1. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a1 Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [Two considerations on TWR] Date Submitted: [July 1, 2005] Source: [Yihong Qi, Huan-Bang Li, Ryuji Kohno, Company: National Institute of Information and Communications Technology ] Contact: Yihong Qi Voice:+81 46 847 5092, E-Mail: yhqi@nict.go.jp] Abstract: [TWR with crystal offset management can effectively improve the ranging accuracy. A separate mode is necessary for precision with mobility] Purpose: [To propose a separate mode for ranging and communications] Notice:This document has been prepared to assist the IEEE P802.15. 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 P802.15.

2. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a2 Two Considerations on TWR Yihong Qi, Huan-Bang Li, Ryuji Kohno National Institute of Information and Communications Technology (NICT)

3. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a3 Outline In TWR with crystal offset tracking (Vern, 05-336r0), knowing an offset difference only is sufficient to reduce the error due to crystal offset. Regarding communication and ranging, a separate mode is necessary for precision ranging with mobility.

4. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a4 Device ADevice B Two equations in two unknowns yield: * US Naval Observatory, Telstar Satellite, circa 1962 http://www.boulder.nist.gov/timefreq/time/twoway.htm Unmatched detect-delays in the two devices may require one-time offset calibration. Unknown propagation delay Unknown clock offset Message 1 Message 2 Multiple measurements of t p and t o yield finer precision & accuracy, and allow frequency offset correction. Figure 1 Two-Way Time Transfer Model from 04-0581r07.

5. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a5 Let ’ s be specific. Here is the error due to crystal offsets. round trip time seen at A Since the turn-round time (e.g., 2ms) is greatly larger than the propagation delay (e.g., 150ns), the round trip time and turn-round time is almost equal. turn-round time seen at B crystal offset at A crystal offset at B

6. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a6 TWR with offset tracking (Vern, 05-336r0) AB AB A measures B’s oscillator drift here B embeds his measurement of A’s oscillator drift as a number in the data. (along with turn-around time) Initiation message Response message

7. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a7 TWR with offset tracking Crystal offset between A and B is tracked. How about the error due to crystal offset now? before: now: since is known It works!

8. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a8 Communication and Ranging A parallel mode (current) With a common header, communication and ranging are done in parallel. A separate ranging mode( to be discussed) When precision ranging with mobility is taken into account, a separate ranging mode is necessary.

9. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a9 Ranging time due to mobility vs. header length e=ranging error, v=node speed, a=maximum percentage of error increase due to the mobility, e.g., a=10%, e=25cm, total error (1+a)e=27.5cm E.g., v=10m/s, a=10%, e=25cm  T<2.5ms Header length for ranging is also on the order of millisecond, 1~ 4ms. Ranging time due to mobility can be on the order of millisecond  No time for communication data!

10. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a10 What are our options for ranging with high mobility? Using shorter preamble  lower ranging accuracy A separate ranging mode Using time-saving ranging methods –TWR with crystal offset tracking –TDOA

11. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a11 Regarding a separate ranging mode A separate ranging mode for long distance ranging (50m) when a node speed is larger than certain threshold, e.g., 3m/s. Otherwise, a parallel ranging and communication mode

12. July, 2005 Doc: IEEE 15-05-0375-01-004a Qi, Li, Kohno (NICT) SlideTG4a12 Advantages of a separate ranging mode Flexibility: data packets and data updating period are designed separately to meet different requirements of ranging and communications. E.g., Communication may use the channel information obtained in the ranging mode. header lengthData updating period communicationsshortlong ranginglongshort