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

RE: [802.21] [802-11TGU] IEEE 802.21 Teleconference Call



Well personally I got back four copies of my last email, all of which were corrupted.  So here it is again in plain text - apologies to anyone who managed to get it correctly one of the previous times...

Mike.

All,

As I spoke against the existing text, I though I ought to suggest some changes.  Attached are some suggestions for re-phrasing the existing bullets.

Mike.


 REPLACE:
-         The definition of ‘link’ in IETF is different from that in IEEE 802. It is understood that within the IETF, ‘link’ is an IP layer notion, whereas within IEEE 802, ‘link’ is a Layer 2 notion. Furthermore the IEEE 802 term “link” refers to the association between the STA and AP, whereas within the IETF, a ‘link’ is the area that could be reached by a link-local address or broadcast address. It is felt that this distinction perhaps could be clarified within the draft to avoid confusion between our two groups.



WITH

-         The definition of ‘link’ used in IETF seems most analogous to the IEEE 802 term “logical link” which is the service provided by two Logical Link Control entities communicating across one or more LAN segments.  Within IEEE 802 the term “link” is used in different ways, and is often used for a LAN segment, such as the association between a STA and an AP in IEEE 802.11.



 REPLACE

An additional concern with this issue, is that the use of the term “link” within the IETF, may refer to multi-hop L2 links, whereas the IEEE 802 term would only apply to a single L2 link. For example, would use of the term ‘Link-down’ imply that all the links are down, or just one hop?



WITH

-         Where a logical link consists of multiple LAN segments, the semantics of terms such as “Link-down” are unclear.  Do they apply to the local segment, intermediate segments, or all segments?



REPLACE

Another terminology issue, concerns the meaning of ‘link up’. It is understood that within the IETF, ‘link up’ refers to the fact that you can use a certain IP address for information exchange over that link.  However, within IEEE 802, ‘link up’ may simply mean that a  radio connection has been established.



WITH

-         With terms such as “link up” it is important to distinguish between the multiple possible meanings, including the establishment of radio communication, the completion of link layer authentication, and the commencement of layer 3 communication after possible determination of a layer 3 address.



REPLACE

The term ‘link quality’ is proposed in IEEE 802.21. This is distinguished from ‘link up/down’. It is felt that this term deals with a layer 3 concept, whereas the use of the terminology ‘link up/down’ is useful to characterize the wired media. For wireless media, it is unnecessary to coerce the ‘on/off’ point of view.  In heterogeneous media, ‘link down/up’ has different meanings to different applications and media types.  Perhaps it should be considered that status is more important. The notion ‘link’ should take the application requirements into considerations. Clarify this paragraph



WITH

-         The definition of “link up/down” is application specific, especially in a wireless environment where mobility may lead to changes in available bandwidth or quality of service.  For this reason IEEE 802.21 prefers to signal “link quality” instead..



REPLACE:

If one considers a scenario of multiple interfaces and multiple applications, it may be a requirement that only one application undergoes a handover.  Hence, in this situation, the use of the term ‘handover finish’, is not necessarily ‘link down’. The handover decision may be related to application requirements, not only ‘link up/down’. Relate to a specific section of the draft.



WITH

-         As the concept of link suitability is application dependent, the decision to initiate handover will also be application dependent.  Where multiple applications are using the same interface, the results of the handover decisions may very, and only a sub-set of applications may handover from one interface to another.





DELETE

-         The draft implies that most handovers are dependent upon the isappearance of the AP.  It is felt that there are other use cases, such as moving out of radio range which should also be considered. Clarification from 802.21 is required. [This has been removed because the draft includes the words “Where a mobile 802.11 STA encounters a series of consecutive non-acknowledged frames, the most likely cause is that the station has moved out of range of the AP.]