NHK Laboratories Note No. 468


Kenichi Tsuchida, Naohiko Iai, Shigeki Moriyama, and Makoto Sasaki
(Digital Broadcasting Network)
    ISDB-TN (Narrow-band Terrestrial Integrated Services Digital Broadcasting) was standardized for terrestrial digital audio and data broadcasting system in Japan after system verification experiments done in November 1999. The transmission scheme is based on Band Segmented Transmission - Orthogonal Frequency Division Multiplexing (BST-OFDM). There are two kinds of bandwidths: a single 430 kHz segment, and an extended version using three such segments. The information bit rate depends on the bandwidth and transmission parameters; it ranges from 280 kbps to 5.3 Mbps. Some data services and two-channel stereo audio service can transmit on a single segment with sufficient robustness transmission parameters.
    System verification experiments showed that the field strength necessary for mobile reception when each OFDM carrier was modulated by DQPSK was 38 dBV/m, and 44 dBV/m when the carriers were modulated by 16QAM.
    Terrestrial digital audio and data broadcasting technology has been researched and developed all over the world. The key requirements for this type of broadcasting system are robustness against multi-path and fading interference during mobile reception, and easy introduction even when frequency bands are congested.
    ISDB-TN (Narrow-band Terrestrial Integrated Services Digital Broadcasting) has been standardized in Japan. It uses segmented Orthogonal Frequency Division Multiplexing (OFDM); at a bandwidth of about 430 kHz. Because this bandwidth is narrow, it takes full advantage of vacant frequencies during congested frequency band situations. In addition, broadcasters can have uniquely allocated own channels, making sharing unnecessary.
    System verification experiments using ISDB-TN were conducted to measure the associated mobile reception characteristics. This report discusses the features of ISDB-TN and the characteristics of mobile reception using single-segment transmission.

    ISDB-TN is a mobile multimedia broadcasting scheme that employs the same transmission method as wide-band ISDB-T, Japan's digital terrestrial television broadcasting scheme. It allows the use of MPEG-2 AAC (Advanced Audio Coding) for audio coding, and MPEG-2 systems for multiplexing. This makes it compatible with many multi-media services. It also uses OFDM as a modulation method. The transmission characteristics relating to the fading environment have been improved by time interleaving.
    Table 1 shows the transmission parameters of ISDB-TN. Transmission is done using either one or three OFDM segments. Three kinds of modes are used, depending on the OFDM carrier spacing. The carrier spacings are approximately 4, 2, and 1 kHz, which translate into effective symbol durations of 252, 504, and 1008 s.
    The modulation scheme for each carrier can be selected from among four types - DPQSK, QPSK, 16QAM, and 64QAM. The guard interval can be selected from among four lengths: 1/4, 1/8, 1/16, and 1/32 of the effective symbol length. Error correction is carried out using concatenated coding of convolutional codes and Reed-Solomon codes; the encoding rate of the convolutional codes can be selected among five steps: 1/2, 2/3, 3/4, 5/6, and 7/8. The time interleave can also be selected from among four possibilities, from 0 msec to about 1 sec.

TABLE 1 - Transmission parameters of ISDB-TN

TABLE 2 - Service examples of ISDB-TN

    The information bit rate of the system ranges between 280 kbps and 5.3 Mbps depending on the combination of transmission parameters. Table 2 shows two examples of services, using the encoding rate of covolutional code 1/2 - most robust in this system.
    In the carrier modulation scheme DQPSK, one stereo audio program, an Electric Programming Guide (EPG), still pictures and files can be broadcast at a bit rate within 330 kbps. In 16QAM, two stereo audio programs, an EPG, a motion picture and files can be broadcast at a bit rate within 660 kbps.

3.1 Set-up and Observation Method
    Field trials were conducted using an experimental transmitter and a vehicle set up to specifically measure transmission characteristics for mobile reception.
    The ISDB-TN signal was generated by the experimental transmitter at Tokyo Tower and distributed in the VHF-band. The transmitter's specifications are given in Table 3.
    The field strengths and bit error rates (BERs) were measured for each combination of transmission parameters while the test vehicle made its way over a trial measurement course. We installed in a vehicle an ISDB-TN receiver, signal-level measuring equipment, error-rate measuring equipment and a computer. A GPS receiver was also installed to allow the position and speed of the vehicle as well as the signal-level and bit error rates to be logged automatically on the computer. Figure 1 shows a block diagram for the mobile measurement setup. A dipole antenna was folded on the roof of the measurement vehicle. Its height was 1.8 m. Measurements were made once per second and the field strength was the average of the received values. The BERs were the average ones per second after the outer code decoding.
    This paper only describes the results of 1-segment transmission, because those of 3-segment transmission can be estimated by allowing for the difference in those bandwidths (430 kHz vs. 1287 kHz).

Figure 1. Block diagram for measurements

TABLE 3 - Specifications of experimental transmitter

3.2 Wide-Area Mobile Reception
    The reception characteristics were measured within an 80 km diameter of the experimental transmitter, to find the field strength necessary for mobile reception. Table 4 shows the measured transmission parameters. The encoding rate for the inner code was set at 1/2, and the modulation schemes used were DQPSK and 16QAM.
    Figure 2 shows the measurement course. For each modulation scheme, the vehicle covered approximately 700 km on expressways and main arterial roads.

TABLE 4 - Transmission parameters for wide-area mobile reception trial

Figure 2. Wide-area mobile reception trial measurement course

3.2.1 Definition of correct reception time rate
    For all given field strength, we defined the "correct reception time rate" as the ratio of the number of samples in which errors did not occur vs. the total number of samples for that field strength. The formula defining the correct reception time rate is shown below.
For any given field strength E,

3.2.2 Required field strength for wide-area mobile reception
    Figure 3 shows the correct reception time rates on the measurement course. The field strengths with correct time reception rates of 95%, 98%, and 99% have been selected from this graph and laid out in Table 5.
    To obtain a correct reception time rate of 99%, DQPSK modulation required a field strength of 38 dBV/m and 16QAM required a field strength of 44 dBV/m. Based on these results, we found that DQPSK was approximately 6 dB more robust than 16QAM in this mobile reception environment.

Figure 3. Correct reception time rates in wide-area mobile reception trial

TABLE 5 - Field strengths needed for specific correct reception time rates (dBV/m)
Correct reception time rate 95% 98% 99%
DQPSK 33 35 38
16QAM 40 41 44

3.2.3 Margin for mobile reception
    We calculated the margin for mobile reception using the required field strengths derived from the results of indoor experiments, and the measured values for noise. Table 6 shows the link budget we derived from our wide-area experiments. The required CN ratio for mobile reception was 18.1 dB for DQPSK, and 24.1 dB for 16QAM.
    The margin for mobile reception can be thought of as the difference between the required CN ratio for mobile reception and that based on Gaussian noise (including equipment deterioration). The margins for mobile reception were 11.6 dB for DQPSK and 13.2 for 16QAM.

TABLE 6 - Link budget derived from wide-area experiments

3.3 Urban-Area Mobile Reception
    Next, we studied the correct reception time rate for mobile reception in medium-size cities under typical traffic conditions. Cities located 20-30 km from the experimental station were selected for this experiments. Table 7 shows the transmission parameters.

TABLE 7 - Transmission parameters for urban area mobile reception trial

3.3.1 Field strengths for each city and correct reception ratio
    Table 8 shows the measured data for each city, including the median and maximum field strength. As the distance from the experimental station increased, the median field strength declined, and the correct-reception ratio also decreased.

TABLE 8 - Field strengths and correct reception time rates for cities in urban-area trial

3.3.2 Required field strength for urban-area mobile reception
    Figure 4 shows the correct reception time rates for all four cities. The required field strengths to obtain correct time reception rates of 95%, 98%, and 99% are shown in Table 9. The required field strengths were 2 or 3 dB lower than those for the wide-area.
    One difference between the wide-area and urban-area experiments was the length of time that the vehicle was motionless due to signal lights, traffic jams, and other factors. The motionless time in the wide-area trials was 15.1% of the total measured time, but as much as 44.0% in the urban-area trials. We derived the correct reception time rates separately for when the vehicle was motionless and when it was moving. These rates are shown in Figure 5. When the vehicle was motionless, the correct reception time rate was high even when the field strength was low. Consequently, the required field strength in the urban areas was relatively low.

Figure 4. Correct reception time rate in urban-area mobile trials

TABLE 9 - Field strengths needed for specific correct reception time rate

Figure 5. Difference in correct reception time rate for both moving and motionless conditions

    We discussed some features of ISDB-TN. The field strengths required for mobile reception in actual operating environments were discussed based on the results of experiments. The field strength required for a correct reception time rate of 99% was 38 dBV/m for DQPSK and 44 dBV/m for 16QAM.
    The mobile reception margins, based on the results of wide-area experiments, ware 11.6 dB for DQPSK and 13.2 dB for 16QAM.


[1] K.Tsuchida et. al., Transmission performance of narrow-band OFDM for DSB. ITE Tech. Rep., Vol. 22, No. 34, pp. 7-12, June 1998
[2] K.Tsuchida et. al., Transmission performance of terrestrial digital audio and data broadcasting in Japan. ITE Tech. Rep., Vol. 24, No. 19, pp. 25-30, February 2000
[3] N.Iai et. al., High performance of narrow-band ISDB-T system for terrestrial digital audio and data broadcasting. NAB 54th Annual Broadcasting Engineering Conference, April 2000

Copyright 2000 NHK (Japan Broadcasting Corporation) All rights reserved. Unauthorized copy of the pages is prohibited.