DRM+ for local radio broadcast - measurements in the MHz band

Transcription

1 DRM+ for local radio broadcast - measurements in the MHz band Authors: Dipl.-Ing. Friederike Maier Institute of Communications Technology University of Hanover Germany Dipl.-Ing. Detlef Pagel State Media Authority Lower Saxony (NLM) Germany April 12, 2010

2 Contents 1 Contents List of abbreviations 1 1 Introduction DRM+ in the MHz band Measurements in the MHz band DRM+ System parameter Encoding Evaluation of the MHz band channel properties Impact of the receiver velocity on the DRM+ reception Flat fading System Setup Equipment Transmission content Measurement parameters Measurements Measurements in an urban environment Measurement of the coverage limit Measurements with 4-QAM Measurements with 16-QAM Measurements with 100 W ERP and 4-QAM Conclusion 17 References 18 A The transmit antenna 18 List of abbreviations BER FAC ICI MSC OFDM PRBS RSNR RSTA SDC Bit Error Rate Fast Access Channel Intercarrier Interference Main Service Channel Orthogonal Frequency Division Multiplex Pseudo Random Bit Sequence Receiver Signal to Noise Ratio Receiver Status Information Service Description Channel

3 1 Introduction 2 1 Introduction DRM+ is an enhancement of the existing DRM (Digital Radio Mondial) standard up to the MHz band. It has been approved in the ETSI DRM standard [1] in 2009 as DRM Mode E for frequencies up to 174 MHz. 1.1 DRM+ in the MHz band With Digital Audio Broadcasting(DAB 1, DAB+) from EUREKA 147 Family, a high-performance digital broadcast system is available. Due to Regional Radiocommunication Conference 2006 (RRC 06), the possibility to use DAB in Region 1 in frequency the MHz band is given. This digital broadcast system enables the transmission of approximately 8 to 20 (depending on coding MPEG 2 or MPEG 4 AAC V2) programs in one multiplex over a 1.5 MHz channel so it can be eciently deployed in large areas such that all broadcast programs sent in one multiplex have the same coverage area. On the other hand, narrow-band broadcast systems like DRM+ are more suitable for local and regional broadcast. In one DRM+ multiplex, up to four programs can be transmitted over a 100 khz channel. This oers the possibility for each broadcaster to have his own coverage area independently of other broadcasters, and an ecient local/regional frequency broadcast can be implemented. What is the situation of the analogue FM radio utilisation? In Germany and Europe the main radio distribution path, the MHz band, oers a large number of programs to the listeners. However, this band is heavily occupied in Central Europe and it is in general not possible to provide any additional broadcasts, analogue or digital, without switching o the existing ones. Hence, the broadcast of digital local and regional programs using the digital radio system DRM+ would be a valuable option. Following this approach it would be necessary to extend the frequency range dedicated to the DRM+ System including the MHz band. 1.2 Measurements in the MHz band To analyse the system performance in the MHz band, a eld trial was carried out within the scope of the pilot project "Digital radio broadcast for local area with the system DRM" which is carried out between the State Media Authority Lower Saxony (NLM) and the Institute of Communications Technology of the University of Hanover. The measurements were made in the city of Hanover, Germany and its sourroundings in winter/spring 2009/10. This report contains a description of the DRM+ system parameters, an evaluation of the channel properties in the MHz band, the system setup and equipment that was used in the trial and the measuring results that were optained in the measuring campaign. 1 Digital System A, ITU-R BS.1114

4 3 Evaluation of the MHz band channel properties 3 2 DRM+ System parameter The DRM+ system parameter are shown in the following table: System parameter Modulation OFDM Data rate kbps Subcarrier modulation 4-/16-QAM Signalbandwith 96 khz Subcarrierspread Hz Number of subcarriers 213 Symbol duration 2.25 ms Guard interval duration 0.25 ms Frame length 100 ms Number of programs Encoding In DRM Mode E with 4-QAM the MSC (Main Service Channel), which contains the user data, has the following protection levels with the corresponding code rates and bit rates: MSC: 4-QAM Protection level Code rate Bit rate [kbit/s] In DRM Mode E with 16-QAM the MSC uses multilevel coding and has the following protection levels with the corresponding overall code rates and bit rates: MSC: 16-QAM Protection level Code rate Bit rate [kbit/s] The SDC (Service Description Channel), which contains signaling data, is modulated with 4-QAM uses the following code rates: SDC: 16-QAM Code Rate The FAC (Fast Access Channel) uses a x code rate of R = Evaluation of the MHz band channel properties As the System was developed for the frequency range from MHz the following section gives an evaluation of the parameter, which could lead to problems with higher transmitting frequencies.

5 3 Evaluation of the MHz band channel properties Impact of the receiver velocity on the DRM+ reception As reception often takes place with moving receiver (in a car for example), the speed at which reception is possible is an important parameter. A moving receiver causes doppler shifts of the OFDM carriers. If this is combined with multipath propagation, paths from dierent directions can cause frequency dependent doppler shifts which results in Intercarrier Interference (ICI). This Interference can be handled as additional near- Gaussian noise [2]. In [3] the upper bounds of the noramlized interference power for a classical channel model over the maximum dopplershifts are given (gure 2 (a)). Additionally the maximum doppler frequency f dopplermax depends on the pilot grid [4]. Figure 1 shows the DRM+ pilot grid. Figure 1: DRM+ pilot grid Considering a carrier spacing f = 444, 44Hz and symbol duration T s =2,5 ms. In time direction the channel will be measured every 4 T s = 10ms, so the channel is sampled at a sampling frequency of 100 Hz. To fullll the sampling theorem the maximum doppler frequency has to fullll the condition: f dopplermax < 50Hz. Figure 2 (b) shows the doppler shifts at dierent carrier frequencies and the limit due to the pilot grid. It shows, that a receiver velocity of 200 km/h can result in dopplershifts up to 42 Hz for a carrier frequency of 230 MHz, up to 33 Hz for a carrier frequency of 170 MHz and up to 19 Hz (a) Normalized ICI power upper bounds (b) Doppler shifts Figure 2: Impact of ICI on an OFDM signal, depending on receiver velocities and carrier frequencies

6 4 System Setup 5 for a carrier frequency of 100 MHz. The resulting normalized power are db for 230 MHz, -20 db for 170 MHz and -24 db for 100 MHz. So this additional noise has to be considered and simulations and eld trials with higher receiver velocities should be made. 3.2 Flat fading The coherence time describes the variation of the channel over the time, a slowly changing channel has a long coherence time, a slowly changing channel a short one. A fast receiver passes the interference pattern of the electromagnetic waves faster, which results in a shorter coherence time. A shorter wavelength results in a higher resolution of the interference pattern in the air an so also in a shorter coherence time. In the DRM+ system bit interleaving is carried out over one frame (100 ms) and cell interleaving over 6 frames (600 ms). For low receiver velocities in band II at fading resulting in signal dropouts occurs because the coherency time/deeps fades lasts longer than the cell interleavers time, so there is no chance to recover the signal with the following error correction. As in the MHz band the wavelength is shorter than in band II, there should be less dropouts due to at fading. Further analyses of the channel properties will be conducted by channel measurements. 4 System Setup The transmitter was located at the University of Hanover, the transmitting antenna was mounted on the roof of the University building (Appelstr.9A, Hannover, GPS: lat: , long: 9.712) at a high of 70 m above ground. Transmitting power is licenced up to 100 W ERP at a frequency of MHz. Most tests were made with 30W ERP. Tests were made with one very roboust 4-QAM modulation and one 16-QAM modulation with hight data rates. The details can be found in the following tables: MSC: 4-QAM Protection level Code rate Bit rate [kbit/s] MSC: 16-QAM Protection level Code rate Bit rate [kbit/s] The 16-QAM MSC Mode oers two SDC encodings, we used the more roboust one. SDC: 16-QAM Code Rate 0.25 For this encodings simulation results of the required Signal to Noise Ratio (SNR) values are available in the DRM ETSI standard [1] in Annex A. 4.1 Equipment The equipment used for the measurements is the following: Fraunhofer Contentserver

7 4 System Setup 6 RFmondial Modulator/Exciter Thomson linear Amplifer (VHF Amplier Band 3, 54 db gain) Transmit antenna: Kathrein K , 5- Element Yagi directional antenna (direction: 120 ) Receive antenna: Kathrein K / BN , Monopole, an antennafactor of = 13 db was measured, the antenna was mounted on the roof of a van at a height of around 2 m HF-Frontend: Rhode & Schwarz ESVB Measuring receiver, 10.7 MHz IF Field strength measurements: ESVB, BW: 300 khz A/D converter Perseus RFmondial Software Receiver Figure 3 shows the DRM+ transmitter. Figure 3: DRM + transmitter for MHz

8 5 Measurements Transmission content The transmission consisted in both modes in an AAC encoded stereo audio stream and a synchronous pseudo random bit sequence (prbs) which was used for the calculation of the bit error rate (BER) (see [5] for details). 4.3 Measurement parameters The following parameters were recorded and analysed during the measurements: Field strength (measured with Rhode & Schwarz ESVB Measuring receiver, BW: 300 khz, RMS) triggered via GPIB every 16th frame (1.6 sec) GPS coordinates Bit error rate Signal to noise ratio (calculated via the time correlation/syncronisation, see [6]) Receiver status information (RSTA), the status of the audio decoding is evaluated, it shows if one or more audio frames are corrupted within one DRM multiplex frame, see [5]) 5 Measurements Figure 4: Measuring locations and routes (mapsource: Bundesamt für Kartographie und Geodäsie)

9 5 Measurements 8 Measurements were conducted with 100 W ERP towards the south and with 30 W ERP in an urban area in the city of Hanover and on one radial route in the direction of the main beam as shown in gure Measurements in an urban environment To analyse the behaviour in an urban environment measurements were conducted in the streets behind the main station of Hanover where a kind of dense urban environment with high buildings and small streets can be found (see gure 5). The average distance to the transmitter is around 2.5 km. Measurements were made in 16-QAM Mode (Coderate 0.5). The rst measurements were conducted with horizontal antenna polarization. As the receiving antenna is vertically mounted on the roof of a car, the antenna was rotated to vertical polarization and the measurement results were compared. The dierent antenna patterns have to be taken into account, but as the location behind the railway station lies in the main beam there should be no big dierence. The measurements of the horizontal polarized transmission were conducted on the 19/01/2010. There was some snow and a lot of ice on the ground. The vertically polarized measurements were conducted the 4/3/2010. There was mostly overcast and little snow on the ground. Figure 5: Pictures of the environment behind the railway station Figure 6 shows the eld strength distribution for the horizontal an vertical polarized transmission. For horizontal transmission the variability of the eldstrength is between 20 and above 50 dbµv/m, for vertical transmission it is mostly above 40 dbµv/m. Figure 7 shows a plot of the receiver status information (rsta) which describes whether all audio frames are ok (green) or one or more audioframes are corrupted (red). For horizontal polarization there are some errors in the reception, for vertical polarizazion the reception was almost errorfree.

10 5 Measurements (a) Horizontal polarization 9 (b) Vertical polarization Figure 6: Field strength in an "dense" urban environment (mapsource: Bundesamt für Kartographie und Geodäsie) (a) Horizontal polarization (b) Vertical polarization Figure 7: Measurement of the audio data in the urban environment (mapsource: Bundesamt für Kartographie und Geodäsie) The gures 8 and 9 o er a more detailed view on the reception parameter. They show a comparision of the eld strength in dbµv /m, the receiver Signal to Noise Ratio (RSNR) in db, the BER and the RSTA (0: all audioframes ok, 1 one or more audio frames corrupted) over the frames (one frame corresponds to 100 ms, for the horizontal polarization only every 16th frame is plotted). While for horizontal polarization the eldstrength is mostly under 50 dbµv /m an SNR is around 20 db with vertical polarization it reaches 60 dbµv /m and an SNR of around 25 db. If the SNR falls below 20 db audioerrors occured.

11 5 Measurements 10 Figure 8: Measurement results in a "dense" urban environment with horizontally polarized transmission Figure 9: Measurement results in a "dense" urban environment with vertically polarized transmission 5.2 Measurement of the coverage limit As vertical polarization worked much better, the coverage limit was tested with vertical polarization. Measurements were conducted in the main beam of the transmitting antenna with 16- and 4-QAM modulation. The measuring day was also the 4/3/2010. The route was chosen on the B65, a rural road lying mostly in the main beam of the transmission. The measurement was continued until audio quality became bad.

12 5 Measurements Measurements with 4-QAM Figure 10 shows the eldsstrength measured on the route. Figure 10: Fieldstrength measurment of the coverage limit 4 QAM Mode (Coderate: 0.33) (mapsource: Bundesamt für Kartographie und Geodäsie) Figure 11 shows the audio status measured on the route. Here it can be seen that with more distance to the transmitter, passing the villages, the audiostate becomes errornous, whereas passing the countryside with almost no obstacles (the countryside in Hannover is quite at) receiving quality is still ok. Figure 12 shows the reception parameter over the frames in the direction away from the transmitter. On the radial route reception was possible down to a eldstrength of around 35 dbµv/m with an SNR of around 12 db. We stopped the measurement at a distance of around 30 km from the transmitter, were audio quality became bad also in the free countryside.

13 5 Measurements 12 Figure 11: Measurment of the coverage limit in 4 QAM Mode (Coderate: 0.33), green: audioframes ok, red: one or more audioframes corrupted (mapsource: Bundesamt für Kartographie und Geodäsie) Figure 12: Reception parameter on the B65 with 30 W ERP and 4 QAM Mode (Coderate: 0.33) (mapsource: Bundesamt für Kartographie und Geodäsie)

14 5 Measurements Measurements with 16-QAM On the way back the 16-QAM mode was measured. Figure 13 shows the eldstrength measured then. Figure 13: Fieldstrength measurment of the coverage limit 16 QAM Mode (Coderate: 0.5) (mapsource: Bundesamt für Kartographie und Geodäsie) Figure 14: Measurment of the coverage limit in 16 QAM Mode (Coderate: 0.5), green: audioframes ok, red: one or more audioframes corrupted (mapsource: Bundesamt für Kartographie und Geodäsie) Figure 14 shows a plot of the receiver status information (rsta). The measurement was stopped at

15 5 Measurements 14 Figure 15: Reception parameter on the B65 with 30 W ERP and 16 QAM Mode (Coderate: 0.5) (mapsource: Bundesamt für Kartographie und Geodäsie) a distance of around 15 km from the transmitter. In gure 15 the eldstrength is plotted in comparision to the receiver Signal to Noise Ratio (RSNR), the Bit Error Rate (BER) and the Receiver Status Information (RSTA) over the frames. It shows that the reception was ok down to a eldstrength of around 48 dbµv/m at a calculated SNR of around 20 db. 5.3 Measurements with 100 W ERP and 4-QAM Another measurement took place on 12th January 2010 in the surroundings of Hanover to test the coverage limit. This day was blue sky, but a lot of snow lying on the ground. This trial was with horizontal polarization and with a power level of 100W ERP. However the measurement route was directly to the south (azimuth: 180 ), so 60 away from the main beam where the E-plane of the antenna (see appendix A) shows a gain of -10 db (so there are 10 W in this direction).

16 5 Measurements 15 Figure 16: Measurment of the eldstrength with 100 W ERP and 4 QAM Mode (Coderate: 0.33) (mapsource: Bundesamt für Kartographie und Geodäsie)

17 5 Measurements 16 Figure 17: Measurment of the coverage limit with 100 W ERP and 4 QAM Mode (Coderate: 0.33), green: audioframes ok, red: one or more audioframes corrupted (mapsource: Bundesamt für Kartographie und Geodäsie)

18 6 Conclusion 17 Figure 18: Reception parameter with 100 W ERP and 4 QAM Mode (Coderate: 0.33) Figure 16 shows the eldstrength measured on the trial, gure 17 shows a plot of the receiver audio status information. The reception was possible with short dropouts up to a distance of around 20 km. The comparision of the two plots and gure 18 with the reception parameters over the frames show that with a eldstrength value of more than around 35 dbµv/m reception is possible without dropouts. 6 Conclusion Tests were made at a frequency of MHz in an urban surrounding and on radial routes passing through the city of Hanover and rural environments. Measurements of the eldstrength, the bit error rate, the calculated signal to noise ratio and the audio status show that in 4-QAM mode with a coderate of 0.33 reception with good audio quality was possible down to a eldstrength of around 35 dbµv/m and a calculated SNR of 12 db. In the 16-QAM mode reception was possible down to around 48 dbµv/m at an SNR of around 20 db. The required SNR values are a bit higher than in the simulation results in [1] (7.3 db for 4-QAM with an urban channel model at 60 km/h and 15.4 for 16-QAM). However the simulations were made with optimal channel estimation and implementation losses are not considered. With an ERP of 30 W reception dropped out in the countryside at a distance of around 30 km in 4-QAM mode and at around 15 km in the 16-QAM mode. Thanks to the NLM (State Media Authority Lower Saxony), RFmondial, Thomson, the Bundesnetzagentur (BNetzA/Federal Network Agency), the DRM Consortium and Fraunhofer IIS and many others for their support and good advices.

19 A The transmit antenna 18 References [1] ETSI. ES , Digital Radio Mondiale (DRM), System Specication [2] P. Robertson and S. Kaiser. The eects of doppler spreads in OFDM(A) mobile radio systems. IEEE Vehicular Technology Conference, [3] Y. Li and L. Cimini. Bound on the interchannel interference of ofdm in time-varying impairments. IEEE Transactions on Communications. [4] H. Schulze and C. Lüders. Theory and Applications of OFDM and CDMA Wideband Wireless Communications. Wiley, [5] ETSI. TS , Digital Radio Mondiale (DRM), Receiver Status and Control Interface (RSCI) [6] K. Ramasubramanian and K. Baum. An OFDM timing recovery scheme with inherent delayspread estimation. IEEE Global Telecommunications Conference, A The transmit antenna The transmit antenna is mounted horizontally and afterwards vertically on the roof of the university building in the Appelstr. 9A in Hanover. Direction is to an azimuth of 120. On the next page, the data sheed including the antenna pattern and gain can be found.

20 Directional Antenna MHz K element broadband Yagi antenna of weather-proof aluminum. Component for low power transmitting antennas. Type No. for input 7-16 female K K N female K K Frequency range MHz MHz (ch = channel) (ch 5 ch 8) (ch 9 ch 12) Dimensions A 930 mm 810 mm B 885 mm 765 mm Gain (ref. to λ/2-dipole) 6 db VSWR < 1.15 Impedance 50 Ω Polarization Horizontal Max. power 100 Watt (higher power upon request) Weight 5 kg Wind load (at 160 km/h) frontal: 114 N 102 N lateral: 102 N 91 N Max. wind velocity 225 km/h Packing size 970 x 200 x 135 mm Lengths see table A B Material: Mounting: Grounding: Combinations: Special features: Weather-proof aluminum. Radiator in fiberglass radome. To pipes of mm by means of mounting clamps, supplied. Via mounting parts. The antenna is especially suitable as a component in arrays to achieve various radiation patterns. The antenna will be shipped dismounted. Radiation Patterns (at mid-band) db 55 in E-plane Horizontal Radiation Pattern db 110 in H-plane Vertical Radiation Pattern 936.A958 Subject to alteration. KATHREIN-Werke KG. Anton-Kathrein-Straße 1 3. P.O. Box D Rosenheim. Germany. Telephone (++49)8031/ Fax (++49)8031/