Re: [6tisch] plugfest synchronization point 1

Xavi,

As discussed during the call today, the contribution to the plugfest of
Linear Technology's (through its Dust Networks product group) will be to
demonstrate end-to-end connectivity between an Internet host and a
low-power node based running a protocol stack based on IEEE802.15.4e and
6LoWPAN.

This demonstration is in collaboration with Cisco. Specifically, we will
interconnect a low-power mesh (based on Linear's SmartMesh IP product line)
to the 6LoWPAN-ND-capable Cisco switch described by Pascal. Interconnection
will be done through an SBC in charge of 6LoWPAN compaction.

Data coming "down" from the Internet will go through the switch which
handles ND, the SBC taking care of 6LoWPAN compaction, and into the
SmartMesh IP low-power mesh. Data going "up" takes the inverse route.

As discussed during the call, time permitting, we can put together 1-2
slides with an overview of the architecture and technology demonstrated for
the next call.

Thomas

On Thu, Feb 6, 2014 at 11:53 PM, Pascal Thubert (pthubert) <
pthubert@cisco.com> wrote:

>  Hello Xavi:
>
>
>
> Cisco's contribution to the plugfest will be a demo of the backbone router
> functionality as presented at IETF 86 in Orlando, limited to the 6LoWPAN ND
> registration (RFC 6775).
>
> I will have a sneak preview of efficient ND support (WiND) for those
> interested. The code will run on an IE3K industrial switch. The idea is to
> interface with one or more 6TiSCH root node(s) over Ethernet.
>
>
>
> In order to connect to the switch acting as backbone router, the root node
> will need to implement RFC 6775 or WiND to advertise (towards the switch)
> any device that it learns (e.g. through RPL DAO) from the 15.4e TSCH side.
>
>
>
> The root will use its own MAC address (the address on the Ethernet
> connection to the switch) as SLLA in the registration, but present a real
> unique identifier for each device in the 64-bit unique ID that it will need
> to infer in some fashion (sadly RPL is missing the unique ID, but if all
> the addresses are based on EUI 64 then that can be done in a trivial
> fashion). The dependency is that a same device claiming a same address must
> always appear with the same unique ID, otherwise the registration may be
> rejected as duplicate if it happens within lifetime.
>
>
>
> The IE3K switch will act as 6LR on its SVI interface, and register the
> device. It will also defend the IPv6 address of the device on the Ethernet
> side using the root MAC address (the SLLA in the NS(ARO)).
>
> As such, the demo will implement
> http://tools.ietf.org/html/draft-thubert-6lowpan-backbone-router-03#section-5.2"ND Proxy Operations" and will enable a PC connected over Ethernet to ping
> a 6TiSCH device effectively forming a single heterogeneous multilink subnet.
>
>
>
> So far, we are in line with the plan. Anyone interested to connect is
> certainly welcome!
>
>
>
> Cheers,
>
>
>
> Pascal
>
>
>
> *From:* 6tisch [mailto:6tisch-bounces@ietf.org] *On Behalf Of *Xavier
> Vilajosana Guillen
> *Sent:* jeudi 6 février 2014 21:34
> *To:* 6tisch@ietf.org
> *Subject:* [6tisch] plugfest synchronization point 1
>
>
>
> Dear all,
>
>
>
> we reached the first synchronization point. I would like to call to
> plugfest participants to provide a very quick status update and raise any
> possible problem during the 6TiSCH call tomorrow. We will have some minutes
> during the call to talk about that and start some planning.
>
>
>
> regards,
> Xavi
>
> _______________________________________________
> 6tisch mailing list
> 6tisch@ietf.org
> https://www.ietf.org/mailman/listinfo/6tisch
>
>