Ranga Rao Makanaboyina* and Mohan Kumar Sadashivaiah. Hans-Ludwig Bloecher

Transcription

1 196 Int. J. Ultra Wideband Communications and Systems, Vol. 1, No. 3, 2010 Interference effects of 24 GHz UWB automotive short-range radar to broadcasting and fixed satellite services in Ku and Ka bands Ranga Rao Makanaboyina* and Mohan Kumar Sadashivaiah Mercedes-Benz Research and evelopment India, Embassy Golf Links Tech Park, omlur, Bangalore, , India *Corresponding author Hans-Ludwig Bloecher aimler AG, Wilhelm-Runge-Str. 11, Ulm, , Germany Abstract: The introduction of the automotive radar was a challenge for spectrum engineering and management primarily because of its out-of-band emissions, and not due to its in-band emissions. The SRR will have a large operating bandwidth of up to 5 GHz and could potentially cause interference to broadcasting and fixed satellite services operating/planned in the nearby Ku and Ka bands. The emissions from 24 GHz systems on-board a vehicle should not cause unwanted interference to these services. Accompanying the market introduction of 24 GHz ultrawide band (UWB) short automotive radar (short-range radar), extensive co-existence studies have been performed between 24 GHz UWB short-range radar and different present and planned RF services. In this paper, an estimate of the interference due to direct-coupled fields from the SRR s on a single vehicle to BSS and FSS systems in Ku and Ka bands, is studied. In addition, aggregate interference from multiple SRR s mounted on single/multiple vehicles are also analysed. This paper introduces a methodology for coexistence studies between 24 GHz UWB SRR and BSS/FSS services. Keywords: short-range radar; SRR; broadcasting satellite services; BSS; fixed satellite services; FSS; electromagnetic compatibility; ultra wideband; UWB; coexistence; vehicular radar. Reference to this paper should be made as follows: Makanaboyina, R.R., Sadashivaiah, M.K. and Bloecher, H-L. (2010) Interference effects of 24 GHz UWB automotive short-range radar to broadcasting and fixed satellite services in Ku and Ka bands, Int. J. Ultra Wideband Communications and Systems, Vol. 1, No. 3, pp Biographical notes: Ranga Rao Makanaboyina received his Masters and Bachelors in Electrical Engineering from Osmania University, India in 1990 and 1988, respectively. From 1990 to 1999, he worked with the efence Research and evelopment Organization, Hyderabad in different Radar System esign studies. Since 1999, he is with Mercedes-Benz Research and evelopment India working in the field of automotive safety systems research and development projects. Mohan Kumar Sadashivaiah received his Masters and Bachelors in Electrical Engineering from Bangalore University, India in 1998 and 1995, respectively. From 1999 till date, he is with Mercedes-Benz Research and evelopment India working in the field of signal processing, short-range radar design for vehicular application, ultra wideband communication and distributed MAC protocols for inter-vehicle communication application. Currently, he is working on vehicle suspension systems, model-based function development and testing of automotive electronics. Hans Ludwig Bloecher received his ipl.-ing. in Electrical Engineering and his r.-ing. in Microwave Engineering from the University of Siegen, Germany. In 1997, he joined aimler-benz Aerospace (now EAS) as a Radar Systems Engineer and leader of technical studies. In 1999, he joined aimlerchrysler AG Corporate Research as an Internal Consultant in the fields of corporate research business development and technology transfer. Since 2001, he is with present aimler AG Group Research & Advanced Engineering as a Senior Researcher, Manager of internal and public funded R& projects and Supervisor of research teams in the Copyright 2010 Inderscience Enterprises Ltd.

2 Interference effects of 24 GHz UWB automotive short-range radar to broadcasting and fixed satellite services 197 field of millimetre wave automotive radar. Since 2002, he represents aimler AG in international spectrum engineering and standardisation gremia and technical bodies. He is the author or co-author of more than 70 contributions to journals and conferences. 1 Introduction The automotive industry has developed systems using short-range radars (SRR) for environment sensing in road safety and intelligent transportation applications. SRR operates at very low power levels with narrow elevation antenna beam widths and full azimuth coverage. The SRR centred at GHz has bandwidths up to 5 GHz and has unintended emissions outside of its band. The out-of-band emissions of SRR can interfere with other licensed services such as fixed systems (FS), radio astronomy services (RAS), Earth exploration satellite services (EESS) and broadcasting satellite service (BSS) in region 1, in 21.4 to 22.0 GHz band and BSS and fixed satellite service (FSS) in GHz band. Region 1 covers Europe and Africa. Region 2 covers North and South America. Region 3 covers Asia and Australia. The interference compatibility of in-band emissions of SRR with FS, RAS and EESS is studied in ECC Report 23 (2003). In this paper, we studied the compatibility of SRR with other satellite services such as BSS and FSS. The following satellite services are considered for studying their electromagnetic compatibility with SRR: FSS in 12 to 18 GHz (Ku-band) and 20 GHz (Ka-band) BSS in 12 to 18 GHz (Ku-band) BSS in 21.4 GHz to 22.0 GHz (Ka-band). BSS has frequency allocation for high definition television (HTV) broadcast near 21 GHz from April 2007 (Proceedings of the conference, 1992). The 21 GHz BSS and 24 GHz SRR overlap in frequency. To study the electromagnetic compatibility between them, estimates of direct-coupled fields due to interference from SRR operating on a single vehicle and aggregate interference from multiple SRR s operating from multiple vehicles are required. If the aggregate interference from multiple SRR s onboard multiple vehicles are within the protection margin of the BSS/FSS system as specified in International Telecommunication Union (ITU) reports/recommendations, then the compatibility between them is established. The study is organised as follows: Section 2 gives details of SRR system specifications taken from reports of regulatory and standardisation bodies, operational characteristics of BSS in Ku/Ka bands and FSS in Ku/Ka bands. Section 3 discusses interference from SRR, onboard a single vehicle, to Ku-band and Ka-band FSS systems. Section 4 discusses the interference from SRR to Ku-band BSS and Ka-band BSS, for single vehicle scenario. Section 5 discusses SRR-BSS (Ka-band) compatibility for multiple-vehicles scenario. Section 6 gives conclusions outlining margins available after establishing electromagnetic compatibility of SRR with BSS and FSS systems. 2 SRR and satellite services 2.1 Short-range radar SRR operates at 24 GHz with operating range of up to 30 m. SRR s are used in active and passive systems to provide assistance to the driver for road safety. Typical automotive applications based on SRR are obstacle avoidance, collision warning, lane departure warning, lane change aid, blind spot detection, parking aid and airbag activation. The function of 24 GHz SRR is to provide speed/position information. The peak power and average powers defined are +20 dbm and 0 dbm, respectively. The SRR sensors that provide high-resolution position information of objects operate at average spectral power density of 41.3dBm/MHz with a bandwidth of 5 GHz. The frequency band of operation of SRR as defined in ETSI standards in draft EN (raft ETSI European Standard EN v1.1.1, ) is: narrowband: GHz to GHz wideband: GHz to GHz. The vertical discrimination antenna pattern for 24 GHz SRR given in Figure 1 is taken from ETSI document EN v1.2.1 (2001). The vertical antenna angle is positioned on 0 for a vector direction parallel to ground. Figure 1 Antenna pattern of SRR (see online version for colours) The mounting height of 24 GHz SRR on the vehicle is limited to a maximum of 1.5 m. However, the mounting height for forward and rearward facing SRR sensors is about 0.5 m for cars.

3 198 R.R. Makanaboyina et al. 2.2 FSS in Ku-band and Ka-band A geostationary orbit (GSO) FSS consists of a space station in a GSO and an earth station (ES) located at a fixed position on the ground. The signals from the Earth-to-space are up linked at a frequency different from the signals from space-to-earth (SE). FSS systems located on GSO orbits in frequency bands 12.5 to GHz and 17.7 to 21.2 GHz is considered for interference analysis. Compliance to ITU recommended values of carrier-to-noise plus interference ( C N + I ) guarantees a sufficient signal-to-noise (SNR) ratio to each of the carriers operating in FSS bands. The methodologies to calculate protection margins from interference into the carriers of FSS are given in detail in Recommendations of ITU-R S (1994). 2.3 BSS in Ku-band BSS in Ku-band has frequency allocation for HTV. BSS-HTV is classified into two types narrowband HTV and wideband HTV. The narrowband HTV requires lower bandwidth and operates in Ku-band transponders at 12 GHz BSS. BSS space stations transmit or retransmit signals that are intended for direct reception by the general public. In the broadcasting-satellite service, the term direct reception encompasses both individual reception and community reception. Radio Regulation 123 (RR123) and RR124 define individual and community reception as follows (RR123 ITU Radio Regulations, 2004; RR124 ITU Radio Regulations, 2004): Individual reception: The reception of emissions from a space station in the broadcasting-satellite service by simple domestic installations and in particular those possessing small antennas. Community reception: The reception of emissions from a space station in the broadcasting-satellite service by receiving equipment, which in some cases may be complex and have antenna larger than those used for individual reception and intended for use by a group of the general public at one location or through a distribution system covering a limited area. The density of BSS receivers is very high and the receivers could be located very close to urban/suburban roads. There is a high possibility of interference from any device that is operated on roads (like SRRs). In case of individual reception, antenna may be mounted on the rooftop of houses/apartments/offices. The antenna mounting height can be as low as 2 m from ground level. The lower the relative height of the BSS antenna with respect to the SRR antenna, higher will be the interference from out-of-band/in-band emissions of SRR. allocations, WARC-92 made an allocation in the bands GHz in Regions 1 and 3 to the BSS on a primary basis from 1 April Resolution 525 of the WARC-92 further states: It shall be understood that prior to 1 April 2007 all existing services in the band GHz in Regions 1 and 3 operating in accordance with the Table of Frequency Allocation shall be continue to operate. After that date they may continue to operate, but they shall neither cause harmful interference to BSS (HTV) systems nor be entitled to claim protection from such systems. 2.5 Interference calculations For all the three satellite services mentioned in the previous sections, the downlink carrier-to-interference can be determined using the following equation: ( ) 1 C = C + + C N ISRR N ISRR where 1 1 C is downlink carrier to noise plus N + I SRR interference ratio ( C N ) C I SRR is downlink carrier to noise ratio is downlink carrier to interference ratio. The degradation in C + due to interference from N I SRR SRR is obtained after simplifying the previous equation as: C N + I C N Figure 2 ( ) SRR 1 = I 1+ N SRR Interference geometry (1) (2) 2.4 BSS in Ka-band The wideband HTV requires wider bandwidth and is proposed for operation at 21 GHz. The World Administrative Radio Conference dealing with frequency

4 Interference effects of 24 GHz UWB automotive short-range radar to broadcasting and fixed satellite services 199 Interference-to-noise ratio from a SRR device to a BSS system (see Figure 2) is calculated using: I SRR = PSRR Abump + iscsrr FSL N (3) Gain ktb F where BSS/ FSS BSS/ FSS P SRR power transmitted by radar in BSS/FSS bandwidth A bump attenuation due to bumper (db) isc SRR discrimination of radar in the direction of BSS/FSS receiver (db) FSL free space loss (db) Gain BSS/ FSS gain of BSS/FSS antenna in the direction of the radar (db) ktbbss/ FSSF noise floor of BSS/FSS receiver. All the interference calculations discussed in next sections are computed for the interference geometry shown in Figure 2. The θ el and θ r are angles of BSS/FSS antenna boresight direction and the SRR antenna boresight direction respectively, from the line of sight (LOS) between them. 3 SRR interference to Ku and Ka band FSS In this section, we use equation (3) and calculate the I SRR /N of SRR to Ku-band and Ka-band FSS systems. The I SRR /N ratios are compared with the available protection margins in their respective bands. 3.1 Apportionment to radio services in FSS For interference compatibility among many users in FSS frequency band, ITU recommends apportionment to the radio sources in the following way: 20% or 25% of 32% or 27% of clear sky noise of ES is the total apportionment 20% or 25% for primary radio services depending on frequency reuse methods 6% for co-primary service such as FS 1% for unlicensed and out-of-band emissions, non-primary radio sources. The values listed in ITU-R S define the maximum allowable interference. For example, the maximum permissible interference-to-noise ratio (I SRR /N) to digital FSS carrier, as specified in Rec ITU-R S.523-1, is 12.2 db. This value corresponds to 6% increase in total noise power due to interference from other radio services sharing the frequency allocations on a co-primary basis such as fixed service (FS) links (Refer Rec ITU-R S.1432). An apportionment ratio of 1% is allotted to emissions from other radio services sharing the frequency allocation on a non-primary basis, unlicensed devices and unwanted emissions including out-of-band emissions. The interference from 24 GHz SRR into FSS at 12 GHz falls under the 1% category. 3.2 Protection criteria for FSS For the interference analysis of FSS in Ku-band, the out-of-band emission of the interferer system (SRR) in the victim receiver (FSS) frequency band is considered. The EIRP of the SRR considered for the study is 61.3 dbm/mhz. This value is based on FCC approval of SRR in the USA. Since three types of services fall under the 1% apportionment of interference, it is safe to assume that out-of-band emissions of SRR could contribute to 0.5% of the total apportionment allotted to unwanted interferences. This means I/N ratio of 23.0 db (corresponding to 0.5%) from out-of-band emissions of SRR, forms the protection criteria for FSS. This ratio is applicable to all types of carriers of FSS mentioned in ITU. I is the power of any interfering system. Hence, we conclude that if the total I SRR /N due to all SRRs operating near the ES is less than 23.0 db, it implies SRR can be compatible with FSS. 3.3 Results of SRR interference to Ku-band and Ka-band FSS Table 1 lists the characteristics of candidate FSS systems. The intensity of interference depends on the antenna pattern of SRR and FSS and the distance between them. Therefore, interference from SRR to these FSS systems is calculated using equation (3), for different relative antenna heights h and separation distances d and is shown in Table 2. The antenna patterns (Figure 3) for the Ku-band and Ka-band FSS receive ES are computed as per Rec ITU-R S.1428 (2001) and are used for the interference calculations. Figure 3 FSS receive ES antenna vertical pattern (see online version for colours)

5 200 R.R. Makanaboyina et al. Table 1 Operational characteristic of FSS Receive ES Ku-band Eutelsat Ka-band Eutelsat Ka-band France Carrier no Freq. (GHz) Altitude (km) Elevation angle Latitude Noise BW (KHz) ,000 Threshold: C/(N + I) (db) % of the year C/(N + I) should be exceeded ES noise temp., K On-axis ES ant. gain (dbi) Antenna diameter (m) Table 2 Frequency Separation distance (m) Interference-to-noise ratio of SRR to Ka band FSS Ka-band France d = 30 d = 30 d = 30 d = 30 Height (m) h = 0 h = 5 h = 10 h = 100 SRR EIRP (dbm/mhz) SRR Tx power (dbm/mhz) SRR ant. off-axis gain (db) Bumper loss (db) FSL (db) Atmospheric loss (db) FSS off-axis gain (db) Received power (dbm/mhz) Reference BW ,000 (KHz) Received power in reference BW (dbm) Interference power dbm (I SRR ) Noise power dbm (N) I SRR /N db The I SRR /N ratio is computed for three relative mounting heights between SRR and FSS/BSS for separation distance of 30 m. The results indicate that SRR can be compatible with FSS with margins of 30 db above the ITU specified values for single vehicle scenario. In multiple vehicle scenarios, the increase in aggregate interference will be offset with the increased distance of the vehicle borne SRR s from the FSS-ES. The mounting heights of the FSS-ES relative to the SRR mounting heights help in reducing I SRR /N ratio to a large extent. To conclude I SRR /N ratio improves by 6 db for every 5 m of relative height of the FSS ES antenna with respect to the SRR antenna. 4 Interference to Ku-band and Ka-band BSS 4.1 Protection margin The total clear sky C/N + I margin above operating threshold value specified by ITU-R BO1444 (2000) provides a margin of 7.6 db. If a decrease in C/N + I margin due to interference from SRR of 0.5 db is assumed to be tolerable, the aggregate interference-to-noise ratio solely due to SRR will be I SRR /N = 10 db. Since information on apportionment of the protection margin was not available for BSS, a 0.5 db decrease in C/N + I due to SRR was assumed tolerable. The compatibility study is based on this assumption. Figure 4 Co-polar and cross-polar receiving ES Notes: Individual reception: ϕ 0 = 2 (nominal half-power beamwidth); community reception: ϕ 0 = 1 (nominal half-power beamwidth); curve A: co-polar component for individual reception without side-lobe suppression (db relative to main beam gain); curve A : co-polar component for community reception without side-lobe suppression (db relative to main beam gain); curve B: cross-polar component for both types of reception (db relative to main beam gain); and curve C: minus the on-axis gain (curve C in this figure illustrates the particular case of an antenna with an on-axis gain of 37 dbi). 4.2 Results of SRR interference to Ku-band BSS The antenna patterns for individual reception and community reception as described in ITU-R BO.652-1

6 Interference effects of 24 GHz UWB automotive short-range radar to broadcasting and fixed satellite services 201 (1991) are shown in Figure 4. An on-axis antenna gain of dbi is used in calculations. The interference to ES BSS receiver (downlink) is considered in the current study. Table 3 shows I SRR/N calculation for BSS antenna height of 0.5 m and a separation distance of 10 m. This will represent a worst-case interference scenario. Calculations are performed for interference from a single SRR device to BSS for relative antenna heights of 2 m and 5 m (Table 4) and separation distance of 10 m, 20 m and 30 m. Table 3 Interference to noise ratio for Ku band BSS Individual reception Community reception BSS antenna height (m) SRR mounting height (m) Separation distance (m) BSS antenna elevation Maximum SRR EIRP in BSS band (dbm/mhz) BSS receiver BW (MHz) Bumper attenuation (db) 3 3 BSS on-axis antenna gain (db) BSS ant. of-axis gain FSL (db) Interference power at the BSS antenna (dbm) Noise figure (db) 5 5 Noise power (dbm) Interference-to-noise ratio at BSS Rx. (db) Table 4 Separation distance (m) I SRR /N for BSS antenna height of 2 m and 5 m I SRR /N (db) (individual reception) I SRR /N (db) (community reception) h = 2 m h = 5 m h = 2 m h = 5 m The I SRR /N to the BSS from SRR is estimated to be at least 22 db in the worst-case scenario of separation distance of 10 m and height of 0.5 m, for both individual reception and community reception. For higher relative heights between SRR and BSS, the I SRR /N is less than 28.0 db (Table 4). Therefore, the SRR and BSS can be concluded as compatible for single vehicle scenario (Makanaboyina et al., 2004). Since the margin of safety available with single vehicle scenario is quite high, the interference study is not repeated for multiple vehicle scenario for Ku band BSS systems. 4.3 SRR interference to Ka-band BSS single vehicle scenario Frequency overlap of SRR and BSS The proposed SRR frequency band is located adjacent to the BSS 21 GHz allocation. Figure 5 shows the power spectral density (PS) for SRR as defined in the ETSI draft standard (raft ETSI European Standard EN v1.1.1, ) and PS of BSS signals. The PS of SRR is constant at 41.3 dbm/mhz from GHz to GHz. The mask slopes down at the rate of 20 db/ghz for frequencies below GHz and frequencies above GHz. In the overlapping region from 21.4 GHz to 22.0 GHz (the band allocated to BSS), the PS varies between 61.3 dbm/mhz to 53.8 dbm/mhz. Figure 5 Spectrum assigned to 24 GHz SRR SRR could potentially interfere with BSS both by in-band emissions ( GHz to 22 GHz) and out-of-band emissions (21.4 GHz to GHz). The protection margin (PM) is different for OOB and IB emissions of SRR. The SRR service is proposed as a secondary service on a non-protected, non-interfering basis. The rules for spectrum sharing depend on whether the interfering system causes in-band or out-of-band emissions. Therefore, the interference assessment is discussed for the two subbands (21.4 to GHz and to 22.0 GHz) separately. The study is done keeping in view the spectrum mask allowed for SRR by FCC (2002) in the USA, is a box-type allocation (Figure 1) whereas the spectrum mask being proposed in Europe (raft ETSI European Standard EN v1.1.1, ) has a sloping allocation near the 22 GHz frequency Service requirements and availability The graceful degradation of HTV reception quality as function of C/N ratio defined in ITU-Report (Report ITU-R BO1075-2, 1998; Report ITU-R BO , 1998) has a fall back quality around 11 db (Figure 6) and service interruption below 5 db.

7 202 R.R. Makanaboyina et al. Figure 6 Graceful degradation of 21 GHz BSS channel ITU recommendations on compatibility, ITU has recommended a minimum I/N value of 20 db for secondary sources. This value is used for interference assessment in this study. The total I SRR /N due to all SRR s operating near the BSS system for multiple vehicle scenarios is estimated. If this estimate is less than the specified protection criteria of 20.0 db, then the electromagnetic compatibility between the SRR and the BSS is established BSS system in 21 GHz ue to the binary structure of the information transmissions, the degradation is expressed as worsening of the system C/N performance and therefore as a reduction of the service availability. ITU report further suggests that C/N degradation due to interference of some 5% to 10% seems to be an acceptable value. This is equivalent to I/N margin of up to 9.5 db. This value of I/N is acceptable for interference to Ka-band BSS from all possible sources of interference. Because of the high frequency of operation, the emitted signals shall be received by small (45 90 cm) individual antenna. There were proposals for ITU recommendations for availability of this service for % of the time during the worst-month for service continuity (Report ITU-R BO , 1998). It is also expected that BSS (HTV) at 21 GHz will have to cope with large rain induced attenuation in most countries Sources of interference The interference to BSS could come from any of the following sources: BSS co-channel and adjacent channel interference co-primary services secondary sources unwanted emissions. The protection ratios defined in ITU Report BO2019 (1999), for BSS co-channel and adjacent channel (digital signals) are of the order 31 db and 15 db, respectively. BSS band does not have co-primary services. Since BSS systems do not allow interference from their own adjacent channels to be greater than 15 db, it is highly unlikely that BSS would like the interference from secondary source such as SRR, to be greater than 15 db. Generally, the apportionment to interference from secondary sources (such as SRR) given by ITU is 1% of the total clear sky noise power. This is equivalent to I/N of 20 db. It may be noted that no linear relationship between the I/N and the available apportionment can be assumed and therefore the I/N of 20 db is only an absolute lower bound criterion used for availability of apportionment estimations. In most of the The characteristics of a reference BSS system chosen for the interference study are based on the system specifications listed in the ITU Report for Wideband HTV operation (Report ITU-R BO , 1998). The parameters relevant for interference estimation are reproduced below: Table 5 Parameters of 21 GHz BSS systems Frequency (GHz) Receiving antenna diameter (cm) 45 NF of converter at home (db) 1.5 Available service time 99% of the worst-month Required total C/N (db) 10 Usable bit rate (Mbits/s) Modulation QPSK with 3/4 convolutional code Bit rate (Mbits/s) Nyquist bandwidth MHz Number of channels/satellite 3 Required BER 10 8 Antenna efficiency 70% Equivalent noise temperature K SRR s characteristics and the associated losses like bumper loss, car shielding and attenuation due to rain, objects such as vegetation/fencing in the line of sight (LOS) are considered (ECC Report 23, 2003). The interference is studied for the following parameters: 1 BSS antenna height: 0 to 5 m 2 distance between cars: 10 to 65 m Antenna gain Antenna pattern/characteristics of receiving equipment for individual and community reception for BSS 21 GHz systems are being studied by ITU under the question number ITU-R 21/6. For the current study, the antenna pattern available for BSS 12 GHz band is used (ITU-R BO.652-1, 1991). The antenna patterns for Individual reception and community reception described in ITU-R BO are used (Figure 5). An on-axis antenna gain of 37.8 dbi corresponding to antenna diameter of 45 cm and efficiency of 70% is used in the calculations.

8 Interference effects of 24 GHz UWB automotive short-range radar to broadcasting and fixed satellite services 203 The antenna diameter and efficiency values are used to calculate the BSS off-axis gain in the direction of the interference path. The noise temperature and bandwidth are used to compute the noise power. For a BSS system with noise figure of 5 db, the noise floor of BSS systems would be 95.6 dbm. This value is used in the I/N computation Interference reduction due to short vegetation/guard rails The presence of short vegetation/shrubs and guardrails or fences around the house or roadway provides considerable attenuation to RF signals from SRR. The reflections from the short vegetation/guard rails are dependent on the angle of incidence with respect to the LOS between BSS and SRR antennas. The area of vegetation/guard rails offering shielding is approximately 0.4 m by 1.0 m. The radar reflectivity indicates fraction of power received in the direction of the transmitting antenna. After accounting for scattering along all directions, the attenuation due to vegetation/guard rails would be at least 10 db (Ulaby and obson, 1989). This value is used for computing interference-to-noise ratio for single vehicle and multiple vehicle scenarios in the following sections. For multiple vehicle scenarios, the BSS antenna height is of main importance. ue to shielding effects the nearest car masks out all other cars by blocking the direct LOS to the BSS antenna that is installed below the blocking vehicle height Results of SRR interference to Ka-Band BSS Table 6 lists BSS and SRR antenna and interference geometry parameters. Table 7 gives interference to noise ratio due to single vehicle scenario for individual reception for both subbands, at BSS height of 0.5 m and separation distance of 10 m. Furthermore both the 61.3 dbm/mhz for OOB and the 53.8 dbm/mhz SRR EIRP emission level are worst-case assumptions that represent the maximum value of the still decreasing slope. Table 8 gives interference-to-noise ratio for BSS antenna heights of 0.5, 2 and 5 m at separation distances of 10 m to 40 m. Table 6 Parameter BSS antenna height (m) SRR antenna height (m) Separation distance (m) BSS antenna elevation (at latitude: 41.2 ; long: 9.5 ) Parameters for interference calculation Frequency GHz GHz Table 7 Parameter Interference-to-noise for 21 GHz BSS Max. EIRP of SRR in BSS band (dbm / MHz) Frequency (GHz) < 61.3 < 53.8 BSS receiver bandwidth (MHz) Bumper attenuation (db) 3 3 SRR antenna gain (db) 0 0 BSS on-axis antenna gain (db) BSS off-axis gain (db) Min. shielding loss from vegetation/guard rail/fences/etc. (db) Free space loss (db) Interference power (dbm/mhz) Interference power (dbm/52 MHz) Noise power at BSS receiver (dbm/52 MHz) Interference-to-noise ratio (db) I SRR /N Protection criterion (I/N) Protection margin (db) Table 8 I SRR /N for BSS for heights h = 0.5 m, 2.0 m and 5.0 m I SRR /N (db) d (m) F = 21.4 to GHz h = 0.5 m h = 2 m h = 5 m f = to 22.0 GHz h = 0.5 m h = 2 m h = 5 m For a typical mounting height of the BSS antenna at 2 m (wall installation) or 10 m (roof installation), a positive margin of more than 13 db (@ 2 m) and 25 db (@ 5 m) is available that allows a sensor aggregation in a 10 m distance of 21 to 295, respectively. Assuming four active sensors per vehicle this results in five to 79 vehicles that must be all located in 10 m distances with no mutual or roadside shielding effects. In real road scenarios, this is impossible.

9 204 R.R. Makanaboyina et al. 5 SRR interference to Ka Band BSS multiple vehicle scenario For automotive safety applications, a number of sensors may be mounted on the exterior of the car. A typical car may have two in the front, two in the rear and one or two on each side with a maximum of four sensors operating at the same time. Studies on aggregate interference to BSS receivers require that the interference is within limits in the worst possible scenario where a number of vehicles may contribute simultaneously to the interference. It is possible that vehicle borne SRR s could be spread within a small area near a BSS receiver. One can visualise a number of scenarios for aggregate interference calculations. Some of them are: Figure 8 Interference scenario of BSS antennas near a T-junction (see online version for colours) BSS antenna located on roof-top of an apartment by the side of a road (Figure 7). BSS antenna located in the first or second floor balcony of an independent building with heights of 5 m or above. Antennas located in the ground floor of an independent house close to the ground. One such scenario is shown below. However, the probability of antennas mounted close to the ground is very less. Figure 7 BSS antenna mounting For interference analysis, vehicles present in SRR antenna view of horizontal sector of 180 are considered. The cars ahead shield radar signals from the cars behind. Three parameters determine the interference to a large extent. They are: the maximum length of queue of cars; the shielding effect; and shielding from intermediate objects such as short vegetation, shrubs, fences or guardrails in front of the house in the direction of interference path. Figure 9 I SRR /N versus separation distance (see online version for colours) 5.1 Aggregate interference scenario The following figure shows the vehicles queuing up near the T-intersection. Assume a single BSS antenna is located in the front side of the house, which is facing directly opposite to the minor road. A number of vehicles are queued up on the major road and the minor road due to some traffic jam (Figure 8). It may be noted that the probability of such a worst-case scenario is very low. The study is carried out assuming there is no fencing. The presence of fencing will provide additional shielding to a great extent due to the low BSS antenna installation height. 5.2 Maximum queue length The cars queuing up near the T-junction contribute to the aggregate interference. Maximum queue length determines the maximum number of cars to be considered that would contribute significantly to the interference. The interference power level varies as inverse square of the separation distance. The plot (Figure 9) shows that the interference from cars located 65 m away from the BSS receiver is 23 db than the car nearest to the BSS receiver. In general, any radar whose contribution is 23 db below the nearest radar (the one closest to the victim antenna) contributing the maximum, can add to interference by less than 0.5% and can be ignored. This factor determines the maximum queue

10 Interference effects of 24 GHz UWB automotive short-range radar to broadcasting and fixed satellite services 205 length. For this aggregate interference analysis, emissions from all SRR s mounted on all cars up to a range of 65 m are considered. 5.3 Shielding effect Shielding of electromagnetic waves by cars in the front provides significant attenuation thereby contributing to lower values of I SRR /N ratio. The closer the car, the higher is the shielding effect. This phenomenon is reported by CEPT in its report (Proceedings of the conference, 1992). The CEPT study includes measurements of attenuation due to shielding for various SRR-BSS separation distances and inter-car separation distance. The measurement results are modelled by a linear curve as shown below (Figure 10). With these parameters, the value of the separation distance is solved using the relation ship: h_ BSS tan tan = 8 a+ 2 d + 5+ a The above condition guarantees that attenuation due to shielding effect is at least 22 db. A plot of relative mounting height versus maximum separation between cars for which the shielding attenuation is 22 db using the previous equation is shown in Figure 12. Figure 12 BSS antenna relative height versus maximum separation distance (see online version for colours) Figure 10 Model representing shielding effect (see online version for colours) Measured Shielding Loss Shielding Model for Calculations a-ar [ ] The attenuation varies depending on the angles α and α r (Figure 10). Attenuation is highest ( 22 db), when the difference between the two angles is more than 8. For the aggregate interference scenario (Figure 11), the maximum separation distance between the cars, for which the shielding attenuation is 22 db, is obtained. This distance is used for interference calculations. Figure 11 Interference geometry showing shielding effect (see online version for colours) The following parameters are used for computing the maximum separation distance between the cars: car length = 5 m car height = 1.4 m distance of rear wheel from the crash guard = 3m (assumed). This means that the cars queued up on the minor road are completely shielded with an additional 22 db attenuation (as the BSS antenna is only mounted at a height of 0.5 m) and the interference due to radars from the cars behind the front of the queue is less than 22 db. The contribution from those following cars to aggregate interference can be considered negligible. 5.4 Results of aggregate interference of SRR to BSS in Ka-band For the aggregate interference, the radars mounted on cars are divided into group A, group B and group C (see Figure 8). Only radars belonging to these groups will contribute to the aggregate interference. One sensor on one side and front OR rear of the car will contribute to the aggregate interference. The locations of those sensors, which contribute to interference, superimposed with the minor and major roads of the T-junction are shown in Figure 13. The BSS antenna is located at origin. The minor road is along the y-axis. The other geometrical parameters of Figure 8 are: The size of crossing path is 8 m 8 m. Width of each lane is 4 m. istance of antenna from the footpath = 8 m and 5 m.

11 206 R.R. Makanaboyina et al. Figure 13 Computed locations of radars contributing to interference (see online version for colours) separation between the cars d = 1 m, 3 m and 5 m queue length of cars at T-junction is varied from 5 m to 65 m in steps of 1 m the distance of BSS antenna to the edge of the road is 5 m clutter from fencing = 10 db (considered as additional transmission loss) noise floor of BSS receiver = 95.6 dbm. SRR Sensors The results are shown in Figure 14. The aggregate interference value varies between 22.4 db to 25.2 db. The effect of sensors on rear side of the cars on aggregate interference is observed to be nil. Figure 14 Aggregate I SRR /N (see online version for colours) BSS Antenna The aggregate interference from SRR s at locations is computed for separation distance of 8 m and 5 m. All computations are done in MATLAB. The traffic scenario is assumed static at a T-junction. The traffic scenario is generated in MATLAB and no special traffic models/generators are used. The computed aggregate interference is given in Table 9. Table 9 Aggregate I SRR /N Antenna height (m) I SRR /N (db) in 21.4 to GHz I SRR /N (db) in to 22.0 GHz d = 8 m d= 5m d = 8m d= 5m Furthermore, both the 61.3 dbm/mhz for OOB and the 53.8 dbm/mhz SRR EIRP emission level are worst-case assumptions that represent the maximum value of the still decreasing slope. The average value in the band is less by 4 db compared to the maximum value and this reduces the computed I SRR /N in the previous tables by at least 4 db. 5.5 Results of aggregate I SRR /N calculation for varying queue length The following parameters are used for computation of aggregate interference: four sensors per car at each corner, inclination 45, AZ coverage 120 height of the BSS receiving antenna = 2 m Worst case: aggregated I SRR /N is 22.4dB for car separation of 1 m with queue length of 65 m. From Figure 14, one may conclude that there is a 2.4 db positive margin with respect to the I SRR /N limit of 20 db. According to the assumptions used in this study, 10 db fencing/shielding loss was taken into account. Without the fencing loss of 10 db, the result will be a negative margin of 7.6 db in the worst case of 1 m-car separation distance. With regard to the mitigation factors mentioned here one can easily conclude that BSS and SRR in the 24 GHz range can coexist with safety margin for the BSS of at least 10.7 db. 5.6 Other mitigation factors 1 Activity factor: The variable activity factor describes all the variable or unexpected deactivations of a SRR device compared to normal compliance measurement mode. This could be due to: 2 The vehicle could be turned OFF during congestion or a short stop and does not emit any radiation.

12 Interference effects of 24 GHz UWB automotive short-range radar to broadcasting and fixed satellite services The vehicular radar device uses only a part of the full ultra wideband (UWB) bandwidth. Some vehicular applications need either less object separation capability (which results in an occupied bandwidth below 1 GHz) or longer detection range (which requires higher emission power that is only available in the 24 GHz ISM band). epending on the chosen radar sensor concept, the different frequency usage modes may be selected via an electronic switch depending on the prevailing road scenario situation. 4 ual operation of a vehicular radar sensor between a wideband mode and a fixed carrier. The sensor is operated for a given time instant in a wideband mode and then switches to a oppler speed measurement for another time interval. The ratio of time-shared is not fixed but depends on the traffic scenario and its prevailing risk. 5 Sensor fusion with other technologies also reduces the number of active 24 GHz wideband radar sensors per vehicle. Possible combination partners are video or laser system or 77 long-range radar/79 GHz SRR. 6 Time of day dependent SRR usage. There are always time periods where almost no vehicles are on the road. 7 Low market penetration levels. 8 The complexity and unpredictability in the scenarios above makes it difficult to express a mathematical exact and universally valid variable activity factor. A probability approach based on both time and frequency coincidence of SRR activity with the EESS in orbit would be necessary to determine the variable activity factor. Even under worst case conditions the mitigation effect of the variable activity factors as listed under items 1 7 are in the order of 7 db. This variable activity factor further improves the margin of electromagnetic compatibility between SRR and BSS. 6 Conclusions The compatibility of SRR with Ku-band and Ka-band BSS, and Ku-band FSS is studied for single vehicle scenario. The interference to Ku-band FSS from SRR is found to be below 46 db for single vehicle scenario. The interference to Ku-band BSS from SRR is found to be below 22.7 db for single vehicle scenario. The interference calculations are done with I/N criteria specified in various ITU recommendations. The compatibility of SRR with BSS in 21 GHz is studied for two subbands. The two subbands are required as the SRR band overlaps with BSS 21 GHz allocation and the EIRP of SRR is different in the two subbands. Currently, no ITU recommendations are in force with regard to sharing between them. ITU report on acceptable C/N protection criteria for BSS is studied. Keeping in of view all the information from the available ITU reports, it is concluded that the minimum acceptable value of I SRR /N from SRR to BSS will be 20 db. This value is based on tolerable emission limits in-force for secondary services. This is an assumption for the study. The I SRR /N is calculated for single vehicle and multiple vehicle scenarios and analysed separately. For both the scenarios, the BSS antenna heights are varied from 0.5 m to 10 m. The separation distances are varied from 10 m to 65 m. The interference scenario for single vehicle scenario is based on an antenna mounted in front of the house near a road. The attenuation due to shielding from vegetation, fences, etc., is considered to be 10 db. For multiple vehicle scenarios, the extreme case of vehicles jamming up near a T-junction is used. The attenuation available from vegetation/fencing near the vicinity of the BSS antenna is considered for I SRR /N calculation. In the first subband 21.4 to GHz, the computations of I SRR /N ratio for single vehicle scenario indicate that full compatibility is given for all cases of BSS antenna mounting heights and all separation distances. Compatibility also exists for BSS protection margin of 20.0 db from OOB emissions of SRR. Compatibility was also shown to exist for a multiple vehicle scenario, with I SRR /N greater than 20.0 db. In the second subband to 22.0 GHz, the compatibility between SRR and BSS is given for all BSS mounting heights of 0.5 m and above. The probability of BSS antennas to be located 5 m or below is extremely low. Compatibility is also shown for separation distances of 10 m and above. Full compatibility is shown between BSS and SRR in this band for single vehicle and multiple vehicle scenarios. Variable activity factor contributes an additional margin of up to 7 db. Additional margin of 7.5 db is available if FCC mask is adopted. From the results, we conclude that SRR interference to BSS is negligible and compatibility exists between SRR and BSS. References raft ETSI European Standard EN v1.1.1 ( ) EMC and radio spectrum matters (ERM); short range devices; road transport and traffic telematics (RTTT); short range radar equipment operating in the 24 GHz band. Part 1: Technical Requirements and Methods of Measurements. ECC Report 23 (2003) Compatibility of automotive collision warning short range radar operating at 24 GHz with FS, EESS and radio astronomy, available at EN v1.2.1 (2001) EMC and radio spectrum matters (ERM); short range devices; road transport and traffic telematics (RTTT); technical characteristics and test methods for radar equipment operating in the 76 GHz to 77 GHz band, available at AAAAAAA. FCC (2002) Report on UWB permissions, 22 April. ITU-R BO (1991) Reference patterns for ES and satellite antennas for the BSS in the 12 GHz band for the associated feeder links in the 14 GHz and 17 GHz bands, available at

13 208 R.R. Makanaboyina et al. ITU-R BO1444 (2000) Protection of the BSS in the 12 GHz band and associated feeder links in the 17 GHz from interference caused by NON-GSO systems, available at ITU-R S.1428 (2001) Reference FSS earth-station radiation patterns for use in interference assessment involving non-gso satellites in frequency bands between 10.7 GHz and 30 GHz, available at AAAAAA. ITU-R S (1994) Carrier-to-interference calculations between networks in the FSS, available at AAAAAAA. Makanaboyina, R., Sadashivaiah, M., Bloecher, H.L., Rollmann, G. and Gopalakrishnan, K. (2004) Compatibility studies of 24 GHz short range radar with FSS and BSS Systems, MIKON-2004, Warsaw, Poland, May. Proceedings of the conference (1992) WARC-92, Malaga-Torremolinos, Spain, February March. Report ITU-R BO (1998) Considerations for the introduction of broadcasting satellite service high definition television systems, available at AAAAAA. Report ITU-R BO.2019 (1999) Interference calculation methods, available at AAAAAAA. Report ITU-R BO (1998) High definition television by satellite, available at RR123 ITU Radio Regulations (2004) Available at RR124 ITU Radio Regulations (2004) Available at Ulaby, F.T. and obson, M.C. (1989) Handbook of Radar Scattering Statistics for Terrain, Table E-7, p.257, Artech House.