Investigations on Correlation Properties of Ultra-Wideband Radio Channels S. Bieder A. Czylwik L. Häring A. Alshabo p. 1
Outline Introduction Measurement Principle Measurement Equipment Verification of Measurements Evaluation of Measurements Measurement Scenarios and Parameter Measurement Results Summary and Future Works p. 2
Introduction Maximum transmission data rate is limited by the channel capacity UWB channel capacity strongly depends on the frequency correlation of channel transfer function (CTF) Frequency correlation of channel transfer function is not known for channels with large bandwidths Investigated tasks: Measurement campaign to determine the transfer function of UWB channels in different scenarios Determination of the correlation function (CF) of the CTF Evaluation of CF to determine the coherence bandwidth p. 3
Measurement Principle Frequency domain based measurements: UWB Antennas Network Analyzer Antenna distance Network analyzer is used to measure scattering parameter of the device under test (DUT) DUT includes effects of: Transmit antenna Mobile propagation channel Receive antenna p. 4
Measurement Equipment Device Manufacturer Type ZVB 20 Network Analyzer Rohde & Schwarz Transmit antenna Receive antenna Institut für Höchstfrequenztechnik und Elektronik University of Karlsruhe, Prof. Wiesbeck Monocone antenna Phase stable measurement cable Rosenberger Micro-Coax UFA 210 db p. 5
Verification of Measurements Frequency domain based measurements result in a high measurement duration (approx. 30 min) Measurements must be carried out in time invariant scenarios Verification idea: Perform 2 measurements in a row for each scenario Calculate the normalized mean absolute difference of the two results Time invariant scenarios small values of Time variant scenarios large values of p. 6
Evaluation of Measurements 1/4 Interpretation of the measured scattering parameter as the effective channel transfer function Determination of the normalized frequency correlation function of the CTF : Reference frequency Distance in frequency p. 7
Evaluation of Measurements 2/4 Approximation of the expectation value by averaging over the frequency: with: Frequency resolution of the measuring points Bandwidth used for averaging p. 8
Evaluation of Measurements 3/4 High influence of the distinct realization of the fast fading of the channel transfer function on the correlation function Reducing the influence: Performing 9 sub-measurements for each scenario Slightly different position of Tx and Rx antenna for each submeasurement The distance of Tx Rx antenna is kept constant Basic characteristic of the CTF remains constant Tx antenna position of the 9 submeasurements 1 2 3 4 5 6 7 8 9 distance of Tx Rx antenna 1 2 3 4 5 6 7 8 9 Rx antenna position of the 9 submeasurements Averaging of the 9 different correlation functions p. 9
Evaluation of Measurements 4/4 For each scenario the frequency dependent coherence bandwidth of the channel transfer function is calculated Definition of the coherence bandwidth: Maximum frequency spacing for which the absolute value of the correlation function still is above a certain threshold value Here: p. 10
Measurement Scenarios and Parameters Measurement locations: medium and large office environment laboratory environment medium size hall 14 different scenarios, line of sight / non line of sight Important Setup Parameters: Start Frequency 2 GHz IF Bandwidth 10 Hz Stop Frequency 10 GHz Number of Measurement Points 16001 Frequency step size 500 khz Sweep Time = approx. Generator Power 12 dbm Measurement Duration 30 min p. 11
Measurement Results Example for of a measured transfer function and its corresponding impulse response for a LOS scenario with antenna distance of d = 5.18 m: p. 12
y Measurement Results LOS, Medium Size Office Environment, 41 m 2 P31 630 cm 57 cm 68 cm P31 door 650 cm d=5 m P31 51 cm P31 BA-010 door Wooden Computer desks Metal Cabinet Wooden Equipment desks Blackboard Position of Transmitting Antenna Position of Receiving Antenna Vector Network Analyser x p. 13
Measurement Results LOS, Medium Size Laboratory Environment, 48 m 2 P38 y 28 cm 668 cm 68 cm BA-012 P38 d=4.05 m P38 714 cm P38 Wooden Computer desks Metal Cabinet 4 cm Wooden Equipment desks Blackboard Position of Transmitting Antenna Position of Receiving Antenna Vector Network Analyser x p. 14
Measurement Results NLOS, Laboratory Environment P37 y 57 cm 28 cm 379 cm 668 cm 68 cm door d=3.8 m P37 P37 714 cm BA-011 door P37 BA-012 Wooden Computer desks Metal Cabinet Wooden Equipment desks Blackboard Position of Transmitting Antenna Position of Receiving Antenna Vector Network Analyser 4 cm x p. 15
y Measurement Results LOS, Conference Room, 82 m 2 P11 1107 cm 739 cm 370 cm P11 d=5.39 m P11 BA 261 P11 door wooden computer desks wooden book cabinet wooden equipment desks Vector Network Analyser Position of Transmitting Antenna x Position of Receiving Antenna p. 16
Measurement Results LOS, Large Hall Environment, 95 m 2 P51 y P52 d= 5 m P52 P52 900 cm door Metal door 1050 cm Metal Ventilation Channel Metal Floor Metal Box Vector Network Analyser Position of Transmitting Antenna Position of Receiving Antenna x p. 17
Summary and Future Works Knowledge of the correlation properties are needed to evaluate the UWB channel capacity To obtain this information Frequency domain based measurement campaign has been performed 14 different scenarios have been investigated For each scenario the correlation function as well as the coherence bandwidth of the CTF has been determined The coherence bandwidth trends to be rather non-frequency selective for typical LOS small/medium size office environments compared with other scenarios To get deeper understanding in the correlation properties, further investigations on additional scenarios are currently in progress p. 18