I know that the sequence number prevents routing loops and I roughly understand how it functions. However, say there is a corruption in one of the nodes in its sequence number for a particular route. Will it be possible for a routing loop to form?
I feel like the answer is no. First of all a loop cannot form during path discovery because every node forwards an RREQ only once. If there exists some path between X---S, then some nodes on the path will have already forwarded the RREQ once and will block the loop from forming.
It is also not possible for X---S---T to have been some old path that has been corrupted to a higher sequence. This is because if X---S---T is an old path, S can either at least go directly from S---T without passing X. A new path path passing through X also cannot exist because otherwise the routing table stored at X will not point back towards X---S---T and as discussed, the loop cannot be formed during route discovery.
Is there some part of the analysis that I'm missing?
Is it possible for a routing loop to form using ADOV (Ad Hoc on Demand Distance Vector Routing) if there is a corruption in the sequence number?
