Ultrawideband Radiation and Propagation

Transcription

1 Ultrawideband Radiation and Propagation by Werner Sörgel, Christian Sturm and Werner Wiesbeck LS telcom Summit July 26

2 UWB Applications high data rate fine resolution multimedia localisation UWB solutions large number of users sensor nets 2

3 Frequency Allocations UWB 3

4 UWB Frequency Allocations in US and EU /MHz FCC UWB Indoor DRAFT ECC 3.1 GHz GHz GHz 6 GHz 9 GHz DRAFT ECC with mitigation DRAFT ECC until frequency in GHz 4

5 Comparison of Shannon Channel Capacity Theoretical Capacity depends on: Bandwidth B Signal to Noise Ratio SNR C = B log 2 ( 1+ SNR) transmit power P Tx distance R ambient noise Rx noise figure N F SNR = P Tx G Tx G Rx N F ktb 2 $ " 1 & ) % 4# ( R. n n = 2.3 Ultra Wideband: distance R in m short range and high data rate communications 5

6 UWB Indoor Propagation Channel by Werner Wiesbeck, Christian Sturm, and Werner Sörgel

7 Definition Radio Channel radio channel = two port formed by two antennas (antenna + propagation + antenna) a 1 b a 2 b 2 channel impulse response h(t) H(f) = b 2 /a 1 extension: MIMO radio channel n-port formed by n antennas " h 11 h 12 L h 1M % $ h h = $ 21 h 22 L h 2M $ M O M $ # h N1 L h NM & MIMO: Multiple Input Multiple Output n = N + M N M 7

8 Indoor Visualization 2D-FDTD Hertzian dipole db Gauss-Monocycle FWHM = 88 ps 3 m 4 m -3 db 8

9 Mono Cone Antenna Theory: Need for infinite ground plane! Reality: only small space available. Pulse excitation 88 ps 9

10 Measured Directional Impulse Response receive θ E(θ,ψ) ~δ(t-r/c) u rx ~ h(t,θ,ψ) θ transmit u tx ~ δ(t) E(θ,ψ) ~d/dt h(t,θ,ψ) 1

11 Impulse Response Parameters Impulse response monocone θ = 44 peak in m/ns peak value Full Width at Half Maximum analytic envelope.4 FWHM in ns.2 FWHM peak value.4 peak in m/ns.2 ringing time in ns θ in degree 11

12 Mono Cone: Measured Gain and Transient Response h n in m/ns.25-5 gain in dbi 5 12

13 Deterministic Channel Modelling by Raytracing by Werner Wiesbeck and Werner Sörgel

14 Modelling with 3D Ray Tracing Tx Rx deterministic modeling approach creation of polygon object model reflection and diffraction path search coherent addition of transmission coefficients 14

15 Simulation Procedure Scenario: Polygon object model Frequency table (e.g. 151 points, GHz) Ray Tracing Algorithm (path search and superposition) Antenna transfer function (complex, polarimetric) Channel transfer function Correlation properties IFFT Channel impulse response Power delay profile Statistics (delay spread,...) 15

16 Sim. Spatial Power Distribution Indoor 2 (lossy) UWB GHz y in m db Y Z X monocone transmit and receive antennas WLAN GHz y in m x in m x in m

17 Power Delay Profile for UWB Transmission frequency range 3.1 to 1.6 GHz simulation at 151 frequencies (5 MHz spacing) IFFT processing with Hanning window Power delay profile Rx Tx Channel attenuation in db time in s 17

18 Antenna Deconvolution by Werner Sörgel and Werner Wiesbeck

19 Path Transfer Functions Tx Rx U oc Rx Tx $ " = 8Z o C & ) % 4# ( 2 ( ) ( ) N $ C Rx G Rx G Tx P Tx * * Rx Rx i,+ 2 i & C Rx + * Rx Rx i=1 % i,+ i ) ( T, T**i T *+i. T +*i T ++i - ( ) ( ) / $ 1 C Tx * * Tx Tx i,+ i & C Tx * * Tx Tx % i,+ i Polarimetric superposition of discrete pathes T weighted with complex antenna pattern C of transmitter and receiver. Reduction to 2D problem: ) ( omni-directional transmitter receiver resolves azimuth angle ψ co-polarized transmitter and receiver U oc Rx = Z Rx N C Tx Z j"h Tx # U Tx Rx p $ H #,i H ##,i C i=1 19

20 Deconvolution: Rotational Method Antenna transfer function pattern H(ψ i ): measured for N discrete angles ψ i ψ i Received signa U RX for discrete antenna orientations ψ j : measured for same set of N discrete angle a 1 a 2... a i... signals (pathes) received signal for a single ψ C,i U j Rx (" C, j ) = " " $ $ $ $ $ MM $ # $ # Rx U 1 b 1 $ Rx Ub 2 2 Rx Ub NN N Rx Z C # a i $ H Rx % (" i, j ) i=1 % " Hh 1 hh 22 L hh NN %% " a 1 % $ $ h h 3 h 1 = $ H 2 H 3 L H 1 ) $ a 2 $ M O $ M & $ $ # Hh N hh N(1 N(1 L hh N(1 N(1 && # a N & impulse response pattern H(ψ j ) ψ A deconvolution by matrix inversion yields vector a (sampled pathes) 2

21 Rotational Measurement with Ridged Horn Scenario B Measured CIR Rx: ridged horn room h/h max in db Tx Rx NWA delay in ns azimuth in deg door monitor metal plane 7.5 m desk / bookshelf wooden desk 21

22 Model of Vivaldi Antenna Vivaldi antenna modelled with rotational method Verification measurement h/h max in db h/h max in db delay in ns azimuth in deg delay in ns azimuth in deg 22

23 Comparison Measured and Modeled Vivaldi Peak envelope (azimuth) Difference meas-model max( DPD db ) h mod -h max in db delay in ns azimuth in deg delay in ns time shift due to offset between center of rotation and center of radiation 2D-mapping for 3D problem 23

24 Antenna Performance for 3-band OFDM CDF(E b /N ) drop out 3-Band-OFDM E b /N in db ridged horn Vivaldi Logper Bowtie (36x31) Conditions: 48 Mbit/s same scenarios (distances 2 und 4 m, LOS) rotation of AUT Tx-power: -1,6 dbm/1,5 GHz (3- band) based on link budget calculation with measured path loss 24

25 Conclusions frontend antenna wave propagation co-design of frontend and antennas matching for noise and power predistortion frequency dependent characteristics fine structure of channel response EIRP considerations transient radiation meas. LOS channel IR propagation scenario Re{h(t)}/h max 1 Vivaldi LPDA BW: 2-12 GHz time in ns 25