1.1 8K Super Hi-Vision format

Transcription

1 NHK STRL is researching a wide range of technologies for 8K Super Hi-Vision (SHV), including video formats and imaging, display, recording, audio, coding, media transport, content protection and transmission systems. We are looking ahead to the start of the regular broadcasting and widespread use of SHV and future broadcasting services beyond SHV. In our research on video formats, we developed a system for the simultaneous production of high dynamic range (HDR) and standard dynamic range (SDR) video and conducted demonstration experiments on HDR live program production. In our work on imaging, we developed a full-featured SHV camera that is compliant with ITU-R Recommendation BT.2100 and supports a 120-Hz frame frequency, 8K full resolution, a bit depth of 12 bits and HDR. We also developed a full-resolution single-chip color camera using a 133-megapixel image sensor that can operate at a 120-Hz frame frequency in interline transfer. We designed a 1.25-inch, full-featured 8K image sensor with 33 megapixels that supports a 240-Hz frame frequency and a bit depth of 14 bits. In our work on displays, we reduced the size of our 9.6-inch monitor by separating into display unit and control unit. We also improved the image quality of our projector by using high-power laser light sources that are twice as bright as conventional equipment, while halving the speckle noise that occurs due to interference of the laser light. For the future 8K display in home use, we developed a 130-inch sheet-type display by combining four thin 4K organic light emitting diode (OLED) panels and demonstrated a future living space at the NHK STRL Open House In our work on recording, we developed a compact memory package and extended the functionality of our compression recorder. We also succeeded in real-time compression of 8K video in 4:2:0 at a 240-Hz frame frequency for the high-speed capture of SHV video. In our work on audio, we developed an adaptive downmixing technique to generate high-quality stereo or 5.1 ch sound signals through the signal processing of 22.2 ch audio signals for the simultaneous production of program audio. We also developed a software-based codec for MPEG-H 3D audio with an eye toward 22.2 ch sound in next-generation terrestrial broadcasting. For the reproduction of 22.2 ch sound, we researched binaural reproduction using line array loudspeakers integrated with a display. Regarding video coding, we investigated the required bit rates for 8K/120-Hz video using High Efficiency Video Coding (HEVC) and started codec development. We also developed elemental technologies for an advanced coding format for next-generation terrestrial broadcasting and proposed a way of improving intra-prediction at an international standardization meeting. Regarding media transport technologies, we investigated the application of MMT for the IP delivery of 8K content and the synchronized presentation of multiple pieces of content. We also researched MMT technologies for next-generation terrestrial broadcasting, which include IP packet multiplexing and an IP transmission scheme for STL/TTL to enable a SFN. In our work on content rights protection and conditional access, we contributed to the standardization of the second-generation conditional access system. Our effort led to a revision of the ARIB Technical Report (ARIB TR-B39) and the addition of specifications for combined receivers. Regarding satellite broadcasting transmission, we worked on the establishment of a new ITU-R Recommendation for ISDB-S3(BO.2098), a transmission system for advanced wide-band satellite broadcasting. We investigated multilevel coded modulation and a way of compensating for nonlinear distortion to increase the capacity and transmission performance of a 12-GHz-band broadcasting satellite. We also researched an array-fed shaped-reflector antenna for a 21-GHz-band satellite broadcasting system as a new satellite channel and a dual-band antenna to receive both 12-GHz and 21-GHz satellite broadcasting. Regarding terrestrial broadcasting transmission, we prototyped a modulator and demodulator that supports hierarchical transmission in which services for fixed reception and those for mobile reception are multiplexed into a single channel. During the Rio Olympic Games, we demonstrated the world s first real-time 8K terrestrial transmission using a 60-Hz HEVC real-time codec in Rio de Janeiro and Tokyo simultaneously. In our work on wireless transmission technologies for program contributions, we researched field pickup units (FPUs) that use the 6/6.4/7/10/10.5/13-GHz band (microwave band) and 42/55-GHz band (millimeter-wave band) with the aim of enabling SHV live broadcasting of emergency reports and sports coverage. We also worked on the standardization of these FPUs. For the purpose of SHV mobile relay broadcasting, such as road race coverage, in the 1.2-GHz/2.3-GHz band we investigated bidirectional adaptive control and a rate-matching technique that improves reliability by adaptively controlling the coding rate of error correction codes according to the varying channel quality. In our work on wired transmission technologies, we researched video synchronization and equipment control technologies to develop IP-based program production and program contribution systems. Regarding our channel bonding technology for cable TV transmissions of SHV, we conducted demonstration experiments using commercial CATV channels, developed a compact receiver. We also investigated baseband transmission aimed at the large-capacity transmissions that can be expected in the future. 4 NHK STRL ANNUAL REPORT 2016

2 1.1 8K Super Hi-Vision format We made progress with our R&D and standardization activities related to the 8K Super Hi-Vision video system. High-dynamic-range television We worked on the standardization of the high-dynamicrange television (HDR-TV) video format. Our effort led to Recommendation ITU-R BT.2100 issued by the International Telecommunication Union Radiocommunication Sector (ITU-R) in July 2016 (1). This Recommendation specifies that the system gamma value in the electro-optical transfer functions (EOTFs) for the display side of the Hybrid Log Gamma (HLG) system should be set according to the peak luminance of the display. To verify the effect of this setting, we conducted video production experiments at different peak luminance with 20 video engineers to compare a case where the system gamma value was varied depending on the peak luminance and a case where it was not. The results showed that changing the system gamma value according to the display peak luminance tends to reduce the dependence of the average picture level (APL) and histogram of video on peak luminance, demonstrating that it is effective for consistent video production independent of viewing conditions (2). 4K/8K test broadcasting started on August 1, 2016 and some of the programs are being offered in the HLG HDR. We contributed to the Next Generation Television & Broadcasting Promotion Forum (NexTV-F) and the Association for Promotion of Advanced Broadcasting Services (A-PAB) for their discussion on HDR operational guidelines for the video level of HDR programs. With the aim of enabling the simultaneous production of standard dynamic range (SDR) and HDR content, we prototyped a camera capable of capturing SDR and HDR video simultaneously (3). This camera is equipped with a common optical iris used for SDR and HDR and an electronic iris for the gain adjustment of SDR. We demonstrated that controlling the iris for SDR while producing HDR video enables the simultaneous shooting of HDR and SDR video. We built an 8K HDR live production system that covers the whole process from capture to display by connecting multiple 8K HDR (HLG) cameras, an existing video switcher, a compression recorder, a text superimposer and an HDR liquid crystal display. Using this system, we conducted experiments to capture persons and other objects with large amounts of contrast and demonstrated live video production in the HLG system (Figure 1) (4). Figure 1. Exhibit of HDR live production (at the NHK STRL Open House 2016) Full-featured 8K program production system We are conducting R&D on program production equipment and systems that support a 120-Hz frame frequency with the goal of achieving full-featured 8K video production. We built a test production system for 8K/120-Hz (5) consisting of a compression recorder, a signal router (6), a waveform monitor (7), a 17-inch compact display, color grading equipment with a down-conversion function (8) and 120-Hz time code equipment (9). Using this system, we confirmed the normal operation of the switching of 8K/120-Hz video signals, transmissions between devices, signal monitoring and the compatibility of the 120-Hz time code with conventional 60-Hz equipment. We also developed a production switcher by increasing the number of inputs and outputs of the signal router and adding image processing and transition effect capabilities. In addition, we built an 8K/120-Hz off-line video editing system. We also verified the feasibility of a program production system that uses the Precision Time Protocol (PTP), which can coordinate clocks precisely within microseconds, as the synchronizing signal (10). Spatial resolution measurement methods for TV cameras We standardized two types of measurement methods for accurately assessing the spatial resolution characteristics of TV cameras. One of them is to capture an HDTV In Mega Cycle Chart, a resolution chart in which bar patterns having multiple spatial frequencies are arranged spatially, and read the amplitude response of the patterns on a waveform monitor. The other method is to capture a slanted-edge chart and calculate the modulation transfer function (MTF) from the edge. We compiled guidelines for these two measurement methods of resolution characteristics in ARIB Technical Report TR-B41 (11). Color rendering properties of LED lighting We conducted experiments to evaluate the color rendering properties of white LED lighting for wide-color-gamut 4K/8K production. On the basis of the results, we proposed to ARIB that the average color rendering index (Ra) of 90 or higher and the special color rendering index (R9) for red of 80 or higher be adopted for recommended values, leading to ARIB Technical Report TR-B40 (12). Standard test sequences We contributed to the production of Series B of ultra-highdefinition/wide-color-gamut standard test sequences provided by the Institute of Image Information and Television Engineers (ITE). We captured sport scenes and other materials that are not included in Series A with our 8K full-resolution camera and edited the images by adding scrolling subtitles. (1) Rec. ITU-R BT.2100: Image parameter values for high dynamic range television for use in production and international programme exchange (2016) (2) Y. Ikeda, Y. Kusakabe, K. Masaoka and Y. Nishida: Effect of Variable System Gamma for Hybrid Log-Gamma HDR Video Production, Proc. IDW/AD 16, pp (2016) (3) Y. Ikeda, K. Masaoka and Y. Nishida: A television camera to capture HDR and SDR images simultaneously, ITE Tech. Rep., Vol.40, No.15, IST , pp (2016) (in Japanese) (4) D. Koide, T. Yamashita, R. Funatsu, N. Shirai, Y. Ikeda, Y. Nishida and T. Ikeda: 8K UHDTV-HDR Live-Program Production -System Construction and Trial Production-, ITE Tech. Rep., Vol.40, No.23, BCT (2016) (in Japanese) (5) T. Hayashida, T. Soeno, J. Yonai, A. Iwasaki, Y. Ikeda, D. Koide, T. Yamashita, Y. Takiguchi, E. Miyashita and Y. Nishida: Development of an 8K Production System With 120 Hz Frame Frequency, SMPTE 2016 Annual Technical Conference and Exhibition (2016) (6) J. Yonai, T. Yamashita and Y. Nishida: Development of a seamless switcher for the full-featured 8K Super Hi-Vision, ITE Annual NHK STRL ANNUAL REPORT

3 Conference, 22D-2, 2016 (in Japanese) (7) T. Soeno, Y. Ikeda and T. Yamashita: Development of 8K Waveform Monitor with U-SDI Signal Analyzer, ITE Tech. Rep., Vol.40, No.14, BCT , 2016 (in Japanese) (8) T. Hayashida and T. Yamashita: Development of Color-grading equipment for the full-featured 8K Super Hi-Vision, ITE Annual Conference, 22D-1, 2016 (in Japanese) (9) T. Soeno, N. Shirai, D. Koide and T. Yamashita: A Time Code Signal Transmission for 4K/8K Program Production with a 120-Hz Frame Frequency, ITE Tech. Rep., Vol.40, No.14, BCT , 2016 (in Japanese) (10) A. Iwasaki, T. Hayashida and T. Yamashita: Study of the Synchronization System for 8K 120fps Production, ITE Annual Conference, 22D-3, 2016 (in Japanese) (11) ARIB Technical Report TR-B41 1.0, Measurement methods for resolution characteristics of television camera systems (2016) (12) ARIB Technical Report TR-B40 1.0, Color rendering indexes and recommended values of white LED illuminant for UHDTV programme production (2016) 1.2 Cameras We are researching and developing for practical full-featured Super Hi-Vision (SHV) cameras and for imaging technologies to be employed at the upcoming Tokyo Olympic Games. Full-featured SHV camera We developed the first full-featured SHV camera conforming to ITU-R Recommendation BT.2100, which was standardized in FY 2016 (1). This camera supports a high dynamic range (HDR) in addition to a 120-Hz frame frequency, 8K full resolution and a bit depth of 12 bits, which were achieved with our previous cameras. It achieved a setting of 1200% dynamic range by supporting the Hybrid Log Gamma (HLG) format of HDR. We captured various evaluation images with this camera and screened them at NAB Show 2016 and the NHK STRL Open House. With the aim of making a full-featured SHV camera more compact and practical, we developed an 8K full-resolution single-chip color camera using a 133-megapixel image sensor (Figure 1) (2). Its camera head weighs only 6.3 kg and the CCU was downsized to the 3U size. Regarding the resolution characteristics, the camera achieved an MTF in excess of 35% with 3,200 TV lines. We also conducted a 120-Hz driving experiment in which 60-Hz image sensors were driven by inter-line scanning in two-line units in the 133-megapixel image sensor. We previously developed a Cube Camera, a compact SHV camera that uses a 120-Hz, 33-megapixel single-chip image sensor. Using this camera, we captured fast-moving sport scenes such as bike races and ice hockey games for content screening at NAB Show We also researched elemental technologies for improving camera performance. For noise characteristics, we conducted subjective evaluation experiments to investigate the required SN ratio of 8K cameras for HDR images (3). The results showed that an SN ratio of 48 db or higher is desirable under the conditions of a viewing distance of 0.75 H, a display luminance of 1000 cd/m2, a system gamma of 1.2 and a frame frequency of 60 Hz. For the color reproduction characteristics, we investigated bit depth degradation, which is caused by the signal value clipping of high-chroma signals during color adjustment with a linear matrix. We devised a method for using signal processing to reduce the degradation and demonstrated its effectiveness through experiments. To suppress flicker in images captured at a 120-Hz frame frequency under an environment in which the brightness of lighting changes with a 50-Hz, we developed a technique by shooting at 240-Hz and reconstruct a flicker-free image. We demonstrated its effectiveness in capture experiments. For autofocus (AF) technology, we proposed a hybrid AF system that combines a sensor-based phase-difference detection AF (PDAF) with focus-aid (FA) signal, and demonstrated its effectiveness through simulations. We are aiming to develop a general-use 8K camcoder combining a camera and recorder for the wide-spread use of SHV cameras. In FY 2016, we developed prototype equipment for verifying the basic functions. The prototype can record 8K/60-Hz video for more than one hour into four SD cards using conventional compression recording technology. Full-featured SHV image sensor We made progress in our development of image sensors for a practical full-featured SHV camera. In FY 2016, we designed a 1.25-inch full-featured SHV image sensor. It has 33 megapixels, each of which has a size of 2.1 μm (photodetecting area of 16.2 mm 9.1 mm), and operates at 240 Hz. The quantization bit depth of its A/D converter is 14. We designed a new three-stage pipelined ADC architecture for the A/D converter to enable high-precision and high-speed operation. We also incorporated a noise reduction processing circuit to reduce random noise and fixed-pattern noise. This research was conducted in cooperation with Shizuoka University. High-speed SHV capturing technology With the aim of enabling the high-speed capture of SHV video at the Tokyo Olympic Games, we made progress in our development of capture equipment, a recorder and a slowmotion player. We began the development of experimental capture equipment that enables the high-speed capture of video with 33 megapixels at a 240-Hz frame frequency. We developed a recorder that can perform the real-time compression of 8K video in 4:2:0 at a 240-Hz frame frequency by the parallel operation of two compression signal processing boards for full-featured SHV. We are also developing a slow-motion system that can record and reproduce video simultaneously. In FY 2016, we verified its operation at 60 Hz. We also conducted experiments on capturing sports scenes such as speed skating in cooperation with Nippon Sport Science University. Electronic variable ND filter We are researching an electronic neutral-density (ND) filter for television cameras that can control incident light continuously and also locally. We previously studied the use of electrochromic (EC) materials with a wide variable range of the transmission rate from 80% to less than 1% for dimming materials and prototyped an EC element consisting of an electrolyte solution of silver nitrate (AgNO3) and copper chloride (CuCl2) inserted between ITO transparent conductive films. In FY 2016, we made the dimming element into a twolayered structure, which doubled the operation speed compared with the conventional one. This research was conducted in cooperation with Murakami Corporation. 6 NHK STRL ANNUAL REPORT 2016

4 (1) K. Kitamura, T. Yasue, T. Soeno, H. Shimamoto: Full-specification 8K Camera System, NAB Broadcast Engineering Conference Proceedings, pp (2016) (2) T. Nakamura, T. Yamasaki, R. Funatsu, H. Shimamoto: Development of a full-resolution 8K single-chip portable camera system, ITE Annual Report, 22E-1, (2016) (in Japanese) (3) R. Funatsu, K. Kitamura, T. Yasue, D. Koide, H. Shimamoto: Development and Image Quality Evaluation of 8K High Dynamic Range Cameras with Hybrid Log-Gamma, Electronic Imaging, Image Quality and System Performance XIV, pp (2017) Figure 1. 8K full-resolution single-chip camera head 1.3 Displays We have made progress in our development of various displays that can handle 8K Super Hi-Vision (SHV) video and continued with our research on large sheet-type displays. Full-featured 8K program production (display-related) For the purpose of evaluating full-featured 8K panels, we prototyped interface evaluation equipment for 8K displays that can drive panels in low-voltage differential signaling (LVDS). Evaluations using a 55-inch 8K panel demonstrated that the panel can operate at 120 Hz and that a display gamma with different characteristics set by users can be applied in addition to those specified in BT.709 (SDR) and BT.2100 (HLG). We also reduced the size of our 9.6-inch 8K monitor that we developed in FY 2015 by separating its display unit from the power/ control unit. SHV sheet-type display technologies We are researching large, lightweight and sheet-type organic light-emitting diode (OLED) displays that can be rolled up and used in the home for showing SHV. An organic light-emitting element is a self-emissive display device and is suitable for a thin display because it does not require a backlight. In FY 2016, we developed a 130-inch large display by arranging four 4K OLED panels that use thin glass substrates. The display, whose thickness is only 2 mm including the back board for fixing the panel, demonstrated the feasibility of an ultrathin display (Figure 1). This display was fabricated in cooperation with LG Display and ASTRODESIGN, Inc. We plan to improve its performance such as panel luminance and 120-Hz operation and to develop a display that can show 8K images with a single panel. Full-featured SHV projector We improved the image quality of our full-featured SHV projector by increasing the brightness and reducing the coherency of its laser light sources. More specifically, we increased the number of laser diodes to increase the brightness and also widened the half-power band width of light sources by adjusting the center wavelength of individual laser diodes to reduce the coherency. As a result, the projector doubled the brightness and halved the speckle contrast while covering 95% of colors in the xy chromaticity diagram of the BT.2020 color gamut. Figure 1. Prototype large sheet-type display 1.4 Recording systems We are researching compression recorders with the aim of developing full-featured SHV recording equipment. In FY 2016, we worked to downsize the solid-state memory package, implement support for a high-speed interface and extend the functionality of the compression recorder (1). To reduce the size of the memory package, we investigated the use of an NVMe interface, which supports a wider band than the conventional interface. The conventional memory package uses a bundle of multiple SSDs having a 6-Gbps SATA interface, which has hampered downsizing efforts. We therefore devised a memory package structure that uses a single SSD instead of a bundle of multiple SSDs by supporting the wide-band NVMe interface (Figure 1). This structure showed the feasibility of a compact memory package with a size reduced to about 1/7 of that of the conventional package. We also conducted experiments using a memory evaluation board and found that parallel writing into the memory chip and an increased writing block size can achieve a writing/reading speed in excess of 20 Gbps, which is equivalent to that of the conventional memory package. The memory package also has versatility as a recording medium because the NVMe interface is an open standard. On the compression recorder side, we implemented the NVMe interface on the substrate firmware that controls recording, and we verified recording and NHK STRL ANNUAL REPORT

5 reproduction operations with a commercial NVMe drive. To extend the functionality of the compression recorder, we developed a decoder IP core supporting 8K ProRes, which is a codec for video editing, and incorporated it into the recorder in order to simplify the input to/output from the editing system. We also implemented external input/output of the audio and time code, which is a necessary function for a program production system. To sustain the quality of the compressed pictures, we improved the color conversion method to make it effective for compression recording without chroma subsampling. We earlier confirmed that the use of a specific color conversion formula for suppressing noise propagation was effective for curbing image degradation when a certain color signal contains large noise. However, we later found that a conventional color conversion formula was more effective for some images. We therefore investigated a method of switching the formulas adaptively according to the identified visual feature value of the image and confirmed through simulations that the method always reduced image quality degradation after compression (2). (1) K. Ogura, K. Kikuchi, T. Kajiyama, E. Miyashita: Full specification NVMe interface Memory controller Memory chips Figure 1. Structure of compact memory package 8K Super Hi-Vision Compression Recorder and Backup system, ITE Technical Report Vol.41, No.5, pp (2017 in Japanese) (2) K. Kikuchi, T. Kajiyama, K. Ogura, E. Miyashita: Adaptive Color Space Transforms for 4:4:4 Video Coding Considering Uncorrelated Noise among Color Components, MMSP2016 Digest, 39 (2016) 1.5 Sound systems providing a strong sense of presence We are researching a 22.2 multichannel sound (22.2 ch sound) system for 8K Super Hi-Vision (SHV). SHV sound simultaneous production system We are studying technologies to produce high-quality 22.2 ch sound efficiently and simultaneously while producing stereo and 5.1 ch sound. We studied a shotgun microphone, which has sharper directivity than a conventional one, and devised a method for predicting the directivity characteristics of the microphones more accurately. Using this method, we designed a 1-m-long acoustic tube and developed a microphone using the acoustic tube (1). We also developed a microphone system consisting of compact shotgun microphones arranged in a spherical array structure (Figure 1) as a way of picking up 22.2 ch sound at a single point. The microphone achieved favorable separation between channels in a wide frequency range (2). We developed an adaptive downmixing technique focusing on the coherence between 22.2 ch audio signals. The technique suppressed variations in loudness levels, which occur with downmixing using constant coefficients, and thus achieved downmixing that can maintain the mixing balance in 22.2 ch. We also developed an audio monitor for U-SDI signals for fullfeatured 8K SHV production systems. For audio coding, we developed a software-based codec based on MPEG-H 3D Audio (3), which is the latest audio coding scheme, with the aim of supporting 22.2 ch sound in nextgeneration terrestrial broadcasting. The results of a subjective evaluation on the sound quality of coded 22.2 ch audio signals demonstrated that broadcast quality could be obtained at bit rates higher than 512 kbit/s (4). method to develop separate designs according to the bandwidth and verified its effectiveness through computer simulations. We also developed a signal processor capable of real-time execution of extended signal processing in cooperation with Sharp Corporation. We also studied ways of improving the performance of binaural reproduction. We confirmed that using loudspeakers located on the side of the listener to reproduce audio signals with a suppressed bandwidth produces a perception of sound coming from behind the listener and proposed this method as a rear enhancement filter (5). We also demonstrated that the sense of spaciousness increased when this method is combined with a simple reproduction technique in which loudspeakers are set only in front. We investigated a method for converting 22.2 ch sound signals into those of an arbitrary number of channels in order to enable 22.2 ch sound reproduction using a home theater system with fewer channels. The results of a listening experiment showed that our proposed fixed downmix coefficients did not undermine spatial impression as much as downmix coefficients minutely calculated to maintain the direction of sound arrival. Reproduction of converted SHV sound We are researching technologies for the easy reproduction of 22.2 ch sound at home. In FY 2015, we formulated a design method for a reproduction controller for transaural reproduction using a line-array loudspeaker system integrated with a flat panel display (Figure 2). In FY 2016, we extended the design Figure 1. Appearance of single-unit microphone 8 NHK STRL ANNUAL REPORT 2016

6 Figure 2. Line-array loudspeaker system integrated with a display We also studied how to make 22.2 ch sound easier to listen to at public viewings and other places. We investigated a way of sequentially correcting the reproduction level based on the level difference between audio signals and ambient noises in each frequency band while considering the inter-channel correlation in each bandwidth of audio signals and demonstrated that this method improved the audibility of 22.2 ch sound (6). Standardization At the International Telecommunication Union Radiocommunication Sector (ITU-R), we contributed to the establishment of a new Recommendation on the sound systems and the order of channels when programs produced by various sound systems, including those for 22.2 ch sound, are transmitted and recorded (7). We also investigated measurement errors caused by the direction of a measurement microphone which occurs in adjusting the sound characteristics of a sound evaluation room or other places and compiled the results into a draft report. In addition, we contributed to revisions of related Recommendations. This includes the addition of a descriptor for specifying downmix coefficients in the Recommendation on sound metadata used by 22.2 ch and other advanced sound systems, the definition of a constant term and the clarification of a calculation method in the Recommendation on loudness measurement methods, as well as the addition of test sound sources including 22.2 ch sound to the report on compliance material for verifying that a loudness meter meets the specifications within loudness measurement methods. At the Association of Radio Industries and Businesses (ARIB), we contributed to a revision of the standards on 22.2 ch sound to comply with the ITU-R Recommendation (8). We also helped to revise the loudness operational guidelines to incorporate 22.2 ch test sound sources. Additionally, we revised a 22.2 ch sound standard with regard to the method of adjusting the room response as well as 5.1 ch sound production guideline. We also produced standard sound sources of 3D multichannel sound for the purpose of promoting 22.2 ch sound. At MPEG, we worked on the incorporation of ARIB standards into an MPEG standard and thus contributed to the international promotion of the 22.2 ch sound broadcast standard. We also participated in a verification test of MPEG-H 3D audio and helped develop the latest coding scheme. (1) Y. Sasaki, T. Nishiguchi, K. Ono, T. Ishii, Y. Chiba, A. Morita: Development of shotgun microphone with extra-long acoustic tube, Audio Engineering Society Convention Paper 9639 (2016) (2) Y. Sasaki, T. Nishiguchi, K. Ono: Development of multichannel single-unit microphone using shotgun microphone array, Proceedings of the 22nd International Congress on Acoustics (ICA2016) (3) ISO/IEC : 2015 (2015) (4) T. Sugimoto, T. Komori: Audio coding of 22.2 ch audio signal using MPEG-H 3D Audio, Autumn meeting of the Acoustical Society of Japan, (2016). (5) T. Hasegawa, S. Oode, K. Ono, K. Iida: Control of a sound image to the rearward direction using band-stop filters and laterally located loudspeakers, Autumn meeting of the Acoustical Society of Japan, (2016). (6) S. Kitajima, T. Sugimoto, K. Ono: Reproduction method of 22.2 multichannel sound in noisy enviornment considering interchannel correlation, IEICE Technical Report, EA (2017) (7) Rec. ITU-R BS.2102: Allocation and ordering of audio channels to formats containing 12-, 16- and 32- tracks of audio (2017) (8) ARIB Standard STD-B59 2.0, Three-dimensional Multichannel Stereophonic Sound System for Programme Production (2016) 1.6 Video coding We are researching video compression techniques for fullfeatured 8K Super Hi-Vision (SHV) and SHV terrestrial broadcasting. Full-featured SHV video coding We investigated the required bit rates for 120-Hz video coding for full-featured SHV broadcasting and verified the image quality. The domestic standard for SHV broadcasting (ARIB STD-B32) employs temporal scalable coding that can partially decode 60-Hz video frames from compressed 120-Hz video streams. We conducted experiments in which we varied the rate of increase in bit amount of an entire 8K/120-Hz video relative to the layer of the 8K/60-Hz video. The results of an evaluation using an objective metric showed that the increasing rate of the bit amount was 10% or less. We verified the subjective quality of 120-Hz and 60-Hz videos and confirmed no significant deterioration from original videos (1). We researched speeding up methods of 8K/120-Hz video coding. We set appropriate block sizes and applied appropriate coding parameters to prediction methods for the differential temporal layer between 60-Hz and 120-Hz videos. This reduced the coding computation time by 17% while maintaining coding efficiency (2). We worked out a basic design of an 8K/120-Hz codec that uses High Efficiency Video Coding (HEVC). The encoder divides an 8K/120-Hz image into eight partitions (four in the spatial direction and two in the temporal direction) and performs encoding by using eight 4K/60-Hz equivalent HEVC encoders simultaneously. To verify the output streams and the effect on image quality caused by the partitioning, we developed simulation software of the encoder. For the decoder, which we decided to implement in software, we developed the core part for decoding processing and confirmed the successful decoding of streams created by the encoder simulation software. We produced evaluation images for R&D on 120-Hz video coding in cooperation with NTT (Nippon Telegraph and Telephone Corporation). We made images that are helpful for evaluating the quality when they are moving at high speed. Video coding for 8K program contributions We evaluated the image quality of coded video in HEVC in order to design system parameters for wireless contribution NHK STRL ANNUAL REPORT

7 links, which we call field pick-up units (FPUs). We tested one to three codecs in tandem and identified through subjective quality evaluation the required bit rates in each stage to meet the quality criteria for program contributions. The results showed that the bit rates of 285 Mbps for 8K video and 145 Mbps for 4K video were required to meet the criteria of ITU-R Recommendation BT.1868 in the three-stage cascade structure, which was the toughest condition. This evaluation was conducted as part of activities of the ARIB JTG (Joint Task Group) on HEVC performance evaluation for program contributions. Effort toward 8K terrestrial broadcasting During the Rio Olympic Games (August 5-21, 2016), we conducted experimental terrestrial transmissions of 8K and held public viewings (PVs) in Rio de Janeiro in cooperation with the Brazilian TV broadcaster TV Globo. In the experiment, 8K programs were compressed to 84 Mbps by using our prototype 8K HEVC encoder (Figure 1) and then transmitted over terrestrial waves. At the reception point, the compressed stream was decoded by our prototype 8K HEVC decoder and then displayed on a 98-inch 8K liquid crystal display (LCD) with 22.2 ch loudspeakers. This experiment presented the immersive sensation of 8K and our compression and transmission technologies to many people both in and outside Brazil. We also conducted similar experimental terrestrial transmissions in Japan and screened the received programs at the Rio Olympic PV venue in STRL. To enable 8K terrestrial transmission at a low bit rate, we developed a video processing to use before HEVC coding. This processing was partially implemented into a device in consideration of interface with 8K video player and HEVC codec. The processing increases the coding efficiency by taking advantage of the noise reduction and band limitation effects of wavelet shrinkage in the spatial domain. This research was conducted as a government-commissioned project from the Ministry of Internal Affairs and Communications titled R&D on Advanced Technologies for Terrestrial Television Broadcasting. Next-generation video coding In FY 2016, we developed technical elements for nextgeneration video coding to enable 8K terrestrial broadcasting. To improve intra prediction technology, we devised an algorithm that improves the accuracy of the intra prediction for chroma signals. We observed an improvement of the coding efficiency by referring to the prediction modes of neighboring chroma blocks without increasing computational complexity (3). We also developed a method for reducing prediction residues in motion compensated prediction. In addition, we devised a method for improving the motion vector coding used in motion compensation. An improvement of the coding efficiency was observed when controlling the derivation of predictors of the motion vector according to the shape of the block partitioning (4). We also devised a video-format adaptive technique as a new element of coding systems. In response to the international standardization of HDR video format, we developed a method for reducing the coding degradation of HDR video that occurs in encoding process, and confirmed its effectiveness through experiments. We proposed some of these methods to an international standardization conference on next-generation video coding technologies. We also provided HDR test sequences of the Hybrid Log-Gamma (HLG) format to promote standardization activities (5). Video coding using super-resolution techniques We completed a real-time video coding system that delivers 8K and 4K video signals simultaneously by using four pairs of 2K/4K super-resolution inter-layer prediction units with hybrid optimization criteria that can approximate perceptual image quality metric (6). This system transmits 4K HEVC video stream accompanied by ancillary data for super-resolving decoded 4K video into 8K resolution. We developed lossless coding of ancillary data, inter-unit synchronization and automatic synchronization of video and ancillary-data streams to demonstrate the connectivity of the total transmission system (Figure 2). This research was performed under the auspices of a program titled Research and Development of Technology Encouraging Effective Utilization of Frequency for Ultra High Definition Satellite and Terrestrial Broadcasting System of the Ministry of Internal Affairs and Communications, Japan. We are also researching video format conversion using super-resolution techniques. We improved spatial processing by introducing a method for selecting the frequency band to be used by considering the color sampling pattern of a camera (7). We also improved temporal processing by using linear-filtering interpolation with contrast compensation in consideration of the spatio-temporal contrast sensitivity of human vision system (8). For gradational processing, we developed a method for interpolating intermediate gradation by introducing a point spread function that imitates the degradation process of an optical system. (1) Y. Sugito, K. Kanda, S. Sakaida: A Study of Required Bit-rate on 8K 120Hz Video Coding, FIT2016, No.3, RI-004, pp (2016) (in Japanese) (2) Y. Sugito, K. Kanda, S. Sakaida: A Study of Parameters for 8K 120Hz Video Coding, ITE Annual Convention 2016, 12B-1 (2016) (in Japanese) (3) S. Nemoto, Y. Matsuo, A. Ichigaya: Chroma Intra Mode Predictor Based on Modes of Neighboring Blocks, International Workshop on Advanced Image Technology 2017 (IWAIT2017) (2017) (4) S. Iwamura, K. Iguchi, A. Ichigaya: Partition-adaptive merge candidate derivation, JVET-D0107 (2016) (5) S. Iwamura, A. Ichigaya: New 4K HDR test sequences with Hybrid Log-Gamma transfer characteristics, JVET-E0086 (2017) (6) T. Misu, S. Iwamura, Y. Matsuo, K. Kanda, S. Sakaida: Real-time Figure 1. Equipment for 8K program compression Figure 2. Coding system using super-resolution reconstruction technologies 10 NHK STRL ANNUAL REPORT 2016

8 8K/4K Video Coding System with Super-resolution Inter-layer Prediction, 32nd Picture Coding Symposium (PCS 2016), I-12 (2016) (7) Y. Matsuo, S. Sakaida: Super-Resolution Method by Registration of Multi-Scale Components Considering Color Sampling Pattern and Frequency Spectrum Power of UHDTV Camera, 18th IEEE International Symposium on Multimedia (ISM 2016), pp (2016) (8) Y. Matsuo, S. Sakaida: Frame-Rate Conversion by Linear-Filtering Interpolation Using Spatio-Temporal Contrast Compensation, 35th IEEE International Conference on Consumer Electronics (ICCE 2017), pp (2017) 1.7 Media transport technologies In our research on media transport technologies for transmitting video and audio, we are investigating MPEG Media Transport (MMT), which can be used for both broadcasting and broadband networks. We are also studying the IP delivery of 8K content, synchronized presentation technology and the use of MMT for next-generation terrestrial broadcasting. Application of MMT technologies We developed a receiver for broadcasting and broadband that can receive both 8K programs from satellite broadcasting and 8K content through IP delivery. We employed MMT as the signal format for IP content delivery as with 8K broadcasting to share the same signal processing circuit in the receiver. We conducted an experiment in which 11 channels of 8K content were delivered over a 10-Gbps-class network and any one of the channels was successfully presented at the receiver. This result demonstrated stable content delivery and reception. We also investigated a transmission protocol for the IP delivery of MMT signals and confirmed that it can handle unicast distribution as well as multicast distribution (1) and that it allows protocol conversion from MMT to HTTP (2). We exhibited these results at the NHK STRL Open House 2016, InterBEE 2016 and other exhibitions. Figure 1. Experiment on multichannel IP delivery of 8K content Next-generation terrestrial broadcasting Aiming for next-generation terrestrial broadcasting, we researched a multiplexing scheme for IP packets that conforms to the channel coding system for terrestrial broadcasting and an IP transmission system used over studio to transmitter links (STLs) and transmitter to transmitter links (TTLs) to enable a single-frequency network (SFN) in terrestrial broadcasting. We compiled our findings into specifications and conducted verifications with a prototype device. Part of this research was conducted as a governmentcommissioned project from the Ministry of Internal Affairs and Communications titled R&D on Advanced Technologies for Terrestrial Television Broadcasting. International standardization related to MMTbased broadcasting We offered to provide MPEG with MTT packet data and an analysis tool as a sample for implementing ISO/IEC TR MMT implementation guidelines and our offer was accepted (5). We also reported on our research on the application of MMT technologies for the IP delivery of 8K content at ITU-R SG6 and ITU-T SG16. (1) S. Aoki et al.: Delivery of 8K Content over Satellite Broadcasting and Broadband, International Workshop on Smart Info-Media Systems in Asia 2016, SS5-2, pp (2016) (2) S. Aoki et al.: A Study on Conversion from MMT Protocol to HTTP, Proceedings of IEICE Communications Society Conference, 2, B-6-10, p.10 (2016) (in Japanese) (3) Y. Kawamura et al.: Implementation of Inter-Terminal Video Synchronization Using MMT on Android, Entertainment Computing 2016, pp (2016) (in Japanese) (4) K. Otsuki et al.: A Study of Stored Formats to enable HTTP Access by designating Absolute Time, ITE Technical Report, Vol.40, No.45, BCT , pp (2016) (in Japanese) (5) ISO/IEC TR :2016: Information technology High efficiency coding and media delivery in heterogeneous environments Part 13: MMT implementation guidelines, 2016 This research was performed under the auspices of a program titled Research and Development of Technology Encouraging Effective Utilization of Frequency for Ultra High Definition Satellite and Terrestrial Broadcasting System of the Ministry of Internal Affairs and Communications, Japan. As an example of the application of MMT technologies, we developed a technique that allows multiple content such as multi-angle images to be delivered over IP and displayed on multiple devices such as TV and tablets in synchronization. We demonstrated that this technique achieves high-accuracy synchronization (3). We also studied a use case in which multiple pieces of content are retrieved from a network based on a designated absolute time and presented in synchronization with each other, and we demonstrated stored formats that can support this use case (4). NHK STRL ANNUAL REPORT

9 1.8 Advanced conditional access system technology We are researching an advanced conditional access system (CAS) that provides rights protection and conditional access to Super Hi-Vision content and working toward its practical application. The advanced CAS technology uses a secure scrambling scheme and supports a CAS module software update function to ensure continuous provision and improved security. We participated in standardization work at the Association of Radio Industries and Businesses (ARIB), which led to the revision of Chapter 5, Operational Guidelines for Advanced Digital Broadcasting Conditional Access System and Receiver Specifications of Technical Report ARIB TR-B39, Operational Guidelines for Advanced Digital Satellite Broadcasting, in December The revised version includes the specifications of a combined receiver for advanced BS digital broadcasting, BS/wide-band CS digital broadcasting and terrestrial digital television broadcasting. 1.9 Satellite broadcasting technology For the widespread use of 8K Super Hi-Vision (SHV), we are working on standardization and improving the performance of 12-GHz-band satellite broadcasting and also researching nextgeneration satellite broadcasting systems such as for 12-GHzband satellite broadcasting with a larger capacity and 21-GHzband satellite broadcasting. Transmission system for advanced wide-band digital satellite broadcasting We worked on the international standardization of a transmission system for satellite broadcasting and researched multilevel coded modulation and cross-polarization interference cancellation to improve transmission performance. We worked on the international standardization of the transmission system for 8K SHV satellite broadcasting, ISDB-S3 (Integrated Services Digital Broadcasting for Satellite, 3rd generation). Our effort led to the establishment of Recommendation BO.2098 at the International Telecommunication Union Radiocommunication Sector (ITU-R) in December We also contributed to the publication of ITU-R Report BO.2397, which describes ultrahigh-definition television satellite transmission experiments and the SHV test satellite broadcasting, which started in August Regarding multilevel coded modulation, we designed 64APSK (amplitude phase shift keying) coded modulation using set partitioning in order to further increase the capacity of 12-GHz satellite broadcasting (1)(2). Computer simulations demonstrated that the required C/N was improved by 0.4 db and 1.0 db compared with those of DVB-S2X 64APSK and Gray Bit error rate 64APSK coded modulation (4/5) DVB-S2X 64APSK (4/5) Gray coded 64QAM (4/5) C/N[dB] Figure 1. Transmission performance of 64APSK coded modulation coded 64QAM (quadrature amplitude modulation), respectively, when a code rate of 4/5 was used (Figure 1). In FY 2015, we proposed an algorithm that cancels interference of cross polarization in 12-GHz-band satellite broadcasting by receiving RHCP (right-hand circular polarization) and LHCP (left-hand circular polarization) signals, simultaneously. In FY 2016, we implemented the algorithm into hardware and verified its effectiveness. Advanced satellite broadcasting systems For the next generation of satellite broadcasting, we designed a 12-GHz-band on-board antenna and a 21-GHz-band arrayfed shaped-reflector antenna. We also developed a 12/21-GHzband dual-polarized receiving antenna and a 12-GHz-band onboard solid-state amplifier. With the aim of increasing the capacity and service availability rate of 12-GHz satellite broadcasting, we investigated a 12-GHz-band antenna that can be mounted on a broadcasting satellite to reduce interference to other countries and increase the output of a satellite transponder. We designed shaped-reflector antennas, one with a single reflector and the other with two reflectors, and compared their radiation patterns. The results showed that the dual reflector antenna is better suited to the requirements of mountability on a launch vehicle and consistency of each radiation pattern in right- and left-hand circular polarization. We applied shaped reflector to the 21-GHz-band on-board array-fed imaging reflector antenna. Using the aperture diameter and the number of elements as parameters, we studied ways of reducing side lobes, equalizing the RF power distribution of feed array elements and decreasing the number of elements. We designed an array-fed shaped-reflector antenna with an aperture diameter of 1.8 m and 19 elements and investigated a 21-GHz-band 300-MHz-class wide-band transponder system. The investigation results showed that the system achieves an annual service availability rate of 99.9% in Tokyo when using a satellite transmission output of 2.3 kw, a quadrature phase shift keying (QPSK) modulation scheme and a code rate of 1/2 (3). We designed a 12/21-GHz-band dual-polarized receiving antenna using a four-element microstrip array antenna for a feed antenna so that 12-GHz-band right- and left-hand circular polarization and 21-GHz satellite broadcasting can be received with a single parabolic antenna. We used a stacked substrate because the 12-GHz band and 21-GHz band have different optimum antenna substrate thicknesses. Simulation results showed that the antenna has a voltage standing wave ratio of 12 NHK STRL ANNUAL REPORT 2016

10 1.2 or less for the 12-GHz band and 1.4 or less for 21-GHz band and can cover both broadcasting bandwidths. To verify the feasibility of a 12/21-GHz-band dual-polarized receiving antenna, we evaluated radiation patterns of an offset parabolic antenna with 45 cm aperture diameter by using design values of the feed antenna. The antenna achieved gains of 33.5 dbi for 12-GHz-band right-hand circular polarization, 33.4 dbi for 12-GHz-band left-hand circular polarization and 37.7 dbi for the 21-GHz band (right-hand), and a cross-polarization discrimination in excess of 31 db. In our work on the development of a 12-GHz-band on-board solid-state amplifier, we researched technologies to reduce nonlinear distortion caused by the satellite transponder with the aim of improving the satellite transmission performance of 16APSK. To improve the nonlinear characteristics of on-board amplifiers, we prototyped a high-power on-board solid-state power amplifier that has higher linearity than a traveling-wave tube amplifier. By using gallium nitride, the amplifier achieved an output power of 120 W in the 12-GHz band. We connected the output of the prototype solid-state power amplifier, which is connected to a linearizer to compensate for the nonlinearity, to a receiver equipped with a reception equalizer, and evaluated the transmission performance of 16APSK. The results demonstrated that the required C/N was 1.1 db better than that of a traveling-wave tube amplifier. This research was funded by the Ministry of Internal Affairs and Communications, Japan, through its program titled Research and Development of Technology Encouraging Effective Utilization of Frequency for Ultra High Definition Satellite and Terrestrial Broadcasting System. To expand the transmission capacity by using 32APSK and improve the service availability rate of 16APSK, we investigated a thermal transportation method for a satellite carrying a 300-W-class traveling-wave tube amplifier or a 100-W-class solid-state power amplifier. This research was conducted in cooperation with the Japan Aerospace Exploration Agency (JAXA). (1) Y. Suzuki, Y. Koizumi, M. Kojima, K. Saito, S. Tanaka: A study on 64APSK Coded Modulation part 2 -Performance enhancement by LDPC coding rate optimization-, Proceedings of the 2016 IEICE Society Conference, B-5-22 (2016) (in Japanese) (2) Y. Koizumi, Y. Suzuki, M. Kojima, K. Saito, S. Tanaka: A study on 64APSK Coded Modulation, IEICE Tech. Rep., Vol.116, No.243, SAT , pp.51-56, Oct (3) S. Nakazawa, M. Nagasaka, S. Tanaka: A Design of a Compact Transponder for a 21-GHz Band Broadcasting Satellite using an Array-fed Shaped Reflector Antenna, Proceedings of the 2017 IEICE General Conference, B-1-1 (2017) (in Japanese) 1.10 Terrestrial broadcasting transmission technology For the terrestrial broadcasting of Super Hi-Vision (SHV), we are researching a next-generation terrestrial broadcasting system, a channel plan and transmission network design. Next-generation terrestrial broadcasting system We are in the process of establishing a preliminary standard for the purpose of migrating the current terrestrial broadcasting services to the next-generation system. In FY 2016, we developed detailed specifications for a hierarchical transmission system that multiplexes the services for fixed reception and those for mobile reception into a single OFDM modulation signal and prototyped a modulator and demodulator. We improved the frequency interleave of the mobile reception required C/N [db] 16QAM(1/2) QPSK(1/2) Maximum Doppler frequency [Hz] ISDB-T Preliminary standard ( ): FEC code rate Figure 1. Comparison of speed tolerance between the preliminary standard and ISDB-T layer, added a power boost function and verified the effectiveness. We also optimized the pilot signal configuration for mobile reception and confirmed a better required C/N and speed tolerance than the current ISDB-T (Integrated Services Digital Broadcasting - Terrestrial) (Figure 1) (1). Future TV receivers are expected to receive broadcasting signals not only from conventional terrestrial and satellite broadcasting but through various channels such as the Internet and Wi-Fi networks. To increase compatibility with the Internet, we developed an MMT (MPEG Media Transport) that supports hierarchical transmission using multiple layers and implemented it into a prototype system. We also investigated a mechanism for retransmitting received broadcasting signals to a Wi-Fi network in a room or vehicle so that broadcast content can be provided to terminals in places where terrestrial broadcasting cannot be received directly. Advanced technologies for terrestrial broadcasting We are researching terrestrial SHV broadcasting under the auspices of the Ministry of Internal Affairs and Communications, Japan as part of its program titled Research and Development for Advanced Digital Terrestrial TV Broadcasting System, which started in FY Using the constellation, forward error correction and MIMO system compliant with the preliminary standard, we began developing a modulator and demodulator that are capable of a larger transmission capacity and hierarchical transmission. As part of this research, we plan to set up an experimental transmission station in Tokyo. In FY 2016, we developed transmission and equipment specifications for this large-scale experimental station. We also investigated the permissible value of interference caused by radio waves from the experimental station because the test broadcasting signal must not affect current terrestrial TV broadcasting receivers used in the same frequency band. This research was conducted in cooperation with Sony NHK STRL ANNUAL REPORT

11 Degradation of required receiver input power in the preliminary standard [db] (256QAM, LDPC code, 2/3 code rate) Figure 2. Comparison of the degradation of transmission characteristics in a multipath environment Corporation, Panasonic Corporation and Tokyo University of Science. Transmission performance evaluation using terrestrial broadcasting propagation characteristics We evaluated the system performance in a multipath environment with the signals (256QAM) for the fix reception by using a preliminary-standard modulator and demodulator that we prototyped in FY We prepared a multipath environment for laboratory experiments by simulating the terrestrial broadcasting propagation channels measured around our laboratory and compared the degradation of the required receiver input power between the preliminary standard and the existing ISDB-T format (2). The results demonstrated the superior performance of the preliminary standard, which exhibited less degradation than ISDB-T (Figure 2). Channel planning ISDB-T is better Preliminary standard is better Degradation of required receiver input power in ISDB-T [db] (64QAM, convolutional code, 3/4 code rate) In FY 2016, we began a study in cooperation with the Engineering Administration Department to estimate the required scale of frequency reallocation (repacking) of existing digital terrestrial broadcasting, which will be necessary when Required received power difference between conventional SFN and advanced SFN [db] (Required received power of conventional SFN/ Required received power of advanced SFN) Field measurement result (STC-SFN) Field measurement result (SFC-SFN) Computer simulation Received power ratio of main waves and SFN waves in receiver input [db] All 25 data Improvement made by STC-SFN 1.8 db (average) Improvement made by SFC-SFN 1.8 db (average) Figure 3. Field measurement results of STC-SFN and SFC-SFN Received Power ratio of main wave and SFN wave is between 5-12 db 2.0 db (average) 2.0 db (average) allocating new channels for SHV nationwide. R&D on effective use of frequency We have been researching single frequency network (SFN) technology to use frequencies more efficiently through a government-commissioned project from the Ministry of Internal Affairs and Communications titled Research and Development of Technology Encouraging Effective Utilization of Frequency for Ultra High Definition Satellite and Terrestrial Broadcasting System, which started in FY In FY 2016, we conducted transmission experiments within the service area of the experimental stations in Hitoyoshi City, Kumamoto Prefecture, to compare the performance of a space-time coding (STC)-SFN, a space-frequency coding (SFC)-SFN and a conventional SFN. The results showed that the STC-SFN and SFC-SFN had about 2 db better required received power than the conventional SFN. We also confirmed that the SFC-SFN had an equal level of reception characteristics to those of the STC- SFN in the propagation environment of Hitoyoshi City (Figure 3) (3). We also conducted a comprehensive connection experiment by combining STC-SFN technology with an 8K HEVC decoder. The results demonstrated that 8K video could be received successfully in an SFN environment about 18 km and 20 km away from the Hitoyoshi station and the Mizukami station, respectively. Collaboration with overseas organizations During the Rio Olympic Games, we demonstrated 8K terrestrial live transmissions using the world s first HEVC realtime codec in Rio de Janeiro and Tokyo simultaneously. In Brazil, the 8K content was transmitted from an experimental station, where our modulator was connected to the transmission equipment of the Brazilian TV broadcaster TV Globo. The transmitted signals were received at a demonstration venue approximately 8 km away for live public viewing of the Games. We also conducted field experiments in Rio de Janeiro in parallel with the demonstration. The results were reported at SET EXPO and added to ITU-R BT.2343, a report describing the results of terrestrial field experiments on ultra-highdefinition television around the world. At the Digital Broadcasting Experts Group (DiBEG), which promotes ISDB-T internationally, we updated a joint document of Associação Brasileira de Normas Técnicas (ABNT) Brazilian technical standards in cooperation with broadcasters in Brazil and other South American countries. We also investigated and reported trends toward next-generation terrestrial broadcasting of other countries at the Japan-Brazil next-generation broadcast study task force. We dispatched one of our researchers to the Polytechnic University of Valencia in Spain to research the use of MIMO for broadcasting and investigate the trend toward next-generation terrestrial broadcasting in Europe. (1) H. Miyasaka, A. Sato, S. Asakura, T. Shitomi, S. Saito, Y. Narikiyo, T. Takeuchi, M. Nakamura, K. Murayama, M. Okano, K. Tsuchida, K. Shibuya: A Study on the Scattered Pilot Patterns for Mobile Reception in the Next Generation Terrestrial Broadcasting, ITE Technical Report, Vol.40, No.30, BCT , pp.5-8 (2016) (in Japanese) (2) A. Sato, T. Shitomi, T. Takeuchi, M. Okano, K. Tsuchida: Transmission Performance over Multipath Environment of Proposed Specification for the Next Generation Terrestrial Broadcasting - Comparison with ISDB-T by Indoor Experiment -, ITE Technical Report, Vol.40, No.45, BCT , pp.5-9 (2016) (in Japanese) (3) S. Saito, T. Shitomi, S Asakura, A. Satou, M. Okano, K. Murayama, K. Tsuchida: 4x2 MIMO Field Test of Advanced SFN Using Space Time Coding For 8K Transmission, IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB 2016), IEEE, 10A-2 (2016) 14 NHK STRL ANNUAL REPORT 2016

12 1.11 Wireless transmission technology for program contributions (FPU) With the goal of realizing Super Hi-Vision (SHV) live broadcasting of emergency reports and sports coverage, we are conducting R&D on field pick-up units (FPUs) for transmitting video and sound materials. In FY 2016, we researched a microwave-band FPU, a millimeter-wave-band FPU and a 1.2-GHz/2.3-GHz-band FPU. Microwave-band FPU We researched a microwave-band FPU that can transmit 8K SHV signals at a 200-Mbps-class transmission rate over a distance of 50 km and worked toward its standardization. This FPU achieves a much larger transmission capacity in the 18- MHz transmission bandwidth, which is also used by current Hi-Vision FPUs, by using dual-polarized multiple-input multiple-output (MIMO) technology and orthogonal frequency division multiplexing (OFDM) with a higher-order modulation scheme (Figure 1). In FY 2016, we upgraded the low-density parity-check (LDPC) codes and the OFDM frame structure of our prototype modulator and demodulator to develop a practical FPU. For the LDPC codes, we employed the 44,880-bit-length codes used in the transmission system for advanced wide-band digital satellite broadcasting (ARIB STD-B44). The new frame structure allows one OFDM frame to contain an integer number of LDPC code blocks. With an optimized number of data carriers and frame length, the structure enabled efficient transmissions. We also contributed to the establishment of a national standard and the standardization at the Association of Radio Industries and Businesses (ARIB) so that microwave-band FPUs can be put into practical use in Millimeter-wave-band FPU We researched a millimeter-wave-band (42-GHz-band) FPU that can transmit 8K SHV signals at a 400-Mbps-class transmission rate and promoted standardization activities. In FY 2016, we fabricated a MIMO-OFDM modulator and demodulator that support a 125-MHz transmission bandwidth, twice the conventional bandwidth (62.5 MHz) (1). By combining it with an RF front-end unit and dual-polarized antenna that we fabricated in FY 2015, we completed a prototype millimeterwave-band FPU (Figure 2). We confirmed through laboratory experiments that the FPU can achieve a maximum transmission rate of 600 Mbps (with a 32QAM subcarrier modulation scheme and 3/4 coding rate). We also contributed to a revision of the ARIB standard on a portable millimeter-wave digital transmission system for television program contribution (ARIB STD-B43) in preparation for its practical use. In addition, we conducted field transmission experiments over a distance of 8 km between the NHK Broadcasting Center in Shibuya and our laboratory by connecting a millimeter-wave-band FPU with an H.265/HEVC codec that the Engineering Administration Department developed in FY 2016 for transmitting 8K program contributions. The results proved the feasibility of 8K SHV signal transmissions. 1.2-GHz/2.3-GHz-band FPU To enable the mobile relay broadcasting of SHV signals by using the 1.2-GHz/2.3-GHz-band, we are researching a MIMO system with adaptive transmission control using the time division duplex (TDD) scheme. In FY 2016, we confirmed that it is possible to sufficiently reduce the influence of feedback delays caused by TDD even in a time-varying channel similar to that used to broadcast a relay road race (2). We applied space-time trellis coding to the system we prototyped in FY 2015 (Figure 3) in order to increase the robustness of down links. We also investigated rate matching, which controls the code rate of error correction coding adaptively according to the varying channel quality. We used concatenated codes of a turbo code and a Reed Solomon (RS) code for forward error correction. Bit puncturing after turbo coding controlled the coding rate between 0.33 and 0.92 depending on the channel quality. Computer simulations using the received data in a radio wave propagation experiment carried out on the Kyoto Ekiden (a marathon relay race) course showed that adaptively controlling the coding rate can prevent a transmission error even when the reception quality deteriorates (3). Additionally, we prototyped 2.3-GHz-band power amplifiers for field transmission experiments of this system and obtained an experimental radio station license. Part of this research was conducted as a governmentcommissioned project from the Ministry of Internal Affairs and Figure 1. Microwave-band FPU prototype Figure 2. Millimeter-wave-band FPU prototype Figure 3. Prototype MIMO system with adaptive transmission control NHK STRL ANNUAL REPORT

13 Communications titled R&D on Highly Efficient Frequency Usage for the Next-Generation Program Contribution Transmission. (1) J. Tsumochi, F. Ito, Y. Matsusaki, H. Kamoda, T. Nakagawa, H. Hamazumi: Development of Wideband Modulator and Demodulator using a MIMO-OFDM Technique for 42 GHz band FPU (in Japanese), ITE Technical Report, Vol.40, No.23, BCT (Jul. 2016) (2) F. Ito, F. Uzawa, K. Mitsuyama, T. Kumagai, N. Iai: A Study on TDD Feedback Scheme for Adaptive Control SVD-MIMO System, 2017 IEICE General Conference (2017) (in Japanese) (3) F. Uzawa, F. Ito, K. Mitsuyama, T. Kumagai, N. Iai: A Study on Ratematching Applicable to Field Pick-up Units for Mobile Relay, 2017 IEICE General Conference (2017) (in Japanese) 1.12 Wired transmission technology We are researching an Internet Protocol (IP) based program production and program contribution system that can be used for 8K programs. We are also studying a channel bonding technology for transmitting 8K programs over cable TV networks and the FTTH digital baseband transmission system. IP based program production and program contribution system We are conducting R&D on a system that applies IP technology to program production and program transmission for a low-cost and efficient workflow. The system is expected to have various advantages by using widespread IP technology, such as allowing the use of high-speed inexpensive IP devices (1) and transmitting signals of various formats for video, sound, synchronization and control over a single IP network. In FY 2016, we evaluated the synchronization performance of the Precision Time Protocol (PTP), a synchronization technique standardized by IEEE1588, as a method for synchronizing video devices connected to an IP network. PTP synchronizes the clock between devices by exchanging time information (PTP packets) between a master device and slave devices and adjusting the time of the slave devices to the highprecision clock of the master device. The clock synchronization performance is affected by the transmission latency of PTP packets, and it varies by the number of Ethernet switches connected between the master and slave devices. We therefore conducted experiments and quantified the relationship between the number of connected switches or the transmission rate of data passing through the switches and the clock synchronization performance. We plan to utilize the findings for future system design. We also implemented the function of PTP clock synchronization into IP transmission equipment for uncompressed 8K signals and confirmed that the equipment is capable of the stable transmission of uncompressed 8K signals. The use of IP technology for program production enables various capabilities such as automatically detecting a device connecting to the IP network and controlling the connection. We participated in activities at the Advanced Media Workflow Association Networked Media Incubator (AMWA NMI), an international organization that discusses and standardizes connection management methods of IP program production systems, and built a test environment in our laboratory for verifying the connection management methods being discussed at the AMWA NMI. To allow the use of connected devices from any of the studios connected over the IP network, we investigated a new mechanism for preventing the contention of device connection management. We implemented this mechanism into our test environment and verified its operation by referring to the control API format adopted by AMWA NMI. Cable TV transmission of 8K signals We continued with our R&D on a channel bonding technology to transmit partitioned 8K signals over multiple channels so that 8K programs can be distributed through existing coaxial cable television networks. In FY 2016, we conducted retransmission experiments on 8K satellite broadcasting by using commercial cable networks and demonstrated the feasibility of 8K program retransmission via existing cable TV networks (2). Viewing 8K satellite broadcasting through cable TV also requires a compact and affordable receiver. To address this need, we developed the world s first compact receiver (tuner) equipped with a demodulator LSI that supports channel bonding technology (Figure 1). Using the same output signals as the demodulated signals of 8K satellite broadcasting, video and audio decoder LSIs for satellite broadcasting receivers can also be used for cable TV receivers. Digital baseband transmission system for FTTH As a way of distributing broadcasts to homes using FTTH (Fiber to the home), we are studying a digital baseband transmission system. In this system, multiple streams of 8K and Hi-Vision broadcasting are multiplexed with 10-Gbpsclass baseband signals by using time division multiplexing (TDM) and transmitted over optical fibers. It can reduce the cost of large-capacity transmission significantly compared with conventional RF transmission. In FY 2016, we prototyped transmission equipment capable of mitigating the difference in transmission latency, which occurs when many streams are multiplexed. We conducted a performance evaluation and demonstration. We also investigated a way of migrating the existing FTTH equipment for RF signal transmission to a digital baseband transmission system in stages. (1) J. Kawamoto, T. Kurakake: Uncompressed 8K Ultra-high Definition Television Transmission over 100G Ethernet in Broadcasting Station, OFC2017, M2I. 5 (2017) (2) K. Uezono, K. Nakajima, H. Matsumoto, Y. Hakamada, T. Kusunoki, T. Kurakake: Retransmission trial of 8K UHDTV satellite broadcasting over an existing commercial CATV line, Proc. of the IEICE Society Conference, B-8-35 (2016) (in Japanese) Figure 1. Compact receiver 16 NHK STRL ANNUAL REPORT 2016