermail/devel/attachments/20130609/b86c44a0/attachment-0001.html > > > > ------------------------------ > > Message: 2 > Date: Sun, 9 Jun 2013 15:27:47 +0800 > From: Yeoh Chun-Yeow <[email protected]> > To: "[email protected]" <[email protected]> > Subject: Re: HWMP active path timeout > Message-ID: > < > caefj984vvbnrsbg8x4danh4nsxbp0mjvicytb6bb1orepuk...@mail.gmail.com> > Content-Type: text/plain; charset=ISO-8859-1 > > > timeout, if a new data packet needs to be sent, then a PREQ is unicasted > to > > the destination but not broadcasted. > Unicast PREQ is only for RANN propagation (or enabling gate > announcement) in o11s. For on-demand approach, the PREQ is still > broadcasted. > > After active path timeout, if the originated mesh node would like to > send traffic to a specific mesh node, if requires to trigger the path > selection again especially for reactive mode. > > Hope this help. > > ----- > Chun-Yeow > > > ------------------------------ > > _______________________________________________ > Devel mailing list > [email protected] > http://lists.open80211s.org/cgi-bin/mailman/listinfo/devel > > > End of Devel Digest, Vol 16, Issue 7 > ************************************ >
_______________________________________________ Devel mailing list [email protected] http://lists.open80211s.org/cgi-bin/mailman/listinfo/devel
