Hi Paul,

Sorry for the delay in my response.

The new concept of link is defined by Jonathan Hui's draft on route-over ND.

As far as I understand, they extended the concept of link from the
traditional IP subnet link to a concept where link is bounded by the
radio reachability. Since the sensors are not connected by the wired
link, the definition, as I understand is the link which encompasses
the neighbors who can send/receive messages to each-other.

Jonathan or David, please correct if this definition is incorrect.

Thanks,
-Samita

On Sun, Aug 17, 2008 at 11:52 PM, Paul Chilton <[EMAIL PROTECTED]> wrote:
> Hi Samita
> This set of design goals looks like it will meet the WG requirements.
> However, I am not sure I understand the bullet point
>
> "Define a link that serves 6lowpan common deployment"
>
> Could you expand on what you mean by this?
> Thanks
> Paul
>
> On Sun, 2008-08-10 at 17:14 -0700, Samita Chakrabarti wrote:
>> Hi All:
>>
>> Since we are starting to work on the combined ND solution for 6lowpan,
>> as one of the members of the ad-hoc ND team, I thought it would be
>> very useful to verify our solution guidelines with the working group.
>> I am starting with a few bullet points, please check if that meets our
>> (6lowpan wg) requirements. Please feel free to suggest modification
>> when it makes sense.
>>
>> ND Design goals:
>>
>>
>> * Neighbor discovery using IPv6 ND protocol or its variant
>> * The solution will use unicast/anycast messages to reduce mulitcast
>> * The solution should look into minimizing number of total ND messages
>> *  Initial boot-strapping information and fault-tolerance should be
>> addressed if possible
>> * This solution must provide methods for IPv6 address allocation in
>> 6lowpan nodes
>> * Define a link that serves 6lowpan common deployment
>> * The design should work with mesh-under(L2) and route-over(L3) routing
>> * The design should be scalable for adding new nodes into the network
>> * The solution should also keep in mind that 6lowpan is a low bW network
>> * The ND design  may not necessarily address mobility in boot-strapping
>>    and address assignment, but it may be extendable to address
>> mobility in future.
>> * The design will use IPv6 auto-configuration as the main method for
>> address assignment; however stateful address configuration will also
>> be considered
>>
>>
>> Thanks,
>> -Samita
>> _______________________________________________
>> 6lowpan mailing list
>> [email protected]
>> https://www.ietf.org/mailman/listinfo/6lowpan
>
> Paul Chilton
> Jennic Ltd
> Furnival Street
> Sheffield, UK
> S10 3DX
>
>
>
_______________________________________________
6lowpan mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lowpan

Reply via email to