Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802SEC] 802.1 comments on the P802.15.10 draft PAR/5C



Colleauges -

This went out in a slightly unfinished state - my apologies. I will send the final version shortly....

Regards,
Tony


On 16 July 2013 15:29, Tony Jeffree <tony@jeffree.co.uk> wrote:
802.1 has studied the 802.15.10 draft PAR and 5C, and need further clarification in order to understand the intent behind this project.

The PAR scope states that the object of the recommended practice is to "route packets dynamically".  While there is no fundamental difference between "bridging frames" and
"routing packets", in the context of IEEE 802, this statement seems to say that 802.15.10 will do something different to 802.1 bridging.  However, this appears to be a bridging PAR, and that being the case, there should be opportunities for 802.15 to leverage the existing Bridging work, rather than creating something new and potentially incompatible with Bridging.

The responses to the compatibility aspects of the 5 Criteria indicate that the recommended practice will not mandate compliance with IEEE Std 802,  IEEE Std 802.1D or IEEE Std 802.1Q.  However, the base standard and this amendment are compliant with the draft D1.5 IEEE 802, which says "... traffic between EUI-64 and EUI-48 addressed networks needs to be routed at a layer above the DLL."  Hence, the current IEEE Std 802.1D and IEEE Std 802.1Q are not applicable as they only support EUI-48.

The quoted statement in IEEE Std 802 points out that you have to go above the Data Link Layer; however, reading the draft PAR, 802.15.10 is staying within the Data Link Layer.  The addresses being used are 64-bit MAC addresses, not IP addresses.  

The 5C response goes on to say that in the case where the 64-bit address space is effectively used as a 48-bit address space, then 802.1 Bridging would be fully supported. However, in this case, there is no work to be done other than the need to specify text for inclusion in IEEE Std 802.1AC to specify how an 802.15 device supports the Internal Sublayer Service and thereby supports 802.1 Bridging.

It seems that there are at least the following possibilities:

- It is the intent of this project to provide a means of bridging between 802.15 devices that support 64-bit addresses. If that is the case, then the re-use of existing 802.1 Bridging technology would make a great deal of sense, rather than inventing something new and potentially incompatible. Going in this direction would probably result in changes to existing 802.1 standards (802.1AC and/or 802.1Q) in order to accommodate different MAC address lengths in the Bridging standards.

- It is the intent of this project to provide a means of bridging between 802.15 devices and other 802 MAC technologies that use 48-bit addresses. If that is the case, then the only way this can be achieved is by encoding a 48-bit address in the 64-bit address fields of the 802.15 MAC, and this would require PARs for (a) changes to 802.1AC to specify how the 802.15 MAC supports the ISS, and (b) potentially, changes to the 802.15 standard to specify how 48-bit addresses are encoded in the address fields of 802.15 frames.

- It is the intent to provide a means of bridging between 802.15 devices that use 64-bit addresses and other 802 MAC technologies that use 48-bit addresses. If that is the case, then bridging is not an option (for the reasons stated in IEEE Std 802) and the only option is to perform routing at layer 3, which is technology that already exists in IETF specifications.

802.1 would like to engage in discussion with 802.15 in order to clarify the intent of this draft PAR before it is submitted to the EC for approval, in the spirit both of the existing P&P and the changes to the P&P regarding PAR submission that are under discussion this week. Some initial discussion took place in 2012, and resulted in Norm Finn giving the following presentation in an 802.15 interim in September 2012:


It is unfortunate that his suggestion of collaboration with 802.1 was not taken up; however, it is not too late for that to be rectified. We currently have a project under way in 802.1 that was the result of a highly successful collaborative activity with 802.11 in order to properly specify 802.11 Bridging, both between wired and wireless media, and among wireless stations (this will be documented in 802.11ak and 802.1Qbz);

IEEE 802.1 has also recently worked very well with 802.16, and is currently executing PARs jointly created with 802.11  to support bridging, .  All of these media, 802.16, 802.11, and 802.3, can look forward to full interoperability with each other in a bridged network.  We se no reason why 802.15 cannot be a part of this joint effort.


Regards,
Tony
​​


---------- This email is sent from the 802 Executive Committee email reflector. This list is maintained by Listserv.