Presentation is loading. Please wait.

Presentation is loading. Please wait.

<January 2002> doc.: IEEE <02/139r0> March, 2008

Similar presentations


Presentation on theme: "<January 2002> doc.: IEEE <02/139r0> March, 2008"— Presentation transcript:

1 <January 2002> doc.: IEEE <02/139r0> March, 2008 Project: IEEE P Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [A Modified Aggregation Process] Date Submitted: [March 16, 2008] Source: [Kyungsup Kwak, Seokho Kim, Xizhi An] Company: [Inha University] Address: [6-141B, Inha University, 253 Yonghyun-dong, Nam-gu, Incheon, , Republic of Korea] Voice: [], FAX: [], Re: [] Abstract: [This document gives some comments on aggregation process in current draft standard. In order to make “3c” standard more compatible with “3” and “3b”, we propose a modified aggregation process that is transparent to MAC layer.] Purpose: [To be considered in IEEE c standard] Notice: This document has been prepared to assist the IEEE P 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 acknowledges and accepts that this contribution becomes the property of IEEE and may be made publicly available by P Inha Univ. Chuck Brabenac, Intel Labs

2 Overview Review of Aggregation Process
March, 2008 Overview Review of Aggregation Process Current Draft (DF1): P c-DF1_Draft_Ammendment.pdf New Proposal of Aggregation Process Modified Aggregation Process New PHY frame formatting Inha Univ.

3 Review of Aggregation Process in DF1
March, 2008 Review of Aggregation Process in DF1 Incompatible with & 3b In section 8.7a, subframe is confused with MPDU. In fact, the fragmentation is the same as section 8.7 Blk-ACK is unnecessary. In fact subframe just is MPDU and ACK is done on the MPDU basis, so Dly-ACK can do this job. Subheader is similar to MAC header, however fragmentation control can not work. Functions of different layers are mixed up! Aggregation shall be realized in PHY layer, “Number of subframes” field is in PHY header. PHY frame is like a kind of “container”, which conveys MPDUs. PHY only passes frames to the MAC that have passed the HCS test. Fragmentation and ACK shall be performed in MAC layer. When MSDU/MCDU is fragmented, the fragmentation control information for each fragment must be set correspondingly. The MAC validates the error free reception of every frame by using FCS. Inha Univ.

4 Current DF1 Section 8.7a Aggregation
March, 2008 Current DF1 Section 8.7a Aggregation Upon receiving an MSDU from FCSL, the MAC may fragment the MSDU into subframes if the length of MSDU exceeds the pPreferred-FragmentSize as described in <xref >. The PHY puts the subframes together, as described in <xref >, to form an aggregated frame that is sent as one PHY transmission over the medium. The number of subframes accommodated in a single aggregated frame is contained in the PHY header as described in <xref >. Inha Univ.

5 Current DF1 12.2.11.2 Supported fragment size encoding
March, 2008 Current DF1 Supported fragment size encoding The encoding of the preferred fragment size for SC mode used in the Capability IE Inha Univ.

6 Current DF1 12.2.4.1.10 Number of subframes
March, 2008 Current DF1 Number of subframes The default setting of the Number of Subframes field shall be 0000 to indicate a non-aggregated frame. When not set to zero, the Number of Subframes field shall indicate the number of subframes minus one. The maximum allowed number of subframes is 16. For an aggregated frame, the frame header shall contain the base header as well as the optional header. Inha Univ.

7 March, 2008 Current DF1 Inha Univ.

8 Current DF1 Fixed field: MPDU-specific: March, 2008 ACK policy
PNID, DestID, SrcID, Stream index MPDU-specific: Frame control / retry = retransmission policy MSDU number = MSDU number Fragment number = Subframe ID Useless? FCS information Inha Univ.

9 Review of Aggregation Process in DF1
March, 2008 Review of Aggregation Process in DF1 Prerequisites of Aggregation ACK policy shall be Dly-ACK or No ACK. All aggregated MPDUs have the same PNID, DestID, SrcID and Stream index. MAC header in PHY frame One fragment control field in MAC header is not enough for multiple fragments. Fragment, MPDU and MAC frame have the same meaning. Subheader shall contain fragment control field for corresponding fragment. Inha Univ.

10 Review of Aggregation Process in DF1
March, 2008 Review of Aggregation Process in DF1 FCS information is useless and shall be removed Only Imm-ACK frame has no frame body, and thus no FCS. All other frames must have FCS. It is the task of MAC layer to decide the existence of FCS. Subframe ID is useless and its function shall be changed. The “Number of subframes” field in PHY header and “Subframe length” in subheader can assure that the aggregated MPDUs can be perfectly de-aggregated. Subframe ID has its meaning in PHY layer, however it is used to realize some functions of ACK and fragment which are in fact tasks of MAC layer. It shall be changed to support fragmentation control that is missed in current aggregation process. Instead, the “fragment number” shall be used and the length of this field shall be 7 bits. Moreover, a whole “Fragmentation control” (3 octets) shall be added in subheader. Inha Univ.

11 Modified Aggregation Process
March, 2008 Modified Aggregation Process Aggregation is one of PHY functions. Transparent to MAC layer Fragmentation control field of every MPDU is maintained. Inha Univ.

12 March, 2008 Current DF1 MAC performs fragmentation, so the fragmentation control field is necessary to every fragment, but there is only one MAC header in the flowchart. The size of MSDU block is misleading. Inha Univ.

13 March, 2008 Current DF1 Inha Univ.

14 New Proposal of Aggregation at source
March, 2008 New Proposal of Aggregation at source Inha Univ.

15 New PHY frame formatting of Aggregation
March, 2008 New PHY frame formatting of Aggregation Inha Univ.

16 MAC Header and Subheader
March, 2008 MAC Header and Subheader Method 1: “Common” MAC header “Common” MAC header contains PNID, DestID, SrcID and Stream index. The Fragmentation Control field is useless. Each subheader contains fragment information. There are 1 MAC header and n subheaders. n: number of subframes Method 2: MAC header of 1st MPDU MAC header is exactly the MAC header of 1st MPDU to be aggregated in one PHY frame. Subheaders are the simplified MAC headers of the rest MPDUs that are aggregated after the 1st one. There are 1 MAC header and (n – 1) subheaders. Note It should be consistent with PHY frame without aggregation. Inha Univ.

17 Summary of Aggregation Process
March, 2008 Summary of Aggregation Process Aggregation shall be regarded as a PHY enhancement Rearrange section 8.7a to section 12 PHY frame format Frame Header MAC header and subheader extraction MAC header compression, a new term? Transparent to MAC layer: make clear the concept of layered protocol stack It is the whole MPDU (not so-called “subframe”) that passes through the PHY SAP. Compatible with & 3b: Fragmentation control is necessary to every MPDU More suitable to Dly-ACK procedure MPDU ID in Dly-ACK frame just is the fragmentation control information. ACK the aggregated MPDUs Burst size = n x number of MPDU aggregated in one PHY frame Inha Univ.

18 MAC header compression
March, 2008 MAC header compression PHY gets multiple MPDUs via PHY SAP. PHY separates MAC header from frame payload (with FCS). PHY converts multiple MAC headers into one (common) MAC header and several subheaders. Subheader has less number of octets than MAC header. MAC header decompression just is the reverse process. Inha Univ.

19 March, 2008 Appendix Inha Univ.

20 March, 2008 c-resolutions-to-comments-discussed-at-taipei.pdf Inha Univ.


Download ppt "<January 2002> doc.: IEEE <02/139r0> March, 2008"

Similar presentations


Ads by Google