Polytec - New York University



EE4414 Multimedia Communication System II, Fall 2005, Yao Wang

Homework 7 Solution (Digital Television)

________________________________________________________________________________

1. Describe the two possible alternatives (compatible vs. simulcast) for migrating from analog TV system to an advanced television system. What are the pros and cons of each? What make the non-compatible all-digital approach win over other options?

With a compatible approach, the new system must be designed so that an analog TV receiver can automatically retrieve a portion of the broadcast signal and play back the audio and video in analog formats. Because the much higher bandwidth of the advanced television signals, it is unlikely the entire signal can be fit into one 6 MHz channel allocated to an analog TV station, using a signal encoding and transmission technique that is compatible with existing analog systems. Therefore, one possibility to achieve compatibility is to split the new high-resolution signal into two, the current analog TV signal sent in one analog TV channel, and an augmented signal (= advanced format signal – analog TV signal) will be sent in a separate band, using either analog or digital or mixed techniques.

With a simulcast approach, the same TV program will be broadcast in two different formats, the current analog format and a new advanced format, simultaneously, using separate frequency bands. The analog format broadcast will be gradually phased out. This alternative gives the developers of the advanced TV systems more flexibility in designing new systems. If the new advanced system can send its entire signal using equal or less channel bandwidth than that for the augmented signal in the proceeding compatible approach, then the overall bandwidth efficiency is at least equal to the compatible approach. This proved to be the case, with the all-digital approach, by employing sophisticated audio coding, video coding, channel coding and digital modulation techniques. For example, with the US ATSC DTV standard, a 6 MHz channel can be used to carry either a single HDTV signal or 4 SDTV signals. The all-digital approach also offers compatibility with other newly developed applications over digital media (DVDs, computers networks, etc.), so that the same digitally compressed TV program (audio and video) can be delivered over the air, DVDs, or the Internet.

2. Describe the major components in the US DTV system. What are the audio and video formats supported? What are the audio and video coding standards used? What channel coding and modulation techniques are used?

The US DTV system includes audio coding, video coding, data multiplexing, channel coding and modulation. The audio has 5.1 channels and is compressed using Dolby AC3 standard, which is compatible with the format used in movie theaters and home theater devices. Several video formats are accepted, 2 HDTV format and 1 SDTV format, and each can operate in different frame rate/scanning format. The video is coded using MPEG2-video standard, using either mp@hl or mp@ml. Multiplexing is done following the MPEG2-system standard. Channel coding is realized by concatenating an outer Reed-Solomn code with a trellis code, with a data interleaver in between, with a total code rate of 0.6 (i.e. 60% of raw bit rate is used for carrying information bits, 40% for error correction). Modulation is accomplished using 8-VSB, which uses 8-ASK for mapping from digital to analog waveforms and use VSB to reduce the bandwidth to 6 MHz total. The raw data rate is 32.3 mbps per 6 MHz channel. The total information bit rate is 32.3*0.6=19.4 mbps per 6 MHz channel.

3. Repeat 2 for the Europe DVB system.

The Europe DVB system also consists of 5 components. Video coding and multiplexing follow the MPEG2-video and MPEG2-system standards, as with the US DTV standard. For audio, stereo sound is the standard format, coded using MPEG2-audio format (but only requires MPEG2 BC mode, which is equivalent to MPEG1 layer 2). Channel coding is quite similar, but the inner code is punctured convolution code. DVB uses a very different modulation technique. It combines QAM with OFDM.

4. How does the GOP structure support channel switching in addition to random access (after recording). What decides the maximum delay after a person switch to a different channel? What is the maximum delay considered acceptable? For fast forward, what determines the slowest forward speed? For rewind, what determines the slowest rewind speed?

Within each GOP, there is one I-picture, several P-pictures, separated by B-pictures in between. Because P-pictures and B-pictures are coded by referencing previously coded I- or P-pictures, these pictures cannot be decoded until the referenced I- or P-pictures are decoded first. Had MPEG2 uses the I-picture mode only for the first frame in a video sequence, channel surfing would not be possible, as the video decoder would not be able to decode the compressed bits in a newly switched channel, unless the switching happened in the very beginning of a video program of the new channel. With the GOP structure, after channel switching, the decoder can start to decode the new channel once an I-picture is identified, which occurs at the end of each GOP. The delay after channel-switching depends on the time when the switching happens: 0 second if it happens at the end of a GOP, half of GOP length if it happens in the middle of a GOP, or the entire GOP length if it happens in the beginning of a GOP. Therefore, the maximum delay is equal to the GOP length (=time between two I-pictures), and the average delay is half of the GOP length. The DTV standard recommends a GOP length of 0.5 second (15 frames for a 30 fps sequence), which is considered acceptable by the majority of the audience.

For fast forward, one can choose to display all the I- and P-pictures for the slowest fast forward. The minimal frame-skip is determined by the distance between P-pictures. For example, if the original sequence is 30 fps, and the GOP length is 0.5 second, and distance between P-pictures is 3 frames, then the slowest fast-forward is 3 times faster (display 1 second video in 0.33 second), by skipping only the B-pictures. (Actually, one can also play selected B-pictures, making it even slower than P-pictures only, but this may yield non-even skip, depending on the number of B-pictures between two P-pictures.) Obviously, one can skip some P-pictures and even I-pictures for faster forward. For example, if decoder decodes only the I-pictures, the speed-up is 15 to 1, if a GOP contains 15 frames.

For rewind, one can only display I-pictures. The slowest rewind is obtained when all the I-pictures are displayed. The slowest speed is thus determined by distance between I-pictures. Thus for a 30 fps video coded with 0.5 second long GOP (1 P-picture in every 15 frames), the slowest rewind is 15 times faster (display 1 second video in 1/15 second). Faster rewind can be accomplished by skipping I-pictures.

5. Describe the major components in a perceptual audio coding system. Draw the block diagram and explain the function of each block.

6. What is the purpose of channel coding? How is this achieved in principle?

The purpose of channel coding is to detect and sometimes correct bit errors in a received bit stream. This is generally achieved by generating redundancy bits based on the information bits. For example, with block coding, for every block of information bits (say k bits), r redundancy bits are generated based on some mathematical formula (like generating “hash bits” using a hash function), yielding a coded block of n=k+r bits. The coded blocks rather than the original information block are transmitted over a noisy channel. The r bits are added so that if some bits in the received coded block of n bits are erroneous, they can be corrected. With a well designed code (e.g. the Reed Solomn code), r redundancy bits enables the correction of t=r/2 error bits.

7. What is the purpose of data interleaving? What are some of the disadvantage of interleaving?

Data interleaving can break up an error burst (many consecutive bits that are wrong) into isolated bit errors in separated locations. Without interleaving, in order to correct a long error burst in one data block, one have to use a very large number of redundancy bits (i.e., t=r/2 >= burst length). By using interleaving and deinterleaving, an error burst in one coded block will be converted to single (or a few) bit errors in many data blocks, which can be corrected by a block code that uses a smaller r. (See example in lecture note)

One disadvantage associated with interleaving is delay in both the transmitter and receiver. The acceptable delay in the overall system is usually the factor that determines the allowable interleaving depth.

8. What is concatenated channel coding? And what is its benefit compared to a single channel code?

a. With concatenated channel coding, we perform two channel coding operation in serial, with an data interleaver in between. This is illustrated below.

[pic]

b. The benefit can best be explained in terms of decoder operation. As mentioned earlier, each channel code has an error correction limit. If the number of errors in a block exceeds the limit, then none of the erroneous bits can be corrected. However, the decoder can usually recognize that there are errors in this block, and consequently mark the whole block as in error and reset them all to zeros (or ones). This essentially creates a long error burst. In the case of concatenated code, after the inner decoder, a data block that contains too many error bits will not be corrected and thus creating long error bursts to the outer decoder. The deinterleaver before the outder decoder will spread this error burst into isolated errors in many blocks, which can be corrected by the outer decoder.

9. Describe the channel coding scheme used in the ATSC DTV system and the DVB system respectively.

Both ATSC and DVB use concatenated channel coding. The outer codes in both systems are Reed-Solomn block codes. The inner code in the DTV system is a trellis code, whereas the inner code in the DVB system is a convolutional code. With the ATSC DTV system, the code rate for both the inner and outer codes are fixed, with a total channel rate of 187/207* 2/3=0.6. With the DVB system, the code rate can be varied based on the channel condition. When channel is worse, more redundancy bits are added, lowering the channel code rate, reducing the information data rate that can be sent in the same channel bandwidth. DVB system thus offers the flexibility of trading off the data rate for robustness to channel errors.

10. Describe how to map digital signal to analog waveform using 4-ASK. Illustrate the waveform corresponding to each 2-bit symbol. For the following sequence of bits, 01001011, sketch the resulting analog signal. (you can assume an arbitrary carrier frequency for this exercise. For example, each symbol contains only one or two cycles of a sinusoid).

11. Describe how to map digital signal to analog waveform using 4-QAM. Illustrate the waveform corresponding to each 2-bit symbol. For the following sequence of bits, 01001011, sketch the resulting analog signal.

12. Describe the principle of 8-VSB modulation scheme used for US DTV.

The 8-VSB technique uses 8-ASK for mapping from digital bits to analog waveform. Each group of 3 bits is mapped to a sinusoid waveform, the amplitude of which depends on the actual 3 bits, and the frequency of which equal to the carrier frequency chosen. This analog waveform is then further band-limited using vestigial sideband modulation (VSB), which retains only a small portion of the lower sideband of the modulated signal. This is realized using a so-called “Nyquist filter”, so that the signal values at the sampling instances are not modified.

13. Describe the principle of the OFDM modulation scheme used for the DVB system.

The DVB system uses QAM for mapping from digital bits to analog waveforms. The actual number of bits per QAM symbol varies from 2 (4-QAM), 4 (16-QAM) to 6 (64-QAM). The analog waveform is then split into many sub-signals using sub-sampling (ex. Sub-signal one takes every 4-th sample, starting from the first sample in a 4-band decomposition), and each sub-signal is modulated to a slightly different carrier frequency. In C-OFDM adopted by DVB, the number of sub-channels is either 2048 (2k mode) or 8192 (8k mode). With the 8k mode, the sample interval in each subsignal is longer, enabling it to overcome a larger delay spread caused by multipath fading.

14. What are the pros and cons of 8-VSB vs. COFDM?

By splitting a signal to many small sub-signals each occupying a different frequency band, OFDM is more immune to frequency selective fading, and interference from adjacent channels. This is because each of the above channel impairment is likely to affect only one or a few sub-signals at the same time. The resulting bit errors on the combined signal are likely to be random and can be effectively corrected using the employed concatenated channel coding technique. OFDM is also very effective in overcoming the difficulty caused by multipath fading encountered in indoor reception. By splitting a signal at a high data rate into multiple subsignals, each subsignal has a lower data rate, so that each sample interval is much longer. By using a guard interval between sample vales, as long as the delay of a received signal is less than the guard interval, the receiver can correctly decode the actual value for each sample interval. On the other hand, 8-VSB is simpler to implement (both at the transmitter and receiver), requires less transmission power for the same coverage area, and is more immune to impulse noise.

-----------------------

Outer

decoder

Deinterleaver

Inner

decoder

Demodulator

[pic]

With the above mapping, the test sequence will have the following waveform:

[pic]

Channel

Modulation

Inner

coder

Interleaver

Outer coder

[pic]

With the above mapping, the test sequence 01001011 will have the following waveform:

[pic]

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download