TDOA Ranging for Tracking Nodes in 802.15.4a

1. Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [TDOA Ranging] Date Submitted: [1 May, 2005] Source: [Vern Brethour] Company [Time Domain Corp.] Address [7057 Old Madison Pike; Suite 250; Huntsville, Alabama 35806; USA] Voice:[(256) 428-6331], FAX: [(256) 922-0387], E-Mail: [vern.brethour@timedomain.com] Re: [802.15 4a.] Abstract: [TDOA ranging protocols are proposed for adoption by 802.15.4a.] Purpose: [To introduce protocols for tracking (positioning) large numbers of nodes using 802.15.4a.] 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. Brethour, Time Domain

2. TDOA Ranging Service for 802.15.4a Proposed to support computing the position of large numbers of nodes. Brethour, Time Domain

3. First; we review the vocabulary established in 15-04-0581-05-004a(The Ranging Subcommittee Report) • TOA Ranging • TDOA Ranging mode 1 (SOI is Rx) • TDOA Ranging mode 2 (SOI is Tx) TOA = Time of Arrival; TDOA = Time Difference of Arrival; SOI = Station of Interest Brethour, Time Domain

4. Positioning from TOA 3 anchors with known positions (at least) are required to retrieve a 2D-position from 3 TOAs Anchor 2 (xA2,yA2) Anchor 1 (xA1,yA1) Mobile (xm,ym) Anchor 3 (xA3,yA3) Estimated Position Measurements Specific Positioning Algorithms TOA Ranging from 04-0581r5 Brethour, Time Domain

5. Characteristics of TOA Ranging • Each range involves a two way (at least) message sequence. • The previous slide showed “anchors”, but in general, for relative positioning, there do not need to be anchors. • When n nodes range to all other n nodes the number of two way exchanges is [n*n + n] / 2. Brethour, Time Domain

6. [n*n + n] / 2 Yucko! • We hate it when we get an n squared term in our transaction count equation. • Even for just a dozen nodes to do a full 2 way exchange between all nodes takes 78 two way exchanges. Brethour, Time Domain

7. What is required of the nodes for TOA ranging? • All nodes need to send and receive. • All nodes must be capable of channel sounding upon receive, • None of the nodes need to keep track of any other nodes’ clock. • In general, TOA is the easiest kind of ranging to support. Brethour, Time Domain

8. What about TDOA ? Two modes from 04-0581r5 to consider. • Mode 1 – The station of interest (SOI) receives multiple reference packets and calculates the TDOA • LORAN-C type operation and the processing burden is on the receiver to run the hyperbolic location algorithms • Mode 2 – The station of interest transmits a reference packet which is received by multiple fixed nodes • The fixed nodes must forward the TDOA information to a central node which then runs the hyperbolic location algorithms reference node reference node SOI SOI Key: Sync Pulse Location Pulse Position Report Key: Sync Pulse Location Pulse TDOA backhaul controller controller Mode 1 - Passive Mode 2 - Active Brethour, Time Domain

9. Let’s look at TDOA mode 2 first. • There are anchor nodes and stations of interest. • Anchor nodes must be able to learn their position (for example by using TOA ranging). • Anchor nodes must additionally be able to synchronize their notion of absolute time. • Stations of Interest do NOT need to ever do a channel sounding. They might only transmit, when it’s their turn, if they do not care about their position. • The number of transactions on the air is one per Station of Interest. (No more n squared terms!) Brethour, Time Domain

10. Now onto TDOA Mode 1. • Again, there are anchor nodes and stations of interest. • Anchor nodes must be able to learn their position (for example by using TOA ranging). • Anchor nodes must additionally be able to synchronize their notion of absolute time. • Stations of Interest must be able to do a channel sounding. They also must synchronize (with offset is okay) to the Anchor’s clock. They might only receive if the SOI does its own solver work. • Number of transactions on the air is only a function of the number of anchors. (No more n terms at all!) Brethour, Time Domain

11. Review: How many one way transmissions to keep track of n SOIs ? • With TOA it’s {n squared plus n}. • With TDOA Mode 2 it’s {n}. • With TDOA Mode 1 it’s {independent of n}. Brethour, Time Domain

12. Let’s compare the SOIs • With Mode 2, the SOI can be very simple. A non-coherent receiver might be adequate. There is on need for an on-board solver. • With Mode 1, the SOI must be as capable as an anchor device. Brethour, Time Domain

13. Let’s compare Anchor Devices • With TOA Ranging, there is no need to have the nodes knowing about how to synchronize clocks. • With TDOA Ranging, the Anchor nodes must have all the capability that TOA nodes have, plus more functionality to establish and maintain synchronized clocks. Brethour, Time Domain

14. What’s so attractive about TDOA Mode 2? • Really Cheap tracked nodes! • This is a great way to get high accuracy positioning of nodes that only have non-coherent receivers. • Supports a very large number of tracked nodes. (on-air burden only grows as n). Brethour, Time Domain

15. We like really cheap tracked nodes? • You Bet! • Warehouse applications can tolerate a fixed set of anchor nodes and those nodes can be more expensive and can even be powered by electricity from a power grid. • Warehouse applications want very cheap tracked nodes that run a very long time on small batteries. Brethour, Time Domain

16. What about TDOA Mode 1? • That’s the technique that needed a tracked node device that is as capable as the anchor node. • The only real advantage for this mode is that it can scale to a super huge number of tracked nodes (no time-on-air issues). • This mode seems less compelling. Brethour, Time Domain

17. Should we support TOA Ranging? • Absolutely! • No need for designated Anchor nodes. • No need for clock synchronization. • The easiest and most versatile way to do positioning for an ad-hoc configuration of a handful of nodes. Brethour, Time Domain

18. Should we support TDOA Mode 2 ranging? • Yes. • It scales to large node count without terrible time-on-air issues. • The super cheap tracked node will allow us to address positioning markets we may not be able to serve with TOA only. Brethour, Time Domain

19. Should we support TDOA Mode 1 ranging? • Maybe (?) • The super high node count scalability might prove useful for something? • The cost of all nodes will come down with time and 15.4a volume success. • It may not cost us much in terms of hardware development: Most of the “stuff” we need will have to be developed anyway, if we support both TOA and TDOA mode 2. Brethour, Time Domain

20. Recommendations: • Decide now to develop the 15.4a ranging standard to support BOTH TOA and TDOA mode 2 ranging protocols. • Keep our eyes open, and decide later to support TDOA mode 1, if it appears that we’ll have it “almost for free” anyway. Brethour, Time Domain