1. 6-Address Scheme for TGs Mesh
September 2006
6-Address Scheme for TGs Mesh
Date:
Authors:
Notice: This document has been prepared to assist IEEE It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein.
Release: The contributor grants a free, irrevocable license to the IEEE to incorporate material contained in this contribution, and any modifications thereof, in the creation of an IEEE Standards publication; to copyright in the IEEE’s name any IEEE Standards publication even though it may include portions of this contribution; and at the IEEE’s sole discretion to permit others to reproduce in whole or in part the resulting IEEE Standards publication. The contributor also acknowledges and accepts that this contribution may be made public by IEEE
Patent Policy and Procedures: The contributor is familiar with the IEEE 802 Patent Policy and Procedures < ieee802.org/guides/bylaws/sb-bylaws.pdf>, including the statement "IEEE standards may include the known use of patent(s), including patent applications, provided the IEEE receives assurance from the patent holder or applicant with respect to patents essential for compliance with both mandatory and optional portions of the standard." Early disclosure to the Working Group of patent information that might be relevant to the standard is essential to reduce the possibility for delays in the development process and increase the likelihood that the draft publication will be approved for publication. Please notify the Chair as early as possible, in written or electronic form, if patented technology (or technology under patent application) might be incorporated into a draft standard being developed within the IEEE Working Group. If you have questions, contact the IEEE Patent Committee Administrator at
K. Kim et al.

2. Outline September 2006 Background
Review of 4-Address Scheme and WDS in
6-Address Scheme
New data frame format
Addressing Examples
MP to MP
STA to STA
Interworking: STA to External STA
Summary and Discussions
Straw Poll
Motion
K. Kim et al.

3. September 2006
Background
This proposal is to resolve several major comments including
CID #183: Rules for handling legacy STAs
CID #205: Encapsulation format to transmit frames via a root MP in HWMP
The current 4-address scheme cannot efficiently support all cases of routing & forwarding in mesh networks where frames between legacy STAs associated with MAPs are delivered through multiple MPs, including redirecting MPPs ( Examples).
The use of proxy tables (e.g., [1,2]) neither handles the case of redirecting MPPs unless carrying extra information, nor is a scalable solution for frame forwarding at intermediate MPs.
The encapsulation proposal in [3] incurs rather large overhead due to unnecessary duplication of several header fields.
The 6-address scheme (e.g., [1]) would be desirable not only for TGs mesh networks but also for other applications in the future.
However, MP to MP end-to-end traffic does not need the 6-address scheme – selective use is possible.
K. Kim et al.

4. Review of 4-Address Scheme and WDS in 802.11 (1)
September 2006
Review of 4-Address Scheme and WDS in (1)
What is the (W)DS in existing standards?
No notion of true multi-hopping through nodes.
A kind of shared media/hub/L2 switching implicitly assumed.
At most 1 hop from each other node
See the example on the right (given in [4]).
Hence not a proper model for the s wireless mesh networks.
K. Kim et al.

5. Review of 4-Address Scheme and WDS in 802.11 (2)
September 2006
Review of 4-Address Scheme and WDS in (2)
TGs mesh networks are different from the WDS described in the current standard.
Multi-hopping through MPs (i.e., more than one hops from a source to a destination in general)
Existence of MPPs that can redirect incoming frames to other MPPs or destination MPs/MAPs by remapping of address fields
In tree-based routing in HWMP
In interworking with outside mesh
Tunneling between portals (e.g., for wireless bridging)
K. Kim et al.

6. 6-Address Scheme (1) Mesh Header September 2006
Octets:2 2 6 6 6 2 6 2
4~16
0-tbd 4
FCS
Frame Control
Dur
Address 1 RA
Address 2 TA
Address 3 DA
Seq Control
Address 4 SA
Qos Control
Mesh Header
Payload
Octets: 1
Mesh E2E Seq Number 2
Time To Live 1 12
Mesh Flags
(Optional) Mesh Addressing
Bit 0: Address Extension (AE)
Bits 1-7: Reserved for future use
Address 5 (6 octets)
Address 6 (6 octets)
These fields are always present in mesh frames.
Mesh Header
K. Kim et al.

7. 11s MAC Header (up to Mesh TTL field)
September 2006
6-Address Scheme (2)
To DS
From DS
AE Flag
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
RA=DA
TA=SA
BSSID
N/A
N/P*
N/P 1
TA=BSSID SA
RA=BSSID DA RA TA
Mesh DA
Mesh SA
* N/P = Not Present
11s MAC Header (up to Mesh TTL field)
Address 5
Address 6
Frame Body
FCS
When the AE flag = 0, all fields have their existing meaning, and there exist no “Address 5” and “Address 6” fields – this assures compatibility with existing hardware and/or firmware.
K. Kim et al.

8. 6-Address Scheme – Address Mapping Principle
September 2006
6-Address Scheme – Address Mapping Principle
The ordering of the addresses should be from the innermost to the outermost “connections” ( More examples)
Address 1 & 2 for endpoints of a link between RX and TX
Address 3 & 4 for endpoints of a mesh path between a destination and a source MP
Including MPPs and MAPs
Address 5 & 6 for endpoints of an (end-to-end) 802 communication
A series of mesh paths connected at MPPs (e.g., TBR in HWMP) or
An 802 path between legacy STAs (including nodes outside the mesh) or
Any mixture of them (e.g., an MP to an STA or vice versa). MP MP
MPP MP MP
link
link
link
link
mesh path
mesh path
802 communication
K. Kim et al.

9. Addressing Examples: Network Configuration
September 2006
Addressing Examples: Network Configuration
STA3
Consider the following three cases for different types of path selection modes/protocols:
STA1 → STA2
Through MP2 and MP3 for HWMP on-demand routing and RA-OLSR
Through MPP for HWMP TBR
MP1 → MP4
Only for HWMP TBR
STA1 → STA3
For all three protocols/modes
Wired L2 Network
MPP (Root)
Mesh Network
MP3
MP2
MP4
MP1
MAP2
MAP1
STA1
STA2
K. Kim et al.

10. Addressing Examples: Operational Assumptions
September 2006
Addressing Examples: Operational Assumptions
Maintenance of STA association information*
HWMP
In on-demand routing**, each MAP/MPP locally maintains its STA association and responds to an RREQ message if any of destination addresses matches one of its associated STAs.
In TBR, the Root maintains a global mapping table for all MPs and STAs in the mesh (i.e., Registration assumed; see 11A of [5]).
RA-OLSR
Each MAP/MPP broadcasts its associated STAs to other MAPs/MPPs using LABA (see 11A of [5]).
Each MAP/MPP maintains LAB & GAB for mapping STAs to their associated MAPs (see 11A of [5]).
** The current RREP messages has no provisioning for this scheme, i.e., it needs to be modified to specify MAP as a (proxy) destination. Address for associated STAs.
* External STA (i.e., in wired networks) association at MPPs can be treated same way as wireless STAs. ** RREQ/RREP messages need to be modified as in slide #9 of [2].
K. Kim et al.

11. Example 1a: STA to STA not via Root Portal
September 2006
Example 1a: STA to STA not via Root Portal
STA1
Address 1
Address 2
Address 3
Address 4
MAP1
STA1
STA2
N/A
MAP1
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
MP2
MAP1
MAP2
STA2
STA1
MP2
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
MP3
MP2
MAP2
MAP1
STA2
STA1
MP3
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
MAP2
MP3
MAP1
STA2
STA1
MAP2
Address 1
Address 2
Address 3
Address 4
STA2
MAP2
STA1
N/A
STA2
K. Kim et al.

12. Example 1b: STA to STA via Root Portal
September 2006
Example 1b: STA to STA via Root Portal
STA1
Address 1
Address 2
Address 3
Address 4
MAP1
STA1
STA2
N/A
MAP1
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
MP2
MAP1
ROOT
STA2
STA1
MP2
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
ROOT
MP2
MAP1
STA2
STA1
Root
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
MP3
ROOT
MAP2
STA2
STA1
MP3
Because of the mapping principle in slide #10, the “Address 4 (source of mesh path)” is to be updated at the Root portal, i.e., “MAP1” -> “Root”; alternatively, we can keep the original mesh source information (MAP1) in “Address 4” so that the destination MAP (MAP2) knows the association between “MAP1” and “STA1”. But, what’s the benefit of the latter approach?
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
MAP2
MP3
ROOT
STA2
STA1
MAP2
Address 1
Address 2
Address 3
Address 4
STA2
MAP2
STA1
N/A
Root maps STA2 to MAP2
STA2
K. Kim et al.

13. Example 2: STA to External STA
September 2006
Example 2: STA to External STA
STA1
Address 1
Address 2
Address 3
Address 4
MAP1
STA1
STA3
N/A
MAP1
Address 1
Address 2
Address 3
Address 4
Address 5*
Address 6*
MP2
MAP1
MPP
STA3
STA1
MP2
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
MPP
MP2
MAP1
STA3
STA1
MPP DA SA
STA3
MPP**
STA3
Non (i.e., Ethernet) frame
* Intermediate MPs (here MP2) don’t have to process these fields. ** Ethernet address of MPP’s interface to a wired network
K. Kim et al.

14. Example 3: MP to MP Via Root Portal
September 2006
Example 3: MP to MP Via Root Portal
MP1
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
MP2
MP1
ROOT
MP4
MP2
Address 1
Address 2
Address 3
Address 4
Address 5
Address 6
Root
MP2
ROOT
MP1
MP4
Root
Address 1
Address 2
Address 3
Address 4
MP3
ROOT
MP4
MP1
MP3
Address 1
Address 2
Address 3
Address 4
MP4
MP3
MP1
MP4
K. Kim et al.

15. Summary and Discussions
September 2006
Summary and Discussions
A number of open comments depend on the proposed scheme.
CID #183, #205, #207, #58, #217, #155, ....
Considered both intra-mesh communications and interworking with external wired networks.
The 6-address scheme protects the origin and destination fields with encryption and thus increases the security of the overall approach.
The use of two additional address fields (i.e., “Address 5/6”) is optional, and their existence is indicated by the AE flag 0.
The 6-address scheme may have applications beyond TGs.
That suggests this feature should be documented as a delta on the baseline standard rather than a mesh specific feature.
Regarding frame aggregation/encapsulation, this proposal is compatible with TGn’s MPDU aggregation scheme ( Details).
To recap two major points of rationale behind this proposal:
Maximize backward compatibility
Increase flexibility in adding more functionality in the future
K. Kim et al.

16. 6-Address Scheme: Frame Format Consideration
September 2006
6-Address Scheme: Frame Format Consideration
Encrypted
Octets:2 2 6 6 6 2 6 2 4 12
0-tbd 4
FCS
Frame Control
Dur
Address 1 RA
Address 2 TA
Address 3 DA
Seq Control
Address 4 SA
Qos Control
Mesh Forwarding Control
Mesh Addressing
Field
Payload
Bits: 0-7
Mesh E2E Seq Number
Bits: 8-23
Time To Live
Bits: 0-7
Address 5 (If AE flag = 1)
Octets:6
Address 6 (If AE flag = 1)
Mesh Flags
Bit 0: Address Extension (AE)
Bits 1-7: Reserved for future use
These fields are always present in mesh frames.
Mesh Forwarding Header
K. Kim et al.

17. 6-Address Scheme Straw Poll
September 2006
6-Address Scheme Straw Poll
Shall we accept the 6-address scheme and prepare texts based on it for possible approval during the September TGs meeting?
Yes: 33
No: 5
Abstain: 7
K. Kim et al.

18. September 2006
Proposed Changes
Changes are proposed in the following clauses (see [6] for details):
7.1.2: Change of Mac frame format
a: Descriptions of the new “Mesh Header” field
11A.4.4: Update of the descriptions of data message forwarding based on mesh forwarding extension
K. Kim et al.

19. September 2006
Motion
Move to adopt the changes to P802.11s/D0.03 provided in the IEEE doc /1464r1.
Moved by:
Seconded by:
Results (Yes/No/Abstain):
K. Kim et al.

20. September 2006
References
L. Chu et al., “ST+UCLA TGs Mesh Network Proposal,” IEEE /0379r0.
H. Gossain et al., “Packet Forwarding for Non-routable Devices in Multi-hop Wireless Mesh,” IEEE /0661r0.
J. Kruys and S. Rahman, “Mesh Encapsulation”, Rev. 3.
D. Engwer, “’WDS’ Clarifications,” IEEE /0710r0.
IEEE P802.11s/D0.03
K. Kim et al., “Proposed Texts for TGs Comment Resolution,” IEEE /1464r1.
K. Kim et al.

21. September 2006
Backup Slides
K. Kim et al.

22. 6-Address Scheme – Examples for Link/Mesh Path/Connection (1)
September 2006
6-Address Scheme – Examples for Link/Mesh Path/Connection (1) MP MP
link (= mesh path) MP MP MP MP MP
link
link
link
link
mesh path
STA
STA
MAP MP MP MP
MAP
link
link
link
link
link
link
mesh path
802 communication
K. Kim et al.

23. 6-Address Scheme – Examples for Link/Mesh Path/Connection (2)
September 2006
6-Address Scheme – Examples for Link/Mesh Path/Connection (2) MP MP
MPP MP MP
link
link
link
link
mesh path
mesh path
802 communication
STA
STA
MAP MP
MPP MP
MAP
link
link
link
link
link
link
mesh path
mesh path
802 communication
K. Kim et al.

24. 6-Address Scheme – Examples for Link/Mesh Path/Connection (3)
September 2006
6-Address Scheme – Examples for Link/Mesh Path/Connection (3)
STA MP MP
MPP MP
MAP
link
link
link
link
link
mesh path
mesh path
802 communication
Wired Network
STA
External STA
MAP1 MP
MPP MP
MPP
link
link
link
link
link
mesh path
mesh path
802 communication
K. Kim et al.

25. A-MPDU Aggregation Robust Structure A purely-MAC function
September 2006
A-MPDU Aggregation
Robust Structure
A purely-MAC function
PHY has no knowledge of MPDU boundaries
Simplest MAC-PHY interface
Control and data MPDUs can be aggregated
Limited to a single rate
summary deck
K. Kim et al.

26. A-MSDU Efficient Structure MSDUs of the same TID can be aggregated
September 2006
A-MSDU
Efficient Structure
MSDUs of the same TID can be aggregated
Error recovery is expensive
K. Kim et al.

27. Bursting Does not support multiple responses from multiple receivers
September 2006
Bursting
No idle gap
Normal ACK policy Np Ns
PSDU Ns
PSDU Ns
PSDU
xIFS Np Ns
PSDU Ns
PSDU Ns
PSDU
Non-normal ACK
Last PSDU bit Np
= N-Preamble Ns
= N-Signal robust encoding rate
Sequence of MPDUs or PSDUs from same transmitter
on-the-fly aggregation
Reduced inter-frame spacing
0 usec if at same Tx power level and PHY configuration
2 usec otherwise (with preamble)
Multiple RAs allowed within the burst
multiple rates within burst
allows varying TX power within burst
Block Ack Request and Block Ack frames allowed within the burst
Does not support multiple responses from multiple receivers
K. Kim et al.

28. On the Use of TGn Frame Aggregation Schemes in TGs
September 2006
On the Use of TGn Frame Aggregation Schemes in TGs
The A-MPDU structure seems simplest and best suited for Mesh applications
The A-MSDU structure requires overhead that seems to add no value
Bursting is something that happens “below the horizon”
Considering that TGn does work on features that TGs can leverage, there is no need for TGs to spend (more) time on aggregation, other than encouraging TGn to make sure the A-MSDU makes it to the TGN amendment
K. Kim et al.