The DVB organization was formed in September 1993. Along these years, several system specifications have been proposed and become standard in the European Telecommunication Standard Institute (ETSI) or European Committee for Electrotechnical Standardisation. The term “OFDM-based DVB” system is often rewritten by the “DVB-T” system. The DVB-T system for terrestrial broadcasting is probably the most complex DVB delivery system. The key feature of this system is the employment of COFDM, which is a very flexible wide-band multicarrier modulation system using different levels of forward error correction, time and frequency
transmitted information is split into a given number (“2k” 1705 or “8k” 6817) of modulated carriers with individual low bit rate, so that the corresponding symbol time becomes larger than the delay spread of the channel. A guard interval (1/4, 1/8, 1/16, 1/32 of the symbol duration) is inserted between successive symbols to avoid ISI and to protect against echoes. Depending on the channel characteristics, different parameters (subcarrier modulation: QPSK, 16QAM, and 64QAM; number of carriers: 2k and 8k;
code rate of inner protection, and guard interval length) can be selected leading to different operation modes. Every mode offers a trade-off between net bit rate and protection of the signal (against fading, echoes, etc.). Depending on the selected operation mode, 60 different net bit rates could be obtained ranging from 5 to 32 Mbps.
The selection of the COFDM modulation system presents two major advantages that make its use very interesting to terrestrial digital video broadcasting:
1. COFDM improves the ruggedness of the system in the presence of artificial (long distance transmitters) or natural (multiple propagation) echoes. The echoes may benefit instead of interfering the signal if they fall inside the guard interval.
2. COFDM provides a considerable degree of immunity to narrow-band interferers as maybe considered the analogue TV signals; and on the other hand it is seen by those analogue signals as white noise, therefore not interfering or having little effect upon them.
These characteristics enable a more efficient use of the spectrum and the introduction of single frequency networks (SFN). Furthermore, some characteristics of a DVB-T system, as shown in Fig. 3.2, will be discussed as below.
3.2.1 Scrambler
This is a method of removing long runs of 0’s or 1’s in the signal which would otherwise give the receiver a problem at the other end of the channel. This is achieved at the bit level. The pseudo random sequences produced from the incoming bit stream are generated by the polynomial given by
14 15
1 x+ +x (3.1)
This is realized, as shown Fig. 3.3, with a shift register. Bits 14 and 15 are fed into an XOR gate, then back to the input and also XOR’ed with the data, when enabled, to give
the output. The initialization sequence is loaded every 8 transport packets. Note that this gives a pseudo random binary sequence (PRBS).
3.2.2 Outer Coder (Reed-Solomon Coder)
Reed-Solomon (RS) coding is mathematically very complicated and will not be explained here in any detail. The aim here is simply to give an overview of the coding technique. The coding is a block level code, i.e., it operates over a block of data. The block of data must therefore be constructed prior to the code operation. This puts an overhead on the system in terms of memory and the need for block synchronization.
The coding adds 16 additional bytes to the 188 byte transport stream packets, making a final transport stream packet size of 204 bytes. This error correction algorithm, as applied to a transport stream, is characterized by the three numbers n, k and l, where n is the number of bytes in the final transport stream, k is the number of bytes of the original transport stream, and t is the number of bytes that can be corrected. We set the parameters to be n=204, k=188, and t= . This is referred to as RS (204, 188, 8) 8 shortened code. The RS code effectively specifies a polynomial by generating a large number of points. The RS code can detect and correct up to errors. The field generator polynomial is used as
(n−k) /2
15
8 4 3 2
( ) 1
P x =x +x +x +x + (3.2)
with the code generator polynomial as below
0 1
( ) ( )( ) ( )
G x = x+α x+α x+α (3.3)
3.2.3 Outer Interleaver (Convolutional Interleaver)
Interleaving is a form of time diversity that is employed to disperse bursts of errors in time. A sequence of data symbols is interleaved before transmission over a bursty channel. If errors occur during transmission, reshaping the original sequence to its original ordering has the effect of spreading the error over time. By spreading the data symbols over time, it is possible to use channel coding, such as RS coding in DVB-T systems, which protects the data symbols from corruption by channel. The interleaver performance depends on the required memory for data storage and the delay in interleaving and deinterleaving, which should be kept as small as possible. In [23],
we obtain that the delay with convolutional interleaver is half the operation time compared with block interleaver and an efficiency of 14.2% for block interleaver and 64% for convolutional interleaver, where efficiency can be defined as the ratio of the length of the smallest burst of errors that can cause the errors correcting capability of the code to be exceeded to the number of memory element used in the interleaver. As stated above, the convolutional interleaver is applied in the outer interleaver of a DVB-T system. The outer coding can correct up to 8 bytes in a transport stream packet.
Clearly, if a burst error condition occurs, i.e. a burst of energy from some noise source, then more than 8 bytes within the same packet could become corrupted. The convolution effectively takes these errors and spreads them out over a number of packets, thus allowing the outer coding to be more effective. Fig. 3.4 shows the outer interleaver and deinterleaver. Data is input from the RS outer coding and output to the convolution inner coder. Therefore, there are 12 individual branches with the largest first-input first-output (FIFO) being of length 187 bytes. Input bytes will therefore be delay by 17, 34, 51, …, 187 bytes, depending on the byte index. Fig. 3.5 shows the data format of passing though the scrambler, outer coder and interleaver.
3.2.4 Inner Coder (Convolutional Coder)
Convolutional coding operates at the bit level rather than block level such as RS coding. This has the advantage of the generator not having to store a whole block of data in expensive memory prior to performing the coding. The input stream is fed into shift register stages which have intermediate output taps after each stage. The input stream and various output taps are modulo two added. The DVB-T system actually uses 6 shift register stages and the generator polynomials are shown as
1 O The architecture as shown in Fig. 3.6 can be considered a state machine. Since there are
6 stages in the DVB-T implementation, this gives rise to 64 states. It is the change from one state to another based on which we can draw the trellis diagram to be applied in the Viterbi decoding algorithm.
3.2.5 Inner Interleaver
Inner interleaver is also performed to basically spread out the errors and so make the inner coding more effective. The inner interleaving consists of bit-wise interleaving followed by symbol interleaving. Both the bit-wise interleaving and the symbol interleaving processes are block-based.
Bit-Wise Interleaver
The input, which consists of up to two bit streams, is demultiplexed into v sub-streams, where v = 2 for QPSK, v = 4 for 16QAM, and v = 6 for 64QAM in Fig.
3.7. The block size is the same for each interleaver, but the interleaving sequence is different in each case. The bit interleaving block size is 126 bits. The block interleaving process is therefore repeated exactly twelve times per OFDM symbol of useful data in the 2k mode and forty-eight times per symbol in the 8k mode. For each bit interleaver, the input bit vector is defined by
,0 ,1 ,2 ,125
( ) ( e , , , e e ..., )e
B e = b b b b (3.5)
where be w, denotes the bit number w of inner bit interleaver e, B e( ) denotes the input vector to inner bit interleaver e, and e ranges from 0 to v-1. The interleaved output
is defined by
where denotes the bit number w of inner bit interleaver output stream e, and is a permutation function which is different for each interleaver. The
is defined as follows for each interleaver
,
The outputs of the v bit interleavers are grouped to form the digital data symbols, such that each symbol of v bits will consist of exactly one bit from each of the v interleavers.
Hence, the output from the bit-wise interleaver is a v bit word y', i.e.
0, 1, -1,
'w ( w, ,..., )w v
y = a a a w (3.8)
Symbol Interleaver
The purpose of the symbol interleaver is to map the 2, 4 or 6 bit words onto one of the OFDM carriers (1512 for 2k mode or 6048 for 8k mode). The interleaved vector
is defined by
R takes the following values:
max The permutation function H q( ) is defined by the following algorithm
max
A schematic block diagram of the algorithm used to generate the permutation function is represented in Fig. 3.8 for the 2k mode
3.2.6 Subcarrier Modulation Mapping and OFDM Frame Structure
The system uses OFDM transmission. All data carriers in one OFDM frame are modulated using either QPSK, 16QAM, 64QAM, non-uniform 16QAM or non-uniform 64QAM constellations. The exact values of the constellation points are given by
MOD (
d =K × +n jm) (3.12)
where KMOD is the normalization factor and z∈ +{n jm} with values of is given by Table 3.2. The basic structure of what is known as an OFDM signal is a variable number of frequency carriers, either 1705 (known as 2k mode) or 6817 (known as 8k mode) as Table 3.3. These carriers are spaced in such a way as to allow them to fit into the 7.61 MHz bandwidth. These carriers can be shown in Fig. 3.9 and described as follows.
, n m
1. Data: with a variable number of bits per carrier
2. Transmission parameter signalling (TPS): transmission information
3. Pilot: for receive synchronization, there are two types both transmitted at boosted power levels
a. Continual: there are 177 tones in 8k mode, and 45 tones in 2k mode. These always are at the same frequency in different OFDM symbols. These pilots are used to compute the CPE. The CPE is an error that is introduced into the signal due to the local oscillator’s phase noise. The CPE is a change in phase of all carriers within an OFDM symbol compared to the next OFDM symbol.
b. Scattered: there are 524 tones in 8k mode, and 131 tones in 2k mode.
Specified insertion pattern within the symbol. These pilots are used to estimate the channel distortion.
On the other hand, the guard interval is a replication of the end of the symbol and is added to the beginning of the symbol. The guard interval length of DVB-T system can be 1/4, 1/8, 1/16 or 1/64 of the symbol duration. There are two main reasons for the insertion of a guard interval. The first one is to combat against the ISI. The other is to allow the receiver to identify the start of a symbol. Finally, Table 3.4 gives simulated BER performance anticipating “perfect channel estimation and no phase noise” for various combinations of channel coding and modulations.