1. MDA Simulation Study: Robustness to non-MDA Interferers
March 2007
doc.: IEEE /0356r0
March 2007
MDA Simulation Study: Robustness to non-MDA Interferers
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
Ye Chen, Motorola
Ye Chen, Motorola

2. March 2007
Abstract
This document presents results from simulation experiments conducted using an OPNET model of mesh deterministic access (MDA).
The impact of interference from non-mesh STA on MDA is investigated using the model.
The concerns raised in comments 534, 536, 1239, 1874, 3549, 3575, 3617 and 5622 may be addressed using the results presented in this document
Ye Chen, Motorola

3. Problem Statement Concerns have been raised about MDA (see LB93)
March 2007
Problem Statement
Concerns have been raised about MDA (see LB93)
What benefit does MDA provide compared to EDCA?
Are the benefits worth the complexity?
Does the presence of non-MDA stations negate any gains?
Can these benefits be realized in real scenarios?
Concerns are understandable, given the limited amount of published information on MDA
What can be done to start addressing these concerns?
Ye Chen, Motorola

4. Status Created an OPNET simulation model for MDA
March 2007
Status
Created an OPNET simulation model for MDA
Used the model to investigate:
If a mesh is exposed to outside interference (e.g. from an overlapping infrastructure BSS) how is the performance of the mesh impacted
If the mesh uses MDA flows to forward traffic between mesh points, how well does MDA hold up against the interference
Goal
By studying the problem we hope to answer some of the issue raised about MDA during letter ballot
Ye Chen, Motorola

5. Use Case Scenario March 2007 Small single channel mesh
Mesh AP forward traffic to/from non-mesh STA
Mesh AP connect to wired DS via mesh links
All devices communicate at 6 Mbps
Mix of devices is 6 mesh AP and 10 non-mesh STA
Mesh is large enough to include hidden terminals
AP6 is hidden from AP2
AP1 is hidden from AP7
AP2 is hidden from AP3
Mesh AP1
Mesh AP3
Mesh AP6
Mesh AP2 / Mesh Portal
Mesh AP7
Wired DS
Ye Chen, Motorola

6. Use Case with Interference
March 2007
Use Case with Interference
Mesh AP2 / Mesh Portal
Mesh AP7
Mesh AP3
Mesh AP1
Mesh AP6
Wired DS
Infrastructure BSS on Same Channel
Infrastructure BSS added to act as interferer, has overlapping coverage with the mesh
Run simulations with and without MDA to study the benefits of MDA in the presence of co-channel interference
Ye Chen, Motorola

7. Model Overview OPNET model
March 2007
Model Overview
OPNET model
Supports EDCA and contention based queuing of QoS traffic
Supports MDA for voice flows
Models CSMA/CA and enhanced NAV for MDA
Mesh AP are at fixed locations, stations move around mesh AP
Transmit rate fixed at 6 Mbps
Typical Outdoor Suburban-WiFi Path Loss Exponent (e.g. ~ 2.6)
Voice traffic model
Constant bit rate, bidirectional RTP/UDP/IP voice traffic flows
8 kbps vocoder, 20 ms frame interval
Voice flows start at random times after start time of simulation
Ye Chen, Motorola

8. MAC Parameters Mesh AP (for EDCA & MDA): (DIFS,3,7)
March 2007
MAC Parameters
Mesh AP (for EDCA & MDA): (DIFS,3,7)
STA: (DIFS+aSlot,7,15)
Mesh AP configured to use the same MAC parameters, irrespective of whether MDA is enabled/disabled
Ye Chen, Motorola

9. Statistics Collected Statistics collected form the model include:
March 2007
Statistics Collected
Statistics collected form the model include:
MAC retransmission attempts
Application Layer packet loss
Channel occupancy (load)
Data collected from the model is used to compare the performance of EDCA versus MDA
Ye Chen, Motorola

10. Without Interference EDCA MDA March 2007
Fig 1. Total retransmission attempts (sum) at AP1
Fig 2. Application layer packet loss rate
MDA
Ye Chen, Motorola

11. With Infrastructure BSS Interference
March 2007
With Infrastructure BSS Interference
Impact of interference on EDCA is more pronounced than on MDA
Fig 3. Total retransmission attempt (sum) at AP1 without interference
Fig 4. Total retransmission attempts (sum) at AP1 with infrastructure BSS interference
Ye Chen, Motorola

12. With Interference (Packet Loss)
March 2007
With Interference (Packet Loss)
Impact of interference on EDCA is more pronounced than on MDA
Fig 5. Application layer packet loss rate without interference
Fig 6. Application layer packet loss rate with BSS interference
Ye Chen, Motorola

13. With Interference (Load)
March 2007
With Interference (Load)
Impact of interference on EDCA is more pronounced than on MDA
Fig 7. CCA busy (percentage) measured at AP2 (Portal) without interference
Fig 8. CCA busy (percentage) measured at AP2 (Portal) with BSS interference
Ye Chen, Motorola

14. Perspective of Infrastructure BSS
March 2007
Perspective of Infrastructure BSS
MDA lowers congestion, and therefore Infrastructure BSS enjoys lower number of retransmit attempts
Fig 9. Total retransmission attempts (sum) at Infrastructure BSS
Ye Chen, Motorola

15. March 2007
Summary
Benefits of MDA at least partially offset its implementation complexity
Benefits of MDA do not depend heavily on MAC parameters
MDA can use same MAC parameters as EDCA
MDA is not more susceptible to impact of overlapping BSS interference than EDCA
MDA can be useful in reducing the packet loss rate of voice flows traversing the mesh
MDA can also reduce congestion on mesh links
Ye Chen, Motorola

16. March 2007
Recommendations
Instead of removing MDA from the draft at this time, TGs should continue to improve the feature
It should be possible to resolve comments 534, 536, 1239, 1874, 3549, 3575, 3617 and 5622 by referencing the results presented in this document
Ye Chen, Motorola

17. Excerpts from LB93 Comment Spreadsheet
March 2007
Backup
Excerpts from LB93 Comment Spreadsheet
Ye Chen, Motorola

18. March 2007
Related LB93 Comments
Comment 1874 “I was wondering the benefit of using MDA, but it is not clear to me. The prioritized access and NAV protection have already been defined in e. If there are rationale to introduce this new protocol, I would like to see them in 9.14”
Comment 3549: “MDA seems to be weak in a hidden node scenario commonly observed mesh, because it is a semi-reservation protocol which is known to break.”
Comment 3617: “I have several issues with MDA. Since MDA is optional and only a subset of MPs implement it, while legacy devices also ignore MDAOP, it is unclear that any QoS can be achieved with MDA. When is it beneficial to use MDA? Are there scenarios where its efficacy was demonstrated? While it looks like a good idea to coordinate access between neighbor MPs, I do not believe the MDA as defined achieves this goal..”
Comment 5622 “It is not clear how efficient the MDA mechanism is and how well it performs in the presence of non-MDA STAs. It looks rather complicated and there is no guarantee that it brings performance advantage in the presence of QSTA traffic.”
Ye Chen, Motorola

19. Related LB93 Comments (continued)
March 2007
Related LB93 Comments (continued)
Comment 534 “Suppose the destination of an MDAOP is busy throughout the MDAOP because of transmissions by EDCA MPs or independent WLANs. The sending MP must respect the other MDAOPs and transmit only during time that has not be allotted to MDAOPs. Unless the MDA MPs access the channel with higher priority parameters than EDCA MPs in a mixed mesh, there would be no reason to pursue MDA.“
Comment 536 “ MDA opportunity (MDAOP): It is stated that "The ranges of values allowed for MDACWMin, MDACWMax, and MDAIFSN parameters are identical to that allowed for EDCA." Does this mean the values of these parameters will be the same? Or does it means that the ranges of these parameters will be the same but MDA-capable MPs may access the channel with higher-priority parameters”
Comment 1239: “MDA is redundant and ill-defined. There is no need for another category of access. Furthermore, the describing text is unclear (references to 11e don't make sense when reading the standard as a monolithic document).”
Ye Chen, Motorola

20. Related LB93 Comments (continued)
March 2007
Related LB93 Comments (continued)
Comment 3575: “The draft includes the MDA feature, which is designed to give MP's lower contention in pre-determined time intervals. However, the MDA feature is misguided: - It is only useful at high load, in which case it will be very difficult to calculate a reasonable schedule of MDAOPs, - The schedule could be set up ahead of time, but it is unlikely to be optimal for the traffic load at some point in the future, - The MDA scheme gives MP's priority over regular STA's (from different BSSs in the same channel) which is unfair and unjustified in an unlicensed environment (note: this is a religious argument), - The MDA scheme has massive complexity for an uncertain return.”
Ye Chen, Motorola