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

Re: [802SEC] Mobile Broadband Wireless Access--distinction between the ECSG and 802.16 SG



Dear SEC Colleagues,

 

As a non-voting member of the SEC and as chair of the MBWA ECSG I feel it appropriate to respond to Roger's assessment of the choices facing the SEC in deciding on the progression of the 802.16 and ECSG PARs on mobility. The options are those stated by Roger, as the SEC is well aware.

 

In my view Roger's note goes somewhat beyond merely outlining the options and tries to provide an assessment of how the two PARs stack up. Roger says that he could "see the issues from a wide angle"; however, his comments appear influenced by the fact that he is the chair of 802.16.   He discusses the relationships between the PARs and implies that the work of MBWA ECSG may infringe on work that may be going on in other bodies (though he does state that this may be the case for both PARs).  In this note I will provide my perspective on the issues Roger has raised, and explain why the MBWA ECSG PAR deserves to be approved and why the project appropriately belongs in the IEEE.

 

First let me reiterate that it has always been the position of the MBWA ECSG participants that a PAR that takes no a priori position about which technology is used to address the vehicular mobility space is broad enough to accommodate the evaluation of all potential technologies including 802.16. The MBWA ECSG split from 802.16 because agreement could not be reached since the 802.16 participants (although not unanimously) decided the use of the 802.16 PHY and MAC must be required - "there was no room in the inn" for any other technology. The PAR (http://grouper.ieee.org/groups/802/mbwa/Dot16_Ar/80216sgm-02_05.pdf) that was rejected by 802.16 in July had no statement about the technology to be adopted and (at the insistence of 802.16 participants) was constrained only to using spectrum licensed for mobile application with no upper limit on the frequency range. That PAR was a true superset of the two subsequent PARs currently under consideration. The current MBWA ECSG PAR addresses the "sweet spot" in mobility, i.e. a maximum operating frequency of 3.5 GHz. As described in the Five Criteria, the project is eminently realizable from both a technological and economic viewpoint as it can utilize technologies already  used for or optimized for high performance, high capacity IP-based mobile services.

 

The MBWA ECSG proposed project is unique in scope both within the currently authorized work in 802 and SDOs other than the IEEE. In contrast to work currently ongoing in the Partnership Projects (3GPP and 3GPP2  http://www.3gpp.org/Management/WorkPlan.htm & http://www.3gpp2.org/Public_html/SC/SC-20010201-010_PMT_Chair,3GPP2_Workplan_V23.xls), the MBWA ECSG PAR is focused on:

 

1.       A Radio Access Network (RAN)  that is optimized to support broadband packet data, specifically to support the transport of IP traffic, and

2.       A pure  packet switched network infrastructure, rather then the current 3G predominantly circuit switched control infrastructure and architecture

 

This emphasis makes the project a perfect complement to existing projects in the IEEE that are focused on an all packet switched environment. That being said, it is obvious that just like in the land-line telephony environment where there is a trend to unify voice and data on a packet switched infrastructure, this trend will also occur in the wireless environment. For that reason, the MBWA ECSG PAR identifies the 3G partnership projects and the ITU as a liaison organization with which cooperation is expected to occur. Liaison overtures have been made, but as a short lived SG it would have been inappropriate to commit to a long lasting relationship at that time. Instead, as the PAR indicates, these relationships will be established when the SEC has approved the PAR and established a new Working Group.

 

 

To address Roger's specific points:

 

(1) It is stated that "802.16 believes that the natural evolution of its broadband wireless access technology is to allow its terminals (in licensed bands below 6 GHz) to move during operation". It is not at all clear how the work proposed is a "natural evolution" for 802.16. The Working Group is chartered to work on fixed broadband wireless and covers both unlicensed and licensed bands. Its current specs are for the 2-11 GHz (.16a) and 10-66 GHz (.16). These are both different from the "licensed spectrum < 6 GHz" band that is targeted with the new PAR. A more natural evolution would have been to allow limited mobility of terminals, with backwards compatibility to fixed terminals, operating in the current bands.  

 

As stated above, the MBWA ECSG PAR advocates the development of a specification that is optimized for macro-cellular mobility and makes no a priori assumptions about technologies to be used. Trade-offs involved in supporting mobility are significantly different than those made to support fixed wireless operation. The MBWA ECSG PAR infringes neither on the work of 802.11, the current work of 802.16 or the work of 3GPP or 3GPP2.

 

(2) In this point it is argued that "station speed" should not be used as a "differentiator" since it has not been used as such in the past in IEEE 802. For a project whose focus is on mobility, however, station speed is an obvious key parameter in the specification of the project.  Any PAR for a mobile standard that does not clearly and adequately address speed is incomplete.

 

The distinction proposed based on speed is actually supported by Roger himself in a suggestion made by him in http://ieee802.org/16/arc/802-16-mobile/msg00032.html advocating a two PAR approach. One PAR would address vehicular mobility. The other PAR, produced by or for 802.16, "would be defined as an amendment to IEEE Std 802.16 to introduce limited mobility and portability. Depending on the WG consensus, this might broaden to repeaters, MAC management, and some of the other issues brought up at the last TGa meeting, as recorded in the minutes (802.16a-02/10)."

 

Roger states that the distinction based on speed would "limit one to low speeds while placing no limits on the other". The MBWA ECSG PAR does specify a challenging design envelope with a maximum speed limit of 250 km/h and a minimum sustained spectral efficiency of 1b/s/Hz/cell. The 802.16 SG charter, on the other hand, was to evaluate what degree of mobility could be supported with the constraints of the existing 802.16 PHY and MAC specifications (comprised of technologies optimized for relatively static channels and wide channel bandwidths) and possible (minor?) enhancements to that PHY and MAC. In other words the terminal speed limit would fall out naturally from the evaluation of the capabilities of existing 802.16 PHY and MAC technology. This evaluation was to have been done during the first two meetings of the new 802.16 SG (http://grouper.ieee.org/groups/802/16/mobile/docs/80216sgm-02_08.pdf ). This evaluation of an 802.16 compliant system was not done.

 

(3) Roger raises the issue of how one or both of these PARs relate to the 3G Partnership Projects and how this might impact the relationship between the Partnership Projects and the IEEE LMSC. In describing the liaison meeting with T1P1, one is left with the impression that the MBWA ECSG project and 802.16 SG project were perceived differently by T1P1.  However,  the description of events at the T1P1 meeting are based on a presentation of an 802.16 SG scope and purpose (limited mobility)   that are not consistent with the changes in  the scope and purpose of the current 802.16 SG PAR under consideration. When the 802.16 WG voted to terminate the efforts of the MBWA in July 2002, it immediately established a Mobile WirelessMAN committee whose charter was to evaluate what mobility characteristics could be supported by 802.16a compliant systems. As stated previously, that evaluation was not done. The 802.16 project presented to T1P1 was an evaluation project and is not representative of the operational characteristics of the final PAR. Here are excerpts from the T1P1 minutes (ftp://ftp.t1.org/T1P1/2002/2P100970.pdf ) characterizing the presentation.

 

"Dr. Marks explained that two new areas of work were being explored in IEEE 802.  The first area of work focuses of expanding the current 802.16 standards to support mobility functions. The second potential area of work, outside of 802.16 would produce standards for Mobile Broadband Wireless at Vehicular Speeds. 

 

Work in 802.16 to extend Standards to include Mobility

 

Dr. Marks noted that at this point this activity is at the study group level, and the actual direction of the work was still under discussion and is evolving.  Dr. Marks noted that 802.16 groups lacked expertise in areas such as hand-off, layers higher than MAC, etc., and T1P1 with extensive knowledge in these technical areas could provide a helping hand." 

 

As stated previously, I believe that the ECSG PAR as defined now is distinct from current work in the Partnership Projects in that it addresses a data optimized air-interface for a pure packet switched infrastructure.

 

In closing, it should be noted that the MBWA ECSG PAR and Five Criteria are the product of seven months of deliberate work, with significant participation and consensus from a large number of interested parties and I would urge you to approve the MBWA ECSG PAR on it's merits and establish a new MBWA Working Group to develop Mobile Broadband Wireless Access standards.   Conversations with leaders in various 3G partnership organizations have actually shown that they would be very much interested in liaising and cooperating with such an effort.

 

I believe an IEEE 802 MBWA project will be the driving force for the development of a world class commercially successful vehicular mobility solution for broadband wireless data that will enhance IEEE LMSC's reputation for excellence and continue the IEEE LMSC's wireless success story.

 

Best Regards,

 

Mark Klerer