[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [manet] RREP generation in AODV



Title: Message
Hi Varnit,
 
Unicast will cause the RREP to travel back along the same path by which the RREQ traveled.  This serves a couple of functions, one of which is assuring that the intermediate nodes can update their tables and another is that it attempts to identify bidirectional links.  It generally achieves this, ignoring the random effects of interference and fading, as well as the possibility that movement or something else would cause an AB link to go bad before the BA link is used for the RREP.
 
John Mullen
 
-----Original Message-----
From: manet-bounces at ietf.org [mailto:manet-bounces at ietf.org] On Behalf Of Varnit Suri
Sent: Tuesday, July 13, 2004 3:26 AM
To: John Mullen; 'manet'
Subject: Re: [manet] RREP generation in AODV

hi
thanks for your time and reply
with reference to your message below, is it that the RREP is going to follow the same path from the destination (or inermediate node) to the originator, as the RREQ did ? if that is the case, then are we not assuming the presence of bi-directional links ?
thanks
varnit
 
----- Original Message -----
Sent: Monday, July 12, 2004 8:46 PM
Subject: RE: [manet] RREP generation in AODV

Hi Varnit,
 
Because the RREPs are unicast back to the previous node in the path, the two counts would be identical.
 
John Mullen 
-----Original Message-----
From: manet-bounces at ietf.org [mailto:manet-bounces at ietf.org] On Behalf Of Varnit Suri
Sent: Monday, July 12, 2004 10:43 AM
To: Koojana Kuladinithi; 'manet'
Subject: Re: [manet] RREP generation in AODV

hi koojana
thanks for ur reply and time...
 

>2. The RFC says: "Once created, the RREP is unicast to the next hop toward the originator of the RREQ, as >indicated by the route table entry for that originator. As the RREP is forwarded back towards the node >>>>which originated the RREQ message, the Hop Count field is incremented by one at each hop. Thus, when >>>the RREP reaches the originator, the Hop Count represents the distance, in hops, of the destination from >the originator." Is the Hop Count field in the RREP not supposed to contain the number of hops from the >>>source to the destination, rather than from the destination to the source ?

Why do you think this way? Anyway, Hop counts from source to the destination is known while processing the RREQ.

Section 5.2 of the RFC: the Hop Count field in the RREP is supposed to contain "The number of hops from the Originator IP Address to the Destination IP Address.".
Section 6.6 (also quoted above as Point 2): "Thus, when the RREP reaches the originator, the Hop Count represents the distance, in hops, of the destination from the originator."
 
Are these two statements not mutually contradictory ? This is because the processing of the RREP (as mentioned in Section 6.6 and quoted above) seems to be such that the Hop Count field of the RREP, when it will finally reach the Originator, will contain the number of hops from the Destination to the Originator, and not vice-versa. This is in direct contradiction with what is specified about the Hop Count field in the format of the RREP in Section 5.2.
 
Also, as you rightly said, when a RREQ reaches the node which issues the RREP (whether the final Destination or an intermediate node), its Hop Count field contains the number of hops from the Originator to the Destination. This is precisely the information that the Source (and not the Destination) needs, so as to enter it in the Hop Count field of its Route Entry for the particular Destination. However, if the RREP contains the number of hops from the Destination to the Source (and not vice-versa), the Originator will never learn of this information. This is why I think that the Hop Count field of the RREP MUST contain the number of hops from the Originator to the Destination, and not vice-versa. Please correct me if I am wrong with this.
thanks for ur time
varnit
----- Original Message -----
Sent: Monday, July 12, 2004 1:01 PM
Subject: RE: [manet] RREP generation in AODV

Hi Varnit,

>This is a query regarding the RREP generation in AODV, from Section 6.6 of its Internet Draft, >RFC 3561. >1. The RFC says: "When generating a RREP message, a node copies the Destination IP >>Address and the Originator Sequence Number from the RREQ message into the corresponding fields >in the RREP message."

>To which field in the RREP does the Originator Sequence Number from the RREQ go ? There is only  >Destination Sequence Number field in the RREP.

This was discussed in the ML. This should read Orginator IP address instead of orginator Sequence number.

>2. The RFC says: "Once created, the RREP is unicast to the next hop toward the originator of the RREQ, as >indicated by the route table entry for that originator. As the RREP is forwarded back towards the node >>>>which originated the RREQ message, the Hop Count field is incremented by one at each hop. Thus, when >>>the RREP reaches the originator, the Hop Count represents the distance, in hops, of the destination from >the originator." Is the Hop Count field in the RREP not supposed to contain the number of hops from the >>>source to the destination, rather than from the destination to the source ?

Why do you think this way? Anyway, Hop counts from source to the destination is known while processing the RREQ.

>3. The RREP is supposed to be Route Reply, ie it is supposed to carry routing information that the source >can use to send packets to the destibation. However it does not have any field that tells the source about >such a route or the next-hop it should choose for a route to the destination. Is it that the source will >>simply take the node from which it first received the RREP as the next-hop for a route to the destination >?

When propagating RREQ, each node creates (source and intermediate) the route towards the source. AODV works to find a minimum hops path between the S (Source) and D (Destination). Therefore, D will only process the RREQ with minimum hops in case if it gets more than one RREQs for an unique discovery. Then D will send RREP over this minimum hops paths that are found during the route discovery process. When receiving RREP over minimum hops path, S will create a route towards the D over this path via next hop. In this way, RREP does not require any other fields. What you asked is needed for DSR (I guess).

Regards

Koojana


Koojana Kuladinithi BSc, MSc
University of Bremen, FB 1 - IKOM/ComNets
Otto-Hahn-Allee NW 1, 28359 Bremen - Germany
Tel.: +49 421 218 8264, Fax: +49 421 218 3601,
Email: koo at comnets.uni-bremen.de
WWW: http://www.comnets.uni-bremen.de/~koo


_______________________________________________
manet mailing list
manet at ietf.org
https://www1.ietf.org/mailman/listinfo/manet