Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs)

1. Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs) Submission Title: TG4a – Discussion on UWB Low Data Rate Date Submitted: May 17, 2006 Source: Gideon Kaplan (Sandlinks Systems) Contact: Gideon Kaplan Voice: +972-54-232-1570, E-Mail: gidi@sandlinks.com Abstract: Analysis of the low data rate option Purpose: To provide a basis for further discussion on the merits / shortcomings of the low rate option, and present a suggestion to delete it from the draft 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. Gideon Kaplan (Sandlinks Systems)

2. Low Data Rate in 15.4a - Discussion Gidi Kaplan IEEE, Jacksonville meeting May 06 Gideon Kaplan (Sandlinks Systems)

3. Agenda • Preface • Mandatory rate vs. Low rate • Capacity at low rates • Is it a low cost radio? • Summary & Conclusions • Acknowledgement for the contribution of Ben Rolfe and Dani Raphaeli. Gideon Kaplan (Sandlinks Systems)

4. Preface • The discussion is related to the operation of UWB (not CSS) • The main data rate in the UWB bands is 0.8Mbps • In many applications, the proposed system, working at such rate with the 15.4 protocol, yields adequate performance in terms of link margin, and network capacity • Why low rates are needed at all? Because in some applications there is a need for extended link margin or for extra “processing gain” • It is indeed not a trivial task to design a wireless system with a very wide range of data rates (like we are trying to accomplish here) • Here, we focus on the question if operation at 100Kbps (the optional low rate) is practical and useful according to the current draft. Gideon Kaplan (Sandlinks Systems)

5. Mandatory Rate vs. Low Rate – a simple approach • Basically, a “mandatory rate” definition (in a system with multiple rate options) dictates that all the network elements start their session at this rate • Only then it is possible to move to any other rate • This approach is simple and is good for interoperability. • In terms of link margin, it makes sense only if you start at the lowest rate – as it has the highest margin • This is not the case in the Beacon design of the 15.4a system • The Beacon mandatory rate is 0.8Mbps • The lowest rate is 0.1Mbps. • So in a system using Beacons – which is usually the better way to achieve overall efficiency – it is not clear what is the purpose of the low rate, if anyway the Beacon signal(s) work at 0.8Mbps. Gideon Kaplan (Sandlinks Systems)

6. An idea for consideration • Consider an idea to operate the Beacon at 0.8Mbps and 0.1Mbps alternatively • We degrade the Beacon’s capacity at the mandatory rate, but allow a better option for low-rate operation • Let’s see if this merits discussion… Gideon Kaplan (Sandlinks Systems)

7. Multi-rate Approach for Peer to Peer Connectivity • In the 15.4a draft, some attention has been given to the above concern • The SHR and PHR preambles now have the possibility to work at lower rates, in peer-to-peer communication • A multi-rate receiver will be capable to detect that the transmission is in low rate • But…one can not assure interoperability any more • There is an added concern about simultaneous operation of two Piconets, one at 100Kbps and the other at 800Kbps. Gideon Kaplan (Sandlinks Systems)

8. Capacity at Low Rate • Suppose one configures the network to operate at 0.1Mbps, in order to achieve maximum distance margin • What is the capacity that the network achieves? • Without CCA, the network operates like an un-slotted Aloha, and will practically achieve not more than ~15Kbps • This is really an awfully low result • Imagine what will happen if the network has to communicate 1000 packets of 60 bytes each – this will take about 40 seconds ! • Even with CCA, I believe that the network will not achieve more than ~50Kbps • This means that the low data rate is really not useful (using 500Mhz to achieve an aggregate rate of X0 Kbps seems like bad engineering). Gideon Kaplan (Sandlinks Systems)

9. Performance / Low Cost Implications • Recall that the main goal of using a low data rate is getting a good distance margin • Furthermore, the purpose of 802.15.4a is to create low-cost Radios • However, the current scheme for low data rates, requires a very good clock, in order to coherently integrate all the UWB pulses in a 10-usec long symbol (at 0.1Mbps rate) • The clock will have a tough requirement both in its accuracy – probably a 5 ppm accuracy will be needed, as well as excellent phase noise • Practically, one will have to sacrifice performance, in order to be able to work with a reasonable (low cost) crystal source, for example – use a non-coherent receiver; So – did we get our original goal?... Gideon Kaplan (Sandlinks Systems)

10. A remark on High Data Rates • The draft includes an option for operation at 27Mbps • It seems that if one considers the preamble lengths of a practical packet, the effective transmission speed will be less than half of this value • It will have a limited value in “getting the packet off the air fast” • It will enable communication only at very short distances. • Food for thought: in the 15.4a space of applications, isn’t 6.8Mbps fast enough?... Gideon Kaplan (Sandlinks Systems)

11. Summary • The current air interface suggested is suitable to medium data rates • However the network design DOES NOT PRACTICALLY SUPPORT THE LOW DATA RATE MODE • This has been shown via several angles of view • At this stage, after so much thought have been invested in the system, it is really hard to find a substantial remedy • The suggestion is to eliminate this mode from the current draft • This will open a future possibility of an alternative suggestion for real “low rate UWB systems” – say, in the range 40Kbps to 250Kbps • This author is ready to participate in a new task group that will target this specific range. • The current standard, in my view, is better and more focused if it specifies 0.8Mbps as the main rate and 6.8Mbps as the optional rate, in the UWB bands. Gideon Kaplan (Sandlinks Systems)