Transmission System for Low -Latency Digital Radio Microphones

Transcription

1 Transmission System for Low -Latency Digital Radio Microphones Regulators in Japan have stipulated that the frequency band used for certain (licensed) radio s (77-86 MHz) should be changed. Moreover, the digital radio s affected by this change have a latency of -5 ms, which presents problems compared with using analog devices. We have studied the use of OFDM modulation, which has superior characteristics in multipath conditions, and have developed a low-latency digital transmission system for radio s operating in the new frequency bands. The system enables high-quality and stable transmission of radio audio signals. In this article, we describe the transmission system studied, including the techniques used to achieve low-latency transmission. 1. Introduction Among the specified radio s *1 currently available, the analog type, which transmits the audio signal using frequency modulation (FM), is in wide use. Analog specified radio s have extremely low latency, compactness, and low power consumption. However, they require a high carrier-to-noise ratio (CNR) and are quite susceptible to environments with noise or interference. The number of s that can be used simultaneously is also limited by the effects of intermodulation distortion. In contrast, digital radio s have advantages including resistance to noise and interference, high sound quality, and the ability to use more of them at the same time. Their main disadvantage is transmission latency ranging from to 5 ms. This can be an obstacle for users, depending on the environment and conditions of use. Meanwhile, the Frequency Reorganization Action Plan 1) released in September, 211 requires that specified radio s migrate from the frequencies they are currently using, which are between 77 and 86 MHz. To enable stable, low-latency transmission of radio signals at the new frequencies *2, we have proposed a low-latency specified digital radio *1 Indicates a type of using radio which requires licensing as an on-land radio station. Also referred to as Type- A radio s. *2 1,24-1,26 MHz (excluding 1,252-1,25 MHz), digital terrestrial television broadcast white space (47-71 MHz), and MHz bands. transmission system that transmits uncompressed audio signals using orthogonal frequency-division multiplexing (OFDM), which has excellent resistance to multipath interference 2)). In this article, we report on this transmission system and the techniques used to achieve low-latency transmission. 2. Transmission System for Low-latency Specified Digital Radio Microphones 2.1 Requirements The requirements for digital wireless s specified in a report from the International Telecommunication Union, Radio Communication Sector (ITU-R) 4) are shown in Table 1. According to these requirements, digital wireless s for studio use must cover audio frequencies from 2 Hz to 2 khz, have a dynamic range of at least 1 db, and have a maximum latency of 1 ms. Moreover, in-ear monitors have sound-quality requirements as to their frequency and dynamic range, as well as a transmission latency of 1 ms or less. The European Telecommunications Standards Institute (ETSI) has also issued specifications on the characteristics of wireless s in the 25 MHz to GHz band 5). In particular, they limit the transmission bandwidth to a maximum of 2 khz for frequencies up to 1 GHz, and a maximum of 6 khz for frequencies over 1 GHz. Considering these requirements for digital specified radio s, we placed the following conditions on our low-latency digital specified radio transmission scheme. - Transmission latency of 1 ms or less. - Transmission bandwidth of 6 khz or less. - Ability to transmit uncompressed audio signals. - Ability to transmit over distances up to 1 m. 2.2 Transmission Modes Considering the variety of scenarios in which radio s are used, the low-latency specified radio transmission method must have high quality, and be resistant to interference, so we decided to provide the following three transmission modes. (The modes, together with their transmission parameters are shown in Table 2.) Table 1: Digital wireless requirements (from ITU-R BS.2161 report) Use Audio freq. characteristic Audio dynamic range Max. allowed latency Studio use 2 Hz to 2 khz 1 db or greater 1 ms In-ear monitor 2 Hz to 15 khz 95 to 1 db 1 ms Broadcast Technology No.56, Spring 214 C NHK STRL 1

2 (1) Standard transmission mode The standard transmission mode transmits high-quality sound using linear pulse-code modulation (linear PCM) with a sampling rate of 48 khz and 24-bit quantization. The sound signal is neither compressed nor decompressed, so latency is very low. The data rate is 1,152 kbps. (2) Interference-tolerant transmission mode The interference-tolerant transmission mode applies instantaneous companding to the sampled digital data to reduce the data rate. Instantaneous companding causes a certain amount of distortion that can be perceived by human hearing, but it is able to reduce the amount of data with very low latency. Typical instantaneous compression rules include the A-law * and μ-law 6)*4, which approximate a log characteristic. This transmission mode allows some degradation in sound quality, uses QPSK carrier modulation for resistance to noise and interference, and is useful when long-distance transmission is necessary. The data rate is 576 kbps. () In-ear monitor transmission mode The in-ear monitor transmission mode is for * An algorithm for efficiently reducing dynamic range. Used mainly in Europe for digital communications. *4 An algorithm for efficiently reducing dynamic range. Used mainly in North America and Japan for digital communications. Provides slightly higher dynamic range than A-law. giving instructions, asking questions, or transmitting audio to performers on stage. This mode requires stereo transmission, so the data rate is reduced with instantaneous companding and two audio channels are transmitted. The total data rate is 1,152 kbps. 2. Transmission Parameters Specified radio s are used in environments, such as studios and halls, that can have a lot of multi-path interference (reflections). We used OFDM modulation to reduce these effects because it has excellent characteristics in such environments. We selected as short a symbol length as possible to reduce latency, and by making it an integer multiple of the uncompressed audio signal sample length (8. μs), we were able to minimize the buffering time of the signal processing. The guard interval length was set to 4.9 μs, which is large enough relative to the multi-path latency. The effective symbol length is 78.4 μs, and the carrier interval is khz. The total number of carriers is 46, and the total transmission bandwidth is khz. For bit interleaving, bits are rearranged only within a single OFDM symbol, which also reduces buffering time. 2.4 OFDM Frame Structure The OFDM frame structure for the low-latency digital specified radio transmission system is shown in Figure 1. Scattered pilots (SP) are introduced once every 15 carriers in the frequency direction, and once every 5 symbols in the time direction. Transmission and multiplexing configuration controls Table 2: Low-latency specified digital radio transmission scheme modes and transmission parameters Mode (1) Standard (2) Interference-tolerant () In-ear monitor Data source coding Transmission path coding Analog audio signal Quantization Sampling rate Data compression Transmitted bits Data rate Parity Error correction code, code rate Carrier modulation Secondary modulation OFDM symbol length Effective symbol length Guard interval length (GI ratio) Carrier interval No. Carriers Total Data SP TMCC CP Transmission bandwidth Interleaving Mono 24-bit Uncompressed 24 bits 1,152 kbps 16QAM Mono Stereo 24-bit 24-bit, 2 channels 48 khz Instantaneous companding Instantaneous companding 12 bits 12 bits, 2 channels 576 kbps 1,152 kbps CRC-2 Cyclic Redundancy Check - 2 Convolution code, 2/ QPSK 16QAM OFDM 8. µs 78.4 µs 4.9 µs (1/16) khz khz Bit, frequency 14 Broadcast Technology No.56, Spring 214 C NHK STRL

3 1 Feature Frequency (carrier no.) 2 Time (symbol no.) Data SP TMCC Figure 1: OFDM frame structure (TMCC) are introduced at carrier numbers 2, 22, and 4. A continual pilot (CP) is positioned at the right end of the bandwidth, and an OFDM frame is composed of 6 OFDM symbols.. Transmission Characteristics and Transmission Latency We checked the transmission characteristics of the system through computational simulation. The simulation system is shown in Figure 2. A 24-bit pseudo-random noise (PN) signal was generated at the transmitter, to which two-bit cyclic-redundancycheck (CRC) parity bits were added and convolution coding applied. The coding rate was 2/. After that, bit interleaving, mapping, and frequency interleaving were performed, and TMCC, SP, and CP data were added to create frames. Finally, the transmission signal was generated by applying an inverse fast Fourier transform (IFFT) and adding a guard interval (GI). We assumed the receiving side would use four-branch spatial diversity, so the simulation included maximal ratio combining *5 of four branches. Channel state information (CSI) based on the subcarrier amplitude value was used to apply weightings for Viterbi decoding, and the CRC parity check was done after the Viterbi decoding. *5 A method which applies weightings to two input signals before combining them, in such a way as to maximize the signal-to-noise ratio (SN ratio) of the combined signal. Generate PN No. Add metadata Convolutioncoding Bitrotation Transmitter antena Mapping Frequency interleaving Frame structure Add GI antena Generate TMCC, SP, CP FFT Complex division Max. ratio combining Pilot removal Frequency deinterleave Demapping Extract SP CSI Reverse bit rotation Weighting Viterbi decoding Figure 2: Simulation system Broadcast Technology No.56, Spring 214 C NHK STRL 15

4 .1 Transmission Characteristics with Additive White Gaussian Noise Figure shows the relationships between per-branch received CNR and bit error rate assuming the (1) standard and () in-ear monitor transmission modes in an additive white Gaussian noise (AWGN) environment. The required CNR was taken to be one for a bit error rate of 1 1-5, and it was 1.8 db for singlebranch reception, 1.8 db for two-branch reception, and 8. db for four-branch reception. These results confirm that four-branch reception provides an improvement of 5.8 db in required CNR compared with single-branch reception. The bit-error rate characteristics for the (2) interference-tolerant transmission mode in an AWGN environment are shown in Figure 4. In this case, the required CNR for single-branch reception was 7.5 db, 4.7 db for two-branch reception, and 2. db for four-branch reception. These results confirm that fourbranch reception provides a 5.5 db improvement in required CNR branch rec Figure : Bit error rates for standard and in-ear monitor transmission modes (16QAM, AWGN environment) branch rec Figure 4: Bit error rates for interference-tolerant transmission mode (QPSK, AWGN environment).2 Transmission Characteristics with Fading Next, we simulated a six-wave Rayleigh fading *6 environment to examine mobile reception with specified radio s. The multiple paths used and the desired-to-undesired ratios (DUR) are shown in Table. The maximum Doppler frequency was set at 2 Hz. *6 The phenomenon in which the received signal level fluctuates under conditions in which the transmitted signal is not received directly. Table : Multipaths used in Rayleigh fading environment simulations Path No branch rec Figure 5: Bit error rate characteristics for standard and in-ear monitor transmission modes (16QAM, Rayleigh fading environment) Delay [ns] DUR [db] branch rec Figure 6: Bit error rate characteristics for interference-tolerant transmission mode (QPSK, Rayleigh fading environment) 16 Broadcast Technology No.56, Spring 214 C NHK STRL

5 Feature The bit-error-rate characteristics for the (1) standard and () in-ear monitor transmission modes in a Rayleigh fading environment are shown in Figure 5, and those for the (2) interference-tolerant transmission mode are shown in Figure 6. The required CNR was 18.4 db for the standard and inear monitor transmission modes (four-branch reception), and 1.6 db for the interference-tolerant transmission mode (four-branch). These results confirm that four-branch diversity combining is particularly effective in fading environments.. Transmission Latency We prototyped equipment incorporating the three transmission modes shown in Table 2. Photographs of the transmitter and receiver are shown in Figure 7. The transmitter in Figure 7 (a) is 9 cm 11 cm 4 cm-small enough to attach to a person s body-and can transmit at up to 5 mw in the 1.2-GHz band. The receiver in Figure 7 (b) is equipped with four RF input circuits and a diversity reception functionality able to perform maximal-ratio combining between the four branches for each subcarrier. The prototype test transmitter and receiver were connected by cable, and the analog audio signal input to the transmitter, together with the analog output signal from the receiver, were observed on an oscilloscope to measure the latency. Figure 8 shows these time waveforms when using the standard transmission mode as an example, and Table 4 lists the latency measurements for each transmission mode. These Table 4: Transmission latency with prototyped test equipment for each transmission mode (1) Standard (16QAM).8 ms (2) Interferencetolerant (QPSK).88 ms () In-ear monitor (16QAM).81 ms measurements show that transmission latencies less than.9 ms were achieved for all transmission modes. This result fulfills the requirement that transmission latency be under 1 ms for low-latency digital radio s. 4. Conclusion We have devised a low-latency digital radio transmission scheme and confirmed its transmission characteristics through computational simulation. The simulations indicated that good results can be obtained in AWGN and fading environments by using four-branch diversity-combining reception. Test equipment was prototyped to verify the transmission method being studied, and measurements of the transmission latency showed that a very low latency,.8 ms for the uncompressed audio transmission modes, was achieved. In the future, we will contribute to the frequency migration of specified radio s by implementing practical low-latency digital radio s. This article was created by making revisions and corrections to the following article in the ITE Journal: Taguchi, Nakamura, Iai, Okano, Hamazumi: A Study of Low Delay Digital Transmission for Specified Radio, ITE Technical Report, Vol. 6, No. 51, BCT212-11, pp (212) (Japanese). (Makoto Taguchi) (a) Transmitter (b) Figure 7: Prototype transmitter and receiver Transmitter input audio signal.8ms output audio signal References 1) MIC Publication: Announcement of the Frequency Reorganization Action Plan (Sept. 211 Ed.), (211) 2) Hamazumi, Nakamura, Iai, Okamoto: Initiatives for a New Transmission System for Digital Radio Microphones - Low-latency OFDM transmission, ITE Technical Report, Vol. 6, No. 15, BCT212-51, pp (212) (Japanese) ) Nakamura, Okamoto, Iai, Hamazumi: A Lowlatency Digital Radio Microphone System, ITE Annual Convention, 2-1 (212) (Japanese) 4) ITU-R BS.2161 Report, Low Delay Audio Coding for Broadcasting Applications, (29) 5) ETSI EN v1..2, Electromagnetic Compatibility and Radio Spectrum Matters (ERM); Wireless Microphones in the 25 MHz to GHz Frequency Range; Part 1: Technical Characteristics and Methods of Measurement, (28) 6) Nikaido, Yamazaki: Digital Audio for the Sound Engineer, Kenrokukan Publishing (Japanese) Figure 8: Example audio signal time waveform (standard transmission mode) Broadcast Technology No.56, Spring 214 C NHK STRL 17