Re: [manet] SMF in Manet and MPR

["Adamson, Robert B CIV USN NRL (5522) Washington DC (USA)" <[e-mail]>]

Hi Henning,

I didn't participate in the discussion you mention but agree that heterogenous MANET is particularly challenging for multicast.  SMF by itself with distributed CDS relay set formation can't optimize (or even approximate it well) without multi-hop signaling.  I have been continuing to experiment (part time as I have opportunity) with the Elastic Multicast concept we presented several years ago.  The current "nrlsmf" implementation includes a variety of Elastic Multicast operating modes.  One of those use the an Elastic Multicast flow advertisement (EM_ADV) message that propagates from multicast sources and builds up an ETX path metric that can be used by "downstream" nodes (and/or receivers) to select (and acknowledge via EM_ACK messages) upstream relays according to that metric.  I also plan to extend the code to support advertising the maximum supportable path bandwidth (and possibly even RTT) for flows in that signaling including report back to the source of available path bandwidth.   That would allow for other metrics to be used in the forwarding tree/mesh selection.  This quickly gets into the application-specific domain if imagine how that information could be used.  It does seem there could be benefit for multiple applications being able leverage a common, underlying signaling approach like this.  Even though my code does not yet use PACKET BB (I was lazy on that while I'm still figuring out what meaningful information signaling messages should include), that could be done and integrated as an add-on to existing MANET signaling.  I.e., as opposed to each application implementing its own independent signaling.

Best regards,

Brian

On 3/21/22, 9:08 AM, "manet on behalf of Henning Rogge" <[e-mail] on behalf of [e-mail]> wrote:

    Hi,

    I am reacting to the talk during the IETF meeting (chat didn't worked
    for me for some reason).

    My trouble with the MPR optimization for SMF (and a bit with SMF in
    general) is that we currently have no good MPR algorithm for
    heterogeneous Manet (Manet with multiple different types of radios).

    The only described MPR algorithm (OLSRv2 appendix) works separately
    for each interface... which means it always selects MPRs on long-range
    (and slow) interfaces, which is really not what we want.

    Of course there is also the problem that SMF completely ignores
    routing metrics... and the problem that we lack a good dataplane for
    SMF, which often results in horrible "raw socket trickery"...

    So in general I would not suggest to people using SMF because I have
    yet to find an use-case where SMF is doing well. Often an
    application-specific forwarding strategy is much better than a generic
    IP one in terms of multicast in Manet.

    Henning Rogge

    _______________________________________________
    manet mailing list
    [e-mail]
    https://www.ietf.org/mailman/listinfo/manet