June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a1 Project: IEEE P Working Group for Wireless Personal Area Networks (WPANs) Submission Title: [Simulations on non-coherent ranging] Date Submitted: [19 June 2005] Source: [Ismail Guvenc, Zafer Sahinoglu, Mitsubishi Electric] Contact: Zafer Sahinoglu Voice:[ , Abstract: [This document provides statistical analysis of IEEE a channels’ energy distribution] Purpose: [To provide input for optimization of system design parameters] Notice:This document has been prepared to assist the IEEE P 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 P

June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a2 Outline Channel energy characteristics Threshold statistics for CM1 and CM2 TOA estimation –Mean Absolute Errors

June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a3 Channel Energy Characteristics Unit energy 1ns raised cosine pulse is passed through CM1 (LOS) and CM2 (NLOS) channels Received signal energy is collected in 4ns integration intervals Statistical delay and amplitude characteristics are derived after averaging over 1000 realizations

June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a4 CM1 –PDF of the delay between the first energy block and strongest energy block is exponentially distributed (90% of the time less than 25ns) –Highest energy can reach 0.9, but at a very small probability It is less than 0.6 at 90% confidence level

June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a5 CM2 –PDF of the delay between the first energy block and strongest energy block is exponentially distributed (90% of the time less than 25ns) –Highest energy can reach 0.45 It is less than 0.25 at 90% confidence level

June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a6 CM4 –PDF of the delay between the first energy block 90% of the time less than 22ns –Highest energy can reach 0.6 It is less than 0.22 (at 90% confidence level)

June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a7 CM8 –PDF of the delay between the first energy block 90% of the time less than 115ns –Highest energy can reach It is less than at 90% confidence level

June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a8 Non-Coherent Receiver Signal processing techniques are applied onto the outputs Z[n] of the ADC to find the first energy arrival BPF( ) 2 LPF / integrator ADC 2-4ns Z [n]

June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a9 Optimum Normalized Thresholds Normalized optimum threshold selection for “first threshold crossing based TOA”

June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a10 Mean Absolute Error Even at optimum threshold settings, when threshold crossing is applied, the MAE is in the order of nanoseconds Complex algorithms are needed to identify the first energy block –In the case of SOP interference, it becomes more challenging CM1CM2 EBN0: {8, 10, 12, 14, 16, 18, 20, 22, 24, 26}

June 19nd, 2005Doc: IEEE a I. Guvenc, Z. Sahinoglu, Mitsubishi Electric SlideTG4a11 Mean Absolute Error Single 4ns unit energy pulse in each frame No IFI Search back with threshold crossing based TOA tracking The MAE is lower bounded by 3-4ns There are very large outliers that increase the MAE 90% confidence level for 2ns error at 4ns integration would not be achievable even at very high SNRs