Coding

The mandatory coding scheme is the tail-biting convolutional coding. Each FEC block is encoded by the binary convolutional encoder, which have native rate of 1/2, a constraint length equal to K=7. The following generator polynomials are used to derive its two coded bits:

Y G

X G

OCT OCT

For 133

For 171

1 1

=

=

The generator is depicted in Figure 2-9.

LSB MSB

15

Figure 2-9 : Convolutional encoder of rate 1/2

Each block is encoded by a tail-biting convolutional encoder, which is achieved by initializing the encoder’s memory with the last data bits of the FEC block being encoded.

The puncturing patterns and serialization order that is used to realize different code rates are defined in Table 2-4. In the figure, “1” means a transmitted bit and “0”

denotes a removed bit, whereas X and Y can be referred to Figure 2-9.

Code Rates 1/2 2/3 3/4

d

free 10 6 5

X

1 10 101

Y

1 11 110

XY

X1Y1 X1Y1Y2 X1Y1Y2X3

Table 2-4 : Puncture configuration

2.2.4 Bit-interleaving

The interleaver is defined by a two-step permutation operation. The first step ensures that adjacent coded bits are mapped onto nonadjacent subcarriers. The second

16

permutation insures that adjacent coded bits are mapped alternately onto the least or most significant bits of the constellation, thus avoiding long runs of lowly reliable bits.

(C) Interleaver

Let Ncpc be the number of coded bits per subcarrier, i.e., 2, 4, or 6 for QPSK, 16-QAM, or 64-QAM, respectively. Let s=Ncpc/2. Within a block of Ncbps (number of coded bits per symbol) bits at transmission, let k be the index of the coded bit before the first permutation, mk be the index of that coded bit after the first and before the second permutation, jk be the index after the second permutation, just prior to modulation mapping, and d be the modulo used for the permutation.

The first step permutation is defined as:

(

^/

)

⋅ mod( ) + ^{( / )} =^{0,1, ,} −¹ =¹⁶

=

N d k floor k d k N d

m

_{k cbps d} L _cbps (2-1)

The second step permutation is defined as:

( )

The de-interleaver, which performs the inverse operation of the interleaver, is also defined by two permutations. Within a received block of Ncbps bits, let j be the index of a received bit before the first permutation, mj be the index of that bit after the first and before the second permutation, and kj be the index after the second

permutation, just prior to delivering the block to the decoder.

The first step permutation is defined as:

17

The second step permutation is defined as:

( )

Repetition coding can be used to further increase signal margin over the original modulation and coding mechanisms. Let the repetition factor be R. For uplink, the allocated slots are repeated R times. For downlink, the number of the allocated slots is in the range of [R×K,R×K＋(R－1)], where K is the number of the required slots before applying the repetition scheme. Here, R can be 2, 4, or 6. For example, when the required number of slots before the repetition is 10 (K=10) and the repetition factor R is 6, then the number of the allocated slots (Ns) for the burst can be from 60 slots to 65 slots.

Thus, the binary data that fit into a region that is repetition coded is reduced by a factor R compared to an unrepeated region of the slots with the same size and FEC code type. After FEC and bit-interleaving, the data is partitioned into slots, and each group of bits designated to fit in a slot will be repeated R times to form R contiguous slots following the normal slot ordering that is used for data mapping.

2.2.6 Modulation

After repetition, the data bits then enter serially to the constellation mapper.

Gray-mapped QPSK, 16-QAM, and 64-QAM are supported, as shown in Figure 2-10.

Each M interleaved bits (M=2, 4, 6) are mapped to the constellation bits b(M-1)-b0 in MSB first order (i.e., the first bit should be mapped to the higher index bit in the

18

constellation), in addition, the M bits shall be ordered MSB first.

Figure 2-10 : Modulation constellations

2.3 Subcarrier Allocation

Subcarrier allocation can be performed according to following subcarrier usages:

partial usage of subchannels (PUSC) and full usage of the subchannels (FUSC) in the downlink, PUSC in the uplink, and other optional usages. PUSC is corresponds to allocate some of the subchannels to the transmitter; FUSC to allocate all subchannels to the transmitter.

The OFDMA frame may include multiple zones include PUSC, FUSC, optional FUSC, and adaptive modulation and coding (AMC). In any case, the first zone must be the PUSC zone to ensure the successful receiption of DL_Frame_Prefix within the FCH and DL-MAP. This is because the SS does not have any message regarding allocation before first zone. Furthermore, the transition between zones is indicated in

19

the DL-MAP by the STC_LD_Zone. No DL-MAP or UL-MAP allocation can span over multiple zones. Figure 2-11 depicts OFDMA frame with multiple zones.

Figure 2-11 : Illustration of OFDMA frame with multiple zones

The downlink PUSC must appear in every frame. This thesis focuses on the subcarrier allocation of DL-PUSC.

2.3.1 Introduction to DL PUSC

The symbol is first divided into basic clusters (half slot) and null carriers are allocated. Then, pilots and data carriers are allocated within each cluster. Table 2-5 summarizes the parameters for DL PUSC.

Parameter Value Comments

Number of DC subcarriers 1 Index 1024 (counting form 0) Number of Guard subcarriers,

Left

184

Number of Guard subcarriers, Right

183

Number of used subcarriers 1681 Number of all subcarriers used within a

20

(Nused) symbol, including all possible allocated pilots and the DC carrier.

Number of subcarriers per cluster

14

Number of cluster 120

Renumbering sequence 1 Used to renumber clusters before allocation to subchannels:

6,108,37,81,31,100,42,116,32,107,30,93,54 ,78,10,75,50,111,58,106,23,105,16,117,39, 95,7,115,25,119,53,71,22,98,28,79,17,63,2 7,72,29,86,5,101,49,104,9,68,1,73,36,74,43 ,62,20,84,52,64,34,60,66,48,97,21,91,40,10 2,56,92,47,90,33,114,18,70,15,110,51,118, 46,83,45,76,57,99,35,67,55,85,59,113,11,8 2,38,88,19,77,3,87,12,89,26,65,41,109,44,6 9,8,61,13,96,14,103,2,80,24,112,4,94,0 Number of data subcarriers in

each symbol per subchannel

24

Number of subchannels 60 Basic permutation sequence 12 (for 12 subchannels)

6,9,4,8,10,11,5,2,7,3,1,0

Basic permutation sequence 8 (for 12 subchannels)

4 7,4,0,2,1,5,3,6

Table 2-5 : OFDMA downlink carrier allocations-PUSC

21

Figure 2-12 shows the partition from a frame to a subcarrier in frequency domain.

One frame of used subcarriers (1680 subcarriers excluding DC) is divided into three segments, each segment of 560 subcarriers is divided into 20 subchannels, each subchannel of 28 subcarriers is divided into two clusters, and each cluster (14

subcarriers) is divided into 12 data subcarriers and 2 pilot subcarriers. The location of the pilot subcarrier is also depicted in the figure. Note that the slot and the cluster are two-dimensional, and one cluster corresponds to exactly a half slot.

Figure 2-12 : Partition for PUSC units

2.3.2 Data Mapping

After modulation, the constellation-mapped data are subsequently allocated onto slots and then to the burst area (OFDMA data region). The subcarrier allocation within a slot uses the algorithms described below, as also depicted in Figure 2-13.

z Modulated data shall span continuous 24 data subcarriers and two consecutive OFDMA symbols. Map the data such that the lowest numbered QAM symbol

1 subchannel 12 data ; 2 pilots

Segment 0

Segment 1

Segment 2

a cluster

20 subchannel

slot

22

maps to the lowest numbered subcarrier and the lowest numbered OFDM symbol.

z Continue the mapping such that the subcarrier index is increased. When the edge of the slot is reached, continue the mapping from the lowest numbered subcarrier in the next available OFDMA symbol.

Figure 2-13 : Mapping subcarriers to a slot

The slots allocation within a burst uses the algorithms described below, and also depicted in Figure 2-14.

z Each slot shall span one subchannel and two OFDMA symbols. Map the slots such that the lowest numbered slot occupies the lowest numbered subchannel in the lowest numbered OFDMA symbol.

z Continue the mapping such that the OFDMA subchannel index is increased.

When the edge of the burst is reached, continue the mapping from the lowest OFDMA symbol

24 data subcarriers

Subcarrier

23

numbered subchannel in the next available OFDMA symbol.

Figure 2-14 : Mapping slots to a burst

2.3.3 Permutation

The permutation is a function similar to the subcarrier interleaving. It exchanges the location of subcarriers to resist frequency-selected channel fading.

One segment is divided into one big group of 12 subchannels and another small group of 8 subchannels. The number of subchannels per group is denoted Nsubchannels

and the number of subcarriers per subchannel is denoted Nsubcarriers. As mentioned, one subchannel contains 24 data subcarriers in PUSC. Then, allocating subcarriers in each group to subchannels for each OFDMA symbol using (2-5), called a permutation formula.

24

{

_{s k subchannels}

}

_subchannels

k s

subchannel

n p n N DL PermBase N

N

sbucarrier

is the subcarrier index of subcarrier k in subchannels, k is the subcarrier-in-subchannel index form the set

[0…

N

_subcarriers-1],

s is the index number of a subchannel, from the set [0…

N

_subchannels-1],

n

k =(

k

+13⋅

s

)mod

N

_subcarriers

s subchannel

N

is the number of subchannels, 12 in even group and 8 in odd group

s subcarrier

N

is the number of data subcarriers per subchannel in each OFDMA symbol,

] [ j

p

_s is the series obtained by rotating basic permutation sequence cyclically to the left s times,

DL_PermBase is an integer ranging from 0 to 31, which is set to preamble IDcell in the first zone and determined by the DL-MAP for other zones

On initialization, an SS must search for the downlink preamble. After finding the preamble, the SS will know the IDcell used for permutation. For example, if IDcell in received preamble is “2”, the even group permutation is performed as that in Figure 2-15.

25

Figure 2-15 : Permutation concept

2.3.4 Renumbering

Since PUSC subcarrier permutation is conducted in each group separately, subcarriers will not be exchanged among groups. The function of the renumbering is to interleave subcarriers among groups. The operation is described below, and also depicted in Figure 2-16.

Renumbering uses a cluster as the basic unit to exchange the logical clusters into physical clusters using (2-6):

0123 4 456 7 789……….

23

0123 4 456 7 789……….

23

0

11

0123456789…….

15

………

28

…………

264

……….

277

………..

287

k=

s =

Subcarrier(k,s)

26 For example, the 2nd (from 0) physical cluster is derived from the 37th logical cluster.

Figure 2-16 : Renumbering concept

2.3.5 Pilot and Subcarrier Randomization

Pilots, similar to the preamble, are known by the receiver before data

transmission. They are inserted into data sequence instead of the beginning of the frame. The receiver can use pilots for synchronization and channel estimation. In PUSC, the pilots are inserted with a regular pattern in a cluster. For the even OFDMA symbol (including first symbol just after preamble), pilots are located on 4th and 8th

Group 0

27

(from 0) subcarriers in a cluster, and for the odd symbol, they are located on 0th and 12th subcarriers in a cluster.

The subcarrier randomization is different from the data randomization. The value of pilots and that of subcarrier randomization are generated to the same PRBS

sequence. They use the same PRBS generator to modulate pilot signals and the randomization sequence. The subcarriers will be multiplied by the sequence to randomize the subcarriers. The PRBS generator depicted in Figure 2-17 should be used to produce a sequence, ‘wk’, where ‘k’ is physical (after renumbering) used subcarrier index. The polynomial for the generator is ‘X¹¹+X⁹+1’.

Figure 2-17 : PRBS generator for pilot modulation and subcarrier randomization The initialization vector of the PRBS generator for both uplink and downlink is designated as b10~b0, such that:

z b0~b4 = Five least significant bits of IDcell as indicated by the frame preamble in the first downlink zone and in the downlink AAS zone with

b0 b1 b2 b3 b4 b5 b6 b7 b8 b9 b10

28

Diversity_Map support, DL_PermBase following STC_DL_Zone_IE, and 5 LSB of DL_PermBase following AAS_DL_IE without Diversity_Map support in the downlink. Five least significant bits of IDcell (as determined by the preamble) in the uplink. For downlink and uplink, b0 is MSB and b4 is LSB, respectively.

z b5~b6 = Set to the segment number + 1 as indicated by the frame preamble in the first downlink zone and in the downlink AAS zone with Diversity_Map support, PRBS_ID as indicated by the STC_DL_Zone_IE or AAS_DL_IE without Diversity_Map support in other downlink zones, and 0b11 in the uplink. For downlink and uplink, b5 is MSB and b6 is LSB, respectively.

z b7~b0 =0b1111 (all ones) in the downlink and four least significant bits of the Frame Number in the uplink. For downlink and uplink, b7 is MSB and b10 is LSB, respectively.

For example, the initialization vector of the PRBS generator is b10~b0 = 10101010101, the result in the sequence wk = 1010101010100000… in the uplink.

A new value is generated by the PRBS generator for every subcarrier up to the highest numbered usable subcarrier, in the order of physical subcarriers, including the DC subcarrier and usable subcarriers that are not allocated. Furthermore, The PRBS generator should be clocked “n+Nused-1” times, n is equal to symbol offset mod 32 and Nused is the number of used subcarriers, where symbol offset is counted from the first symbol in each zone. This is because the beginning of the PRBS sequence of the symbol would be shifted by its symbol offset. Consider DL PUSC, let w0,w1,w2,…

be the bits generated after the initialization. The subcarriers of the first symbol in the zone (with symbol offset of zero) use the bits w0,w1,w2,…,w1680. For the second

29

symbol (with symbol offset of 1) use the bits w1,w2,w3,…,w1681. Thus, if there are 30 symbols in first PUSC zone, the PRBS generator should be clocked

“30+1681-1=1710” times.

After generating wk, the pilot subcarrier ck should be modulated according to following equations.

z For the mandatory tile structure in the uplink, and for the other allocations such as TUSC1/TUSC2 structures in the downlink:

0

z In all permutations except uplink PUSC and downlink TUSC1, each pilot shall be transmitted with a power boosted of 2.5 dB higher than the average

non-boosted power of each data tone:

0

Where pk is the pilot’s polarity for spatial division multiple access (SDMA) allocations in AMC AAS zone, and p = 1 otherwise.

The subcarrier randomization is accomplished as each modulated data subcarrier is multiplied by the factor ⎟

⎠

⎜ ⎞

⎝⎛ −

w

_k 2

2 1 . In DL PUSC, we can combine the pilot and

randomization by inserting the value of 4/3 into pilot subcarriers and then multiplying both pilot and data subcarriers with the same factor ⎟

⎠

Figure 2-18 illustrates the OFDMA PHY transmitter architecture in DL PUSC.

The channel coding includes concatenation, data randomization, convolutional

30

encoding, bit-interleaving, repetition encoding, and modulation. The subcarrier allocation includes data mapping, permutation, renumbering, pilot insertion, and subcarrier randomization. Then, passing the resultant signal through IFFT and adding preamble, we can complete the whole baseband processing for transmission.

Figure 2-18 : Baseband transmitter

Figure 2-19 illustrates the OFDMA PHY receiver architecture in DL PUSC. It consists of an inner and outer receiver. The outer receiver reverses the operation

Segment 0

Pilot Adding & Randomization

blo bl bl

Segment 1 Segment 2

Renumbering [1680]

MAC

IFFT [2048]

IFFT IFFT

Preamble Adding

Preamble Preamble

31

conducted in the transmitter, while the inner receiver performs synchronization, channel estimation, and equalization.

Figure 2-19 : Baseband receiver Segment 0

Pilot Adding & Randomization

blo bl bl

Segment 1 Segment 2

Renumbering [1680]

MAC

FFT [2048]

FFT FFT

Inner Receiver

32

Chapter 3: Synchronization

3.1 Preamble Structure

The preamble, having a specific pattern and containing symbols known to the receiver, is usually used for synchronization. Here synchronization is defined in a broader sense, which includes packet detection, frequency offset estimation, timing offset estimation, and channel estimation. The length and contents of the preamble are specially designed such that information for synchronization is just enough avoiding any unnecessary overhead.

3.1.1 Frequency Domain Data

As mentioned, the thesis focuses on the FFT size of 2048 (IEEE 802.16-2004).

The data transmitted in the preamble are defined in the frequency domain. Equation (3-1) shows subcarrier allocation for the preamble data, which is modulated with a boosted BPSK modulation scheme with a specific pseudo-noise (PN) code described below.

k n rrierSet

PreambleCa

_n = +3⋅ (3-1)

where

rrierSet

n

PreambleCa

specifies all subcarriers allocated to the specific preamble

N

in the number of the preamble carrier-set indexed 0~2

K

is a running index 0~567

Each segment uses one type of preamble out of three sets:

z Segment 0 uses preamble carrier-set 0

z Segment 1 uses preamble carrier-set 1

33

z Segment 2 uses preamble carrier-set 2

Thus, each segment eventually use one third of subcarriers for preamble transmission.

In the case of segment 0, the DC carrier will not be used and the corresponding PN will be discarded. In other words, DC carrier should always be zeroed. For the preamble symbol, there will be 172 guard band subcarriers on the left and right sides of its used spectrum. Figure 3-1 illustrates the distribution of three carrier-sets.

Figure 3-1 : Preamble structure in the frequency domain

The PN series modulating the preamble carrier-set are defined in Table 309[2]

[3] . Note that the preamble series table includes PN sequence in a Hexadecimal format. In PUSC, there are total 114 preamble series; the actual series used depends on the designated segment and the IDcell parameter. Each segment can have 38 different preamble series. The series specified is mapped onto the preamble

subcarriers in ascending order. The PN series is converted into to a binary series (wk) with values of plus and minus one (0 mapped to +1 and 1 mapped to -1).

3.1.2 Time Domain Waveform

Once the preamble data in the frequency domain are defined, the preamble waveform in the time domain can be obtained with IFFT. Since subcarriers allocated for the preamble series corresponds to a downsampled version of available subcarriers

Preamble carrier-set 0

Preamble carrier-set 1

172 174 176 1021 1023 1025 1027 1700 1702 1704

Preamble carrier-set 2

34

(the downsampling factor is three), the time-domain preamble waveform is periodic (three periods in the preamble), as depicted in Figure 3-2.

Figure 3-2 : Preamble structure in time domain

Note that the preamble is not exactly periodic. This is because the IFFT size is 2048, which is not divisible by three. In other words, the preamble contains three signal sections, which are similar but not exactly the same. Even though, the periodic property of the preamble can still be used for the purpose of

synchronization.

3.2 Channel Model

The channel between BS and SS can be roughly divided into two types. The first one is the actual signal propagation channel. In wireless systems, this corresponds to the mulipath fading channel, and the packet timing offset. The second one is the non-ideal effect of analog and RF components, which will introduce distortions into the received signal. The channel model we used is a combination of these two channel effects. For the second type of channel effect, we consider the sampling timing offset and the carrier frequency offset. Solutions have been developed to cope with these effects and will be described in next section.

Preamble waveform (2048 point)

A1 A2 A3 Data

D1 D2 D3

A1＝A2＝A3 D1≠D2≠D3

35

3.2.1 Multipath Fading Channel

Multipath fading occurs when the transmitted signal arrives at the receiver via multiple propagation paths having different delays and different intensities. These signal components may add destructively or constructively in the receiver. Note that the multipath channel will extend the transmitted signal period, as shown in Figure 3-3, which may result in inter-symbol interference (ISI) in the OFDM system. The received signal can be described in the time domain as:

]

where r is received signal, his channel impulse response, sis transmitted signal,wis AWGN,

i is path index, and

nis sample index. In the frequency domain, the received signal can be written as:

]

where R [k] is the received frequency-domain signal, H [k] is the channel frequency response, S[k] is the transmitted frequency-domain signal, and W[k] is AWGN, all for subcarrier k.

Figure 3-3 : Multipath fading effect Transmitted

Delay, scale, and add

36

IEEE 802.16 task force does not specify channel models for simulations. We then use the spatial channel model (SCM) provided by the 3GPP. The channel environment of 3G systems, defined for wireless metropolitan area, is similar to that of 802.16. The SCM channel model can support fixed and mobile wireless multipath fading channels. Three different types of models are defined:

z Suburban macro-cell: The cell coverage is 1~6 Km. The BS antennas are set above rooftop height, ranging from 10m to 80 m. The average height is 32 m.

The SS’s mobile speed is between 0~250 Km/hr.

z Urban macro-cell: The cell coverage is 1~6 Km. The BS antennas are above rooftop height, their height is 10~80 m, and the average height is 32 m. The SS’s speed is 0~250 Km/hr. Note that these settings are the same as those in suburban macro-cell; others may be different.

z Urban micro-cell (approximately 1Km distance from BS to BS): The cell coverage is 0.3~0.5 Km. The BS antennas are set at rooftop height; the average height is 12.5 m. The SS’s speed is between 0~120 Km/hr.

3.2.2 Timing Offset

In the actual operation, the receiver cannot know when the packet will arrive.

Thus, it has to perform some kind of detection. After that, the receiver has to

synchronize the received symbol boundary. Timing offset also called “symbol offset”

synchronize the received symbol boundary. Timing offset also called “symbol offset”

在文檔中 IEEE 802.16 OFDMA基頻傳收機設計與實現 (頁 28-0)