Date Submitted: November 11, 2005] Project: IEEE P802.15 Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [Numerical results on the ranging packet with superimposed preamble] Date Submitted: November 11, 2005] Source: [Yihong Qi, Huan-Bang Li, Shinsuke Hara, Bin Zhen and 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: [Numerical results using the ranging packet with superimposed preamble are presented. These results include extra energy consumption, ranging accuracy improvement and carrier sense probability.] Purpose: [To propose a new ranging packet structure] 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. TG4a

Numerical results on the ranging packet with superimposed preamble Yihong Qi, Huan-Bang Li, Shinsuke Hara, Bin Zhen and Ryuji Kohno National Institute of Information and Communications Technology (NICT) TG4a

Outlines Review of the current ranging packet Problems with the current packet Low throughput with CSMA or ALOHA No flexibility for a variety of ranging applications Review of the proposed packet with superimposed preamble Numerical results with the proposed packet Carrier sense (CS) probability Ranging accuracy improvement Extra energy consumption TG4a

Current ranging packet Preamble Header Payload Data on timing, crystal offset, etc Acquisition; channel sounding Control information contents modulation length average PRF preamble Acquisition; channel sounding Ternary code (determined) three lengths: 500us, 1ms, 4ms (proposed) same (determined) header Control information TBD payload Data on timing, crystal offset, etc TG4a

Problem with ranging cf: current solution (?) A variety of ranging applications require tradeoff among Ranging accuracy VS. Mobility Update rate On-air time cf: current solution (?) three preamble lengths: 500us, 1ms and 4ms. TG4a

Problem with ALOHA Low throughput From Bin Zhen’s Doc 619r0 10 -1 1 0.05 0.1 0.15 0.2 offered load normalized payload throughput preamble-only CS whole packet CS pure aloha Low throughput From Bin Zhen’s Doc 619r0 1ms preamble, 0.25ms data(1Mbps, 32 bytes data) TG4a

The proposed ranging packet Preamble + (code multiplexing) Preamble Header Payload Preamble is superimposed in the header and payload section. TG4a

An Illustration of the proposed packet power level Preamble section Header and payload sections s2 s2 s2 s1 s1 s1 s1 s1 superimposed preamble s1 t s1: spreading code for preamble symbol; s2: spreading code for data symbols TG4a

Key point 1 The power level of the superimposed preamble is much lower than that of the regular preamble and data sequences. For minimizing the resulting interference to data demodulation and SOP. Power reduction of 7~15dB examined in the numerical results TG4a

Key point 2. The superimposed preamble is constructed by repetition of one preamble symbol s1. Hence, potential ``spreading gain” is high. E.g., with symbol duration 1us and the payload section of 256us, ``spreading gain”=14dB. TG4a

Advantages Extra preambles for channel sounding/ranging is available. Improve ranging accuracy Provide flexibility to an application to determine its tradeoff among ranging accuracy, mobility, update rate, etc. Superimposed preamble embeds ``regularity” in the header and payload section. CS (carrier sense) of an entire packet is possible, hence can improve the throughput performance. TG4a

Numerical Results TG4a

Parameters and assumptions Pulse: Bandwidth 500MHz with duration 2ns, Peak PRF= 494MHz; average PRF=16MHz Preamble: Ternary code with length 31, symbol duration=1us Total length: 0.5ms, 1ms, 4ms Data Modulation: 2 PPM + BPSK, Walsh code with length 8, symbol duration=0.5us Data length: 256us Assumptions: The power of the regular preamble and data sequences of the header and payload sections is same. The power of the superimposed preamble is 7~13dB lower than the regular level. TG4a

Ratio of extra energy consumed by the superimposed preamble Observation: Extra energy consumption is low, less than 5% for power reduction of less than 8dB. Power reduction = pulse energy of data / pulse energy of the superimposed preamble. TG4a

An illustration of the extra energy consumption Regular preamble Header and payload Less than 5% Superimposed preamble 0.5, 1, 4ms 0.25ms TG4a

Improvement in ranging accuracy TG4a

How to approximate the ranging accuracy improvement Assumption: Coherent ranging estimate the variance of a ranging estimate proportional to 1/SNR. TG4a

Improvement in ranging accuracy L= number of symbols used for channel sounding/ranging in the preamble section. Observation: Accuracy improvement around 40~70% is available due to small number of symbols actually used for channel sounding in the regular preamble section. TG4a

Rationale of the accuracy improvement 0.5, 1, 4ms 0.25ms Regular preamble data Superimposed preamble Small number of symbols in the regular preamble section is used for ranging. TG4a

Carrier sense (CS) probability TG4a

A CS scheme Receiving a signal within certain time window Matched filter to the ternary code S BPF ``de-spreading” ``Energy detection” CS is to use the energy detector to determine whether the received signal contains a preamble structure. The energy detector is to test whether the computed energy is over certain threshold or not. TG4a

Key points of the CS scheme The preamble is a periodic signal with period Ts1, where Ts1 is the duration of the preamble symbol s1. After passing a multipath channel, the received preamble is still a periodic signal with period Ts1. regardless symbol synchronization, regardless multipath distortion The CS scheme exploits the ``spreading gain” of the periodic received signal. TG4a

Successful CS probability with Eb/N0=10dB and false alarm around 5% CS window=16us Observation: Successful CS probability is over 70% for power reduction of less than 10dB. TG4a

Successful CS probability with Eb/N0=10dB and false alarm around 5% CS window=25us Observation: Successful CS probability is over 90% for power reduction of less than 10dB. Power reduction = pulse energy of data / pulse energy of the superimposed preamble. TG4a

Including crystal offset Analytically For CS window of 16us, crystal offset 40ppm corresponds to timing error 0.64ns, i.e., 30% of pulse duration 2ns For CS window of 25us, crystal offset 40ppm corresponds to timing error 1ns, i.e., 50% of pulse duration 2ns Limited effect on CS performance TG4a

Successful CS probability with Eb/N0=10dB, false alarm around 5% and +40ppm crystal offset CS window=16us Observation: Successful CS probability is over 70% for power reduction of less than 10dB. TG4a

Successful CS probability with Eb/N0=10dB, false alarm around 5% and +40ppm crystal offset CS window=25us Observation: Successful CS probability is almost over 80% for power reduction of more than 10dB. TG4a

On non-coherent issue The proposed packet has little interference on the current non-coherent ranging and data detection because of Quasi-orthogonal between the payload and the superimposed preamble codes Power of the superimposed preamble is sufficiently reduced. TG4a

Conclusions The proposed packet with superimposed preambles of power reduction 10dB Low interference to data detection and SOP Low extra energy consumption of less than 5% Good ranging accuracy improvement of 40%~70% Effective CS performance, success probability 85~95% with CS window of 25us and crystal offset 40ppm No confliction with the non-coherent reception TG4a