Presentation is loading. Please wait.

Presentation is loading. Please wait.

19 June 2002 doc.: IEEE 802.15-02/296r0 Knut Odman, XtremeSpectrum Inc.Slide 1Submission Project: IEEE P802.15 Working Group for Wireless Personal Area.

Similar presentations


Presentation on theme: "19 June 2002 doc.: IEEE 802.15-02/296r0 Knut Odman, XtremeSpectrum Inc.Slide 1Submission Project: IEEE P802.15 Working Group for Wireless Personal Area."— Presentation transcript:

1 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 1Submission Project: IEEE P Working Group for Wireless Personal Area Networks (WPANs) Submission Title: A simplified and unified power save scheme Date Submitted: 1 July, 2002 Source: Knut Odman Company: XtremeSpectrum Inc. Address: 8133 Leesburg Pike, Suite 700, Vienna, VA Voice: , FAX: , Re: Draft P /D10 Abstract:This presentation identifies and solves problems with the current handover process. Purpose:For implementation in 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

2 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 2Submission Problem 1 – Subrate -I SF 1 SF 4 SF 3 SF 2 DEV 1 has a sub-rate every SF CTR DEV 2 wants this GTS every beacon. Is it available?

3 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 3Submission scheduler Problem 1 – Subrate -II GTS1 GTS2ATS1 GTSATS Treat sub-rate as hi-priority ATS Possible but complicated. Interval not guaranteed. superframe What is the purpose?

4 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 4Submission Problem 2 – APS Not related to transmission needs No synchronized wakeup. When should it read PCTM? PNC would have to buffer CTR that may be of no interest once they are allocated. What is the purpose?

5 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 5Submission Problem 3 – SPS -I SF 1 SF 2 GTS1 What does PNC do with a passive GTS? It cannot reallocate it.

6 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 6Submission Problem 3 – SPS - II SF1 SF2 SF5SF4SF3 DEV1DEV2DEV4 How do I broadcast? How does DEV4 send to DEV1? How is handover done?

7 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 7Submission Problem 3 – SPS - III Suspend/ revoke GTS… Join/leave SPS… No global sync… CTRI, SPSI, Active, AWAKE, Sleep… The complexity is prohibitive What is the purpose?

8 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 8Submission Experiences from other protocols Other protocols not using PNC forwarding either have announcement beacons where the PNC wakes up dest. DEV, or announcement slots where the source DEVs wake up their destination. What makes us so special?

9 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 9Submission Experiences from research There is a tradeoff between communication and computation. If the algorithm is distributed, computation gets smaller but the control overhead increases. Rule of thumb: Keep it simple.

10 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 10Submission QoS in the sleep? Dont expect QoS to be perfectly maintained in sleep mode. Responsiveness: Does it matter if you get the slot in 10s or s? Throughput: If you need full QoS, why not wake up, establish a stream, send data, kill stream and go back to bed? The overhead isnt bigger than the current SPS. Jitter: Cant be guaranteed anyway. If it doesnt happen all the time its asynchronous!

11 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 11Submission Solution - I Remove subrate, APS and SPS from standard. Introduce PNC regulated wake beacons. The PNC has a down counter to count down to the next wake beacon. In increased traffic, wake beacons can come more often. DEVs join and leave sleep mode. PNC sets bitmap in all beacons. Use ATS reservation. If dest. in PS, CTA will not come until next (or a following) wake beacon.

12 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 12Submission Destination Dev2PNC asleep Set Dev2 bit in beacon PS bitmap Terminate streams to and from DEV2

13 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 13Submission Source Dev1PNC CTR (DEV2, stream 0) No PCTM needed. The CTA will be there in the wake beacon DEV2 in PS? Enqueue (PSQ, data) Enqueue (AsynchQ, data) truefalse No isochronous data to PS DEVs

14 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 14Submission PNC SF1 SF2 SF5SF4SF3 Wake beacon CTA 1->2 CTA with Dev2 as Destination PNC can chain more than one wake- beacon after each other if needed

15 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 15Submission Destination Dev2PNC awake Reset Dev2 bit in beacon PS bitmap

16 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 16Submission Beacon elements octets : Information IElength = 1-9wake-beacon countdown PS bitmap

17 19 June 2002 doc.: IEEE /296r0 Knut Odman, XtremeSpectrum Inc.Slide 17Submission Thats it! Do we really have to make it more complicated than that?


Download ppt "19 June 2002 doc.: IEEE 802.15-02/296r0 Knut Odman, XtremeSpectrum Inc.Slide 1Submission Project: IEEE P802.15 Working Group for Wireless Personal Area."

Similar presentations


Ads by Google