Contract no.: 248231

More Safety for All by Radar Interferernce Mitigation

D2.6 - Multi-interference modeling and effects

Report type Deliverable

Work Group WP2

Dissemination level Public

Version number Version 1.0

Date 24-08-2012

Lead Partner Karlsruhe Institute of Technology

Project Coordinator Dr. Martin Kunert Robert Bosch GmbH

Daimlerstrasse 6

71229 Leonberg

Phone +49 (0)711 811 37468

martin.kunert2@de.bosch.com

MOSARIM No.248231 2012-08-24

2/20

copyright 2012

the MOSARIM Consortium

Authors

Name Company

Tom Schipper Karlsruhe Institute of Technology

Revision chart and history log

Version Date Reason

0.1 2012-07-01 Initial version

0.9 2012-08-18 Final content added

V0.99 2012-08-23 Version for review

V0.999 2012-08-24 Peer Review by Daimler & KIT

V1.0 2012-08-24 Final version for submission

MOSARIM No.248231 2012-08-24

3/20

Table of contents Table of contents ........................................................................................................................ 3 1. INTRODUCTION .............................................................................................................. 5 2. DESCRIPTION OF THE SCENARIO SETUP ................................................................. 6

2.1. Antenna placement for cars and vans .......................................................................... 6 2.2. Considered traffic scenarios (M,A,B) and description of the data processing applied 6

3. SIMULATION RESULTS OF SCENARIO M ................................................................. 8 3.1. PDFs as a function of the radar penetration rate ......................................................... 8 3.2. CDFs as a function of the radar penetration rate ......................................................... 9

4. SIMULATION RESULTS OF SCENARIO A ................................................................ 10 4.1.1. PDFs as a function of the radar penetration rate ................................................ 10 4.1.2. CDFs as a function of the radar penetration rate ................................................ 11

5. COMPARISON BETWEEN SCENARIOS A, B AND M FOR A RADAR PENETRATION RATE OF 50% .................................................................................... 12

5.1. PDFs .......................................................................................................................... 12 5.2. CDFs .......................................................................................................................... 13 5.3. The Number of involved interference sources for a penetration rate of 50% ........... 13

6. INTERPRETATION OF THESE SIMULATION RESULTS ........................................ 16 6.1. Influence of the radar penetration rate ....................................................................... 16 6.2. Influence of a multi-path generating environment .................................................... 16 6.3. The Number of relevant interfering sources .............................................................. 16

7. CONCLUSIONS & OUTLOOK ..................................................................................... 18 8. REFERENCES ................................................................................................................. 19 9. ABBREVIATIONS .......................................................................................................... 20

MOSARIM No.248231 2012-08-24

4/20

Executive Summary Interference effects from multiple sources are investigated with the simulator tool-chain in this report. An important result of this deliverable is the statement that based on simulations carried out so far, maximum interference contributions mainly occur in near vicinity of a target only and are usually caused by a rather limited number of interfering sources. Further tests with much higher traffic densities will show, if this behaviour holds even true under such worst case conditions with maximum interference potential.

MOSARIM No.248231 2012-08-24

5/20

1. INTRODUCTION Deliverable D2.5 has handled the situation that only a single radar interferes with a single victim receiver. Deliverable D2.6 now takes an in-depth look at very complex road scenarios, with respect to their immanent interference potential. Fig.1, taken from deliverable D1.2 [D12], shows such a scenario.

Fig. 1: Scenarios with high density of radars

On the basis of probability density functions (PDF) and cumulative distribution functions (CDF), two different road scenarios will be investigated. Also a comparison between a Matlab™ only simulation and the ray-tracing method will be performed to visualize the influence of multi-path effects. Results of this deliverable will be the basis for workpackage 5: “Deriving guidelines and recommendations for efficient mutual radar interference mitigation procedures.”

MOSARIM No.248231 2012-08-24

6/20

2. DESCRIPTION OF THE SCENARIO SETUP

2.1. Antenna placement for cars and vans

Fig. 2: Antenna setup for cars and vans Trucks are considered with forward looking radars only, similar to these of cars and vans.

2.2. Considered traffic scenarios (M,A,B) and description of the data processing applied

Within this deliverable, a street with four driving lanes is considered, but without guardrail. Scenario A (Fig. 3) has no boundaries but the street itself; while scenario B (Fig. 4) represents a street canyon. For comparison, a simple simulation setup in Matlab™ was created, including two-path propagation with incoherent add up of signals (being comparable to scenario A); no wave propagation phenomena were taken into account. This is named scenario M. The ray-tracing based simulations consider reflections including modified Fresnel reflection coefficients, diffraction being based on the uniform theory of diffraction as well as scattering [GENG][REICH]. Antennas are considered in all simulations as variants and correspond to the data shown in Fig. 3. The commercial traffic flow simulator Vissim is used to generate a realistic traffic behavior [VISSIM]. The traffic behavior is characterized by the traffic flow rate of 2000 vehicles per hour (i.e. 1000 vehicles per hour per lane) and a realistic vehicle composition on the basis of traffic censuses [TCBW], which is used as an input for Vissim. The radar penetration rate is defined as the ratio between cars equipped with radars to the total number of existing cars within a certain street section (here 500 m). The radar penetration rate

30° 30°

All antennas are placed 0.5 m above ground and have an EIRP of 20 dBm (24.125 GHz)

LCA beamwidths ±45° azimuth, ±24° elevation, 0 dBi

±12° azimuth, 8° elevation, 0 dBi ACC beamwidths

MOSARIM No.248231 2012-08-24

7/20

(named “pr” in all figures or RPR in normal text) is swept from 10% to 100% as a parameter in scenario M and A. Scenario B is simulated for a fixed 50% penetration rate, due to the high computation time. The total number of snapshots per scenario is about 4500. The sources [HIPPEL][MAURER] have been used to choose the material parameters shown in table 1.

Frequency ε’ ε’’ σ Asphalt 25 GHz 2.5 0.003 0.4 Concrete 25 GHz 2.35 0.002 1 Glass 25 GHz 5.04 0.086 0 Metal/PEC 25 GHz 1 106 0

Tab. 1: Used material parameters for simulations

Fig. 3: Scenario A+M: street without guardrail Fig. 4: Scenario B: Street canyon

MOSARIM No.248231 2012-08-24

8/20

3. SIMULATION RESULTS OF SCENARIO M The interference power shown in all figures is the result of an incoherent summing up of the specific contributions from all radars in that very scenario at the antenna port of the victim radar. The default polarization is horizontal-horizontal (HH-pol). Antenna gains are 0 dBi as described in Fig. 2, the transmit power is 20 dBm. Because of that, the results can be directly scaled to the readers needs. If a vehicle is equipped with radar, both front- and back mounted radars are considered. There is no vehicle, which is only “half equipped”.

3.1. PDFs as a function of the radar penetration rate

Fig. 5: Scenario M, front mounted radar

Fig. 6: Scenario M, back left mounted radar

MOSARIM No.248231 2012-08-24

9/20

Fig. 7: Scenario M, back right mounted radar

3.2. CDFs as a function of the radar penetration rate

Fig. 8: Scenario M: Comparison of CDFs for two different penetration rates

MOSARIM No.248231 2012-08-24

10/20

4. SIMULATION RESULTS OF SCENARIO A

4.1.1. PDFs as a function of the radar penetration rate

Fig. 9: Scenario A: front mounted radar

Fig. 10: Scenario A, back left mounted radar

Fig. 11: Scenario A, back right mounted radar

MOSARIM No.248231 2012-08-24

11/20

4.1.2. CDFs as a function of the radar penetration rate

Fig. 12: Scenario A: Comparison of CDFs for two different penetration rates

MOSARIM No.248231 2012-08-24

12/20

5. COMPARISON BETWEEN SCENARIOS A, B AND M FOR A RADAR PENETRATION RATE OF 50%

5.1. PDFs

Fig. 13: Comparison of PDFs for a front mounted radar; penetration rate 50%.

Fig. 14: Comparison of PDFs for a back left mounted radar; penetration rate 50%.

Fig. 15: Comparison of PDFs for a back right mounted radar; penetration rate 50%.

MOSARIM No.248231 2012-08-24

13/20

5.2. CDFs

Fig. 16: Comparison of CDFs for a front mounted radar; penetration rate 50%.

Fig. 17: Comparison of CDFs for a back left mounted radar; penetration rate 50%.

Fig. 18: Comparison of CDFs for a back right mounted radar; penetration rate 50%.

5.3. The Number of involved interference sources for a penetration rate of 50%

MOSARIM No.248231 2012-08-24

14/20

The following three plots Fig. 19-21 show the influence of the number of interfering radars per power class for 3 different use-cases: a front mounted victim radar, a back left mounted victim radar and a back right mounted victim radar. The number of classes is 50, subdivided linearly from -200 dBW to -70 dBW. The power is shown at the receiver antenna port with an antenna gain of 0 dBi and 20 dBm transmit power of each interfering radar. For the scenarios A, B and M, the number of radars within a single snapshot is about 75 (maximum 95) at a RPR of 50%. 100 randomly chosen snapshots for each scenario have been drawn in the figures. The total number of snapshots per scenario is about 4 500. The green lines in Fig. 19-21 mark the maximum interference power levels reached by single interference sources.

Fig. 19: Existing interfering sources for the different scenarios; front mounted radar

Fig. 20: Existing interfering sources for the different scenarios; back left mounted radar

MOSARIM No.248231 2012-08-24

15/20

Fig. 21: Existing interfering sources for the different scenarios; back right mounted radar

MOSARIM No.248231 2012-08-24

16/20

6. INTERPRETATION OF THESE SIMULATION RESULTS

6.1. Influence of the radar penetration rate Forward looking radars: For front mounted radars, Scenario A (with ray-tracing) as well as Scenario M (no ray-tracing, only two-path) shows a similar maximum interference power level of -87 dBW at a penetration rate of 100%. Further, the maximum interference power stays more or less the same for Scenario A, where the maximum interference power scales with an increasing penetration rate in Scenario M from -95 dBW to -87 dBW. The distribution’s peak moves from -108 dBW to -92 dBW for Scenario A and from -105 dBW to -94 dBW for Scenario M. Examples for reading the CDFs: Scenario M: Based on Fig. 8, with a RPR of 10%, the chance for a received interference power level of -110 dBW or higher is 90%. For a RPR of 100%, the chance for a received interference power level of -98 dBW or higher is 90%. Scenario A: Based on Fig. 12, for a RPR of 10%, the chance for a received interference power level of -115 dBW or higher is 90%. For a RPR of 100%, the chance for a received interference power level of -96 dBW or higher is 90%. Backward/sideward looking radars: The amount of received interference power is getting higher with an increasing RPR. The back left mounted radars scales “best” with the RPR. The reason for that is the looking-direction towards the middle of the street, getting interfered by many radars. The peak of its distribution moves from -105 dBW to about -90 dBW for a variation of a RPR from 10% to 100%. The maximum total received interference power of about -75 dBW keeps the same for all penetration rates and scenarios presented here. The chance for this maximum interference level slightly increases with increasing penetration rate. The CDFs can be used similar to the example above.

6.2. Influence of a multi-path generating environment The multi-path generating characteristics of Scenario B leads to increased interference power levels from radars, being farther away. This effect can nicely be recognized in Fig. 15. The magenta colored distribution shown (corresponding to Scenario B) loses some contributions at classes of lower powers (-105 dBW to -100 dBW), but gets them back at higher classes (-100 dBW to -90 dBW). However, such a multipath-generating environment does not change the maximum interference power levels for the scenarios presented here.

6.3. The Number of relevant interfering sources

MOSARIM No.248231 2012-08-24

17/20

If the modulation scheme of a single interferer is known, the effective interference floor in frequency domain can be estimated [D25]. An incoherent summing up of the different effective interference floors in frequency domain (based on the number of different interferers taken out from Fig. 19, 20, 21) leads to a final, estimated interference floor. In the here presented scenarios, the maximum number of interfering sources, for a level of up to 10 dB below the strongest path of each snapshot, is: 2 for a back left mounted radar, 4 for a back right mounted radar, and 10 for a front mounted radar. All this indicates, that there are no high numbers of equal, dominant paths possible within the scenarios presented here. Nevertheless, it clearly depends on the size of a target whether an interference source might be negligible or not.

MOSARIM No.248231 2012-08-24

18/20

7. CONCLUSIONS & OUTLOOK An environment, which creates a higher amount of multipath, leads to higher received interference power from sources far away. However, it could be shown, that the absolute maximum interference power levels are caused merely by sources within nearer vicinity of the victim. For the canyon scenario a RPR of 50% was chosen for the time being. Further simulations must be performed to check the situation for a 100% RPR. Also tunnel and intersection scenarios should be investigated next, as well as the influence of an increased traffic density. As a final goal, a combination of the statistical data and the interference floor prediction is intended. This result will enable the estimation and visualization of a certain interference floor in the presence of any waveform-combination on common streets via PDFs and CDFs. The simulation tools developed so far have now reached a quite comfortable status, where they can be applied, with small modifications only, to solve very complex interference issues and scenarios.

MOSARIM No.248231 2012-08-24

19/20

8. REFERENCES [D12] The MOSARIM Consortium, Deliverable D1.2 - ”Specification of

relevant scenarios, applications and traffic conditions,” Workpackage WP1: General Interference Risk Assessment, www.mosarim.eu, June 2010

[GENG] Geng, N., Wiesbeck, W., “Planungsmethoden für die

Mobilkommunikation: Funknetzplanung unter realen physikalischen Ausbreitungsbedingungen“, Springer 1998

[REICH] L. Reichardt, J. Maurer, T. Fügen, and T. Zwick, ”Virtual Drive: A

Complete V2X Communication and Radar System Simulator for Optimization of Multiple Antenna Systems,” 2011.05, Proceedings of the IEEE , vol. 99, pp. 1295-1310, IEEE, July 2011.

[VISSIM] VISSIM, macroscopic simulation tool for traffic flow modeling, PTV

Planung Transport Verkehr AG, Karlsruhe, Germany, www.ptv-vision.com

[TCBW] Straßenverkehrszentrale Baden-Würtemberg, “Bundesweite Straßen-

verkehrszählungen”, regularly actualized. www.svz-bw.de/bundesweite_zaehlung.html

[HIPPEL] A. R. von Hippel, “Dielectric Materials and Applications”, originally

published: Cambridge: Technology Press of MIT, 1954; new edition 2005. ISBN 0-89006-805-4.

[MAURER] J. Maurer, “Strahlenoptisches Kanalmodell für die Fahrzeug-Fahrzeug-

Funkkommunikation”, Forschungsberichte aus dem Institut für Höchstfrequenztechnik und Elektronik der Universität Karlsruhe; Band 43, Karlsruhe (TH), 2005

[D25] The MOSARIM Consortium, Deliverable D2.5 - ”Simulation of effects

and impact of environment, traffic participants and infrastructure,” Workpackage WP2: Simulation of radar interference mechanisms, www.mosarim.eu, August 2012

MOSARIM No.248231 2012-08-24

20/20

9. ABBREVIATIONS ACC Automatic Cruise Control BSD Blind Spot Detection CDF Cumulative distribution function EIRP Equivalent Radiated Isotropic Power LCA Lange Change Assist PEC Perfect Electric Conductor PDF Probability density function pol Polarization RPR/pr Radar penetration rate HH/hh/pp horizontally polarized transmit-antenna, horizontally polarized receive antenna VV/vv/tt vertically polarized transmit-antenna, vertically polarized receive antenna HV/hv/pt vertically polarized transmit-antenna, horizontally polarized receive antenna VH/vh/tp horizontally polarized transmit-antenna, vertically polarized receive antenna