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: [CCA of UWB channel] Date Submitted: [November, 2005] Source: [Bin Zhen, Huan-Bang Li, Yihong Qi, Ryuji Kohno, Company: National Institute of Information and Communications Technology ] Contact: Bin Zhen Voice:+81 46 847 5445, E-Mail: zhen.bin@nict.go.jp] Abstract: [To enable CSMA in the 15.4a network ] Purpose: [MAC] 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. CCA of UWB channel Bin Zhen, Huan-Bang Li, Yihong Qi and Ryuji Kohno

3. Motivations • Although 4a nodes are low duty cycle, the network load is not light. Traffic in sensor network is usually burst style. • Inactive period • Pending data from coordinator to multiple nodes • An event is usually sensed and reported by multiple sensors • The UWB spectrum mask of EU and Japan requests interference mitigation technology, e.g. DAA. • To follow the architecture of 15.4

4. Basic ideas • Because of low radiation power, sparse and short nature of UWB pulse, CCA of UWB channel is difficult. • Frame preamble is of some regular structure. But the data portion is pretty random. • Data portion is hard to be sensed • To enable CCA at any time, we must introduce regular structure to the whole UWB frame • Superimposed preamble

5. CCA using preamble • After multipath channel, the periodicity of preamble still remains. • This enables time average processing without synchronization • Spreading gain of preamble pluses another benefit

6. CCA using preamble (cont.) Original spectrum Polluted spectrum after channel Real experimental data: 30.875MHz PRF, 31 bit ternary code

7. CCA using preamble (cont.) Preamble chip buried in the noise Preamble chip after processing Noise background Cable connected Indoor channel, 10 meters

8. Superimposed preamble • Continue preamble using spread code S1 during data portion which using other orthogonal spread codes • CDMA style multiplexing • Whenever performing carrier sense using spread code S1, preamble can be sensed Ternary code S2 Ternary code S3 Ternary code S1 Header Payload Channel Preamble + Preamble Ternary code S1 Total frame duration

9. Reference protocols • Pure ALOHA • Preamble-only CSMA • Assume only the frame preamble can be sensed. The data part is sensed as free channel. • FSK-style CSMA • Continue to sense the media until preamble is found or the end of packet duration • All-frame CSMA • Idea case where all frame can be sensed as CCA in narrow-band system

10. 6 preamble-only preamble-only all-frame 5 0.25 all-frame FSK FSK aloha aloha 4 0.2 superimposed superimposed power/good-transmission 3 normalized payload throughput 0.15 2 0.1 1 0.05 0 0 -1 0 10 10 -1 0 10 10 offered load offered load Performance analysis 0.5ms preamble, Tcs=0.032ms, 1Mbps, 32 bytes data CS spends the same power as transmission No CS error

11. 3.5 1 preamble-only all-frame 3 FSK 0.8 aloha 2.5 superimposed 0.6 normalized payload throughput normalized power/good transmission 2 preamble-only 0.4 1.5 all-frame FSK 0.2 aloha 1 superimposed 0.5 0 -1 -1 0 0 10 10 10 10 offered load Normalized performance • After inserting preamble into payload portion, network payload throughput increases • Even in light network load, ALOHA and preamble-only CSMA spends more power than superimposed CSMA

12. 6 all-frame all-frame 5 0.25 aloha aloha superimposed superimposed superimposed 4 superimposed 0.2 power/good-transmission 3 normalized payload throughput 0.15 2 0.1 1 0.05 0 0 -1 0 10 10 -1 0 10 10 offered load offered load Performance with CCA error CCA probability= 0.8 CCA probability= 0.95 CCA probability= 0.95 CCA probability= 0.8 • With only 80% detection probability, the maximal through is doubled compared with that of ALOHA

13. Conclusions • Pure ALOHA gives worst performance. It only works when offered load is less then 0.2. • Only sensing the frame preamble is not enough • Performance of FSK style CSMA is poor • Sensing the whole UWB packet is possible by adding regular structure • Superimposed preamble • Even with CCA error, network performance is still better than ALOHA • Benefits • PHY layer implementation • Good throughput and low power consumption in heavy network load.