Chapter 2 OFDM System Overview
2.2 Guard Interval
Guard interval is often inserted in the data stream to reduce the inter symbol interference caused by multipath propagation [15]. Guard interval is using constructed with cyclic prefix. The OFDM symbol with guard interval is shown in Figure 2.2- 1.
Generally, the length of guard interval should be longer than RMS delay spread of the channel. ISI only damages the information within guard interval. Since the signal is transmitted over multipath channel, the received signal may contain some delayed replica. We show two cases here. Case1 is two-ray multipath without ISI as shown in Figure 2.2- 2. Case2 is two-ray multipath with ISI as shown in Figure 2.2- 3.
Figure 2.2- 1 OFDM symbol with guard interval
Figure 2.2- 2 Two-ray multipath without ISI
Figure 2.2- 3 Two-ray multipath with ISI
Note that the component of guard interval can not be inserted with zeros. If we insert zeros within guard interval, the signal will lose orthogonality between every two components of the subcarriers which will result in a large degrade in the system performance after DFT operation. Since cyclic prefix still has non-zero power, system should also increase the transmit power when guard interval is longer.
The starting position of FFT window must be set properly to avoid ISI from either
the starting of guard interval of the longest delay ray and the end of guard interval of preceding wave. The delay of the longest delay ray that can avoid ISI is the length of one guard interval. The ideal position for setting FFT window is the beginning of the payload or the end of guard interval of the preceding wave so that the system can admit the longest delay wave. If the symbol timing position is set after the ideal position, ISI with the next symbol occurs, the performance of the receiver is then drastically affected.
In summary, the guard interval using cyclic prefix not only preserves the mutual orthogonality between the subcarriers but also prevents the adjacent symbols from ISI.
Here we lists the advantages and disadvantage of OFDM [14], [15] and [19] in Table 2.2- 1..
Table 2.2- 1 Advantages and disadvantages of OFDM
Advantage Disadvantage
Immunity to delay spread. With symbol and frequency synchronization problems.
Resistance to frequency selective fading.
Need FFT units at transmitter and receiver, higher complexity of
computations.
Simple equalization. Sensitive to carrier frequency offset.
Efficient bandwidth usage. With high peak-to-average power ratio.
Chapter 3 IEEE 802.16e Standard
EEE 802.16 standard [2], [3] is for Line-of-Sight (LOS) application, it utilizes 10 Ghz-66 Ghz licensed band. Its channel bandwidths of 25 Mhz or
28 Mhz are typical. With carrier frequency below 11 Ghz, it can support
near-LOS and non-LOS (NLOS), at raw data rate exceeding 120 Mb/s.3.1 Overview of WiMAX
Worldwide interoperability for microwave access (WiMAX) is the common name associated to IEEE802.16 standard. WiMAX is also a broadband wireless communication technology designed for high rate data transmission. The application of WiMAX can be applied in DSL, cable modem and so on. WiMAX does not need lots of wiring cost, and it can support wide range wireless service. WiMAX shows great
I
promise as the “last-mile” solution for bringing high-speed internet access into homes and businesses. WiMAX technology has advantages of high data rate and beyond the restriction of terrains, and has the character of mobility. WiMAX was specified in 2003.
Intel, Alvarion, Samsung, Fujitsu, Nokia and Siemens all support this standard.
IEEE Std. 802.16e is an enhanced version of IEEE Std. 802.16-2004. It can support both fixed and mobile wireless systems, and is compatible with IEEE Std.
802.16-2004. The differences between IEEE Std. 802.16-2004 (802.16d) and IEEE Std.
802.16e are listed in Table 3.1- 1.
Table 3.1- 1 Comparison with IEEE Std.802.16d and IEEE Std. 802.16e
Standard IEEE Std. 802.16-2004 (802.16d) IEEE Std. 802.16e
Publish date
June, 2004 December, 2005
Frequency
band
2 ~ 66 Ghz 2 ~ 6 Ghz
Channel
NLOS NLOS
Data rate
75 Mbps (bandwidth is 20 Mhz) 15 Mbps (bandwidth is 5 Mhz)
Modulation type
BPSK, QPSK, 16-QAM,
64-QAM BPSK, QPSK, 16-QAM, 64-QAM
bandwidth
1.5 ~ 20 Mhz 1.5 ~ 20 Mhz
Mobility
Fixed Mobile
The advantages of the WiMAX are summarized as:
a.) Suitable in both NLOS and LOS environments.
b.) High spectrum efficiency and high data rate.
c.) Support the QoS of voice and image.
Due to the cost and complexity associated with traditional wired cable could not provide satisfactory coverage and cause gaps in the broadband coverage, WiMAX is expected to extend the potential of Wi-Fi to far longer distances [1].
3.2 WirelessMAN-OFDM PHY
The wirelessMAN-OFDM PHY is based on OFDM modulation and designed for NLOS operation in the frequency bands below 11 Ghz. On initialization, a subscriber station (SS) should search all possible values of cyclic prefix (CP) until it finds the cyclic prefix being used by base station (BS). Once a specific cyclic prefix duration has been selected by the base station for operation on the downlink, it should not be changed. Changing the CP would force all the subscriber stations to resynchronize to the base station. Inverse Fourier transforming creates the OFDM waveform. The time duration
T is referred to as useful symbol time. A copy of the last
bT of the useful
g symbol period, called cyclic prefix, is use to against multipath delay spread and maintaining the orthogonality of the tones. The time structure of OFDM symbol is show in Figure 3.2- 1. Figure 3.2- 2 illustrates a frequency description of OFDM signals.Figure 3.2- 1 Symbol time structure of OFDM
Figure 3.2- 2 Frequency description of OFDM
Data subcarriers: For data transmission.
Pilot subcarriers: For various estimation purposes.
Null subcarriers: No transmission at all, for guard bands, non-active subcarriers and DC subcarrier.
3.2.1 OFDM symbol parameters and transmitted signal
Equation (3-1) specifies the transmitted signal voltage to the antenna, as a function of time during any OFDM symbol.
}
where
c → is a complex number; the data to be transmitted on the subcarrier whose
kfrequency offset index is k, during the subject OFDM symbol.
The parameters of the transmitted OFDM signal expressed as Equation (3-1) are given in Table 3.2.1- 1.
Table 3.2.1- 1 OFDM symbol parameters
Parameter Value
NFFT / Nused
256 / 200
G
1/4, 1/8, 1/16, 1/32
Frequency offset indices of guard subcarriers
-128,-127,…,-101 +101,+102,…,127
Frequency offset indices of pilot carriers
-88, -63, -38, -13, 13, 38, 63, 88
n
For channel bandwidth that are a multiple of:1.75Mhz then n=8/7, 1.5Mhz then n=86/75,
1.25Mhz then n=144/125, 2.75Mhz then n=316/275, 2.0Mhz then n=57/50; otherwise
specified then n=8/7.
3.2.1.1 Randomization
Data randomization is performed on each burst of data on the downlink and uplink. The randomization is performed in each data block, the randomizer shall be used independently. If the amount of data to transmit does not fit exactly the amount of data allocated, padding of 0xFF shall be added to the end of the transmission block.
The preambles are not randomized, the randomizer sequences is applied only to information bits. A pseudo random binary sequence (PRBS) generator is shown in Figure 3.2.1.1- 1.
Figure 3.2.1.1- 1 PRBS for data randomization
3.2.1.2 FEC
An FEC, consisting of the concatenation of a Reed-Solomon (RS) outer code and a rate-compatible convolutional inner code, are supported on both uplink and downlink.
Support BTC and CTC is optional.
The BPSK-1/2 should always be used as the coding mode when requesting access to the network and in the FEC burst. The encoding is performed by first passing the data in block format through the RS encoder and then passing it through the zero-terminating convolutional encoder.
3.2.1.3 Interleaving
All encoded data bits are interleaved by a block interleaver with a block size corresponding to the number of coded bits per the allocated subchannels per OFDM symbol. The interlevavr is defined by a two step permutation. The first one ensures that adjacent coded bits are mapped onto nonadjacent subcarriers. The second permutation insures that adjacent coded bits are mapped alternately onto less or more significant bits of the constellation, thus avoiding long runs of lowly reliable bits. Table 3.2.1.3- 1 gives the block size of the bit interleaver.
Table 3.2.1.3- 1 Block sizes of the bit interleaver
3.2.1.4 Data Modulation
After bit interleaving, the data bits are sent serially to the constellation mapper [16]. The contellations of BPSK, Gray-mapped QPSK, 16-QAM and 64-QAM are shown in Figure 3.2.1.4- 1, whereas the support of 64-QAM is optional for license-exempt bands. The constellations must be normalized by multiplying the constellation point with the indicated factor c to achieve equal average power. For each modulation,
b denotes the LSB.
0The constellation-mapped data must be subsequently modulation onto all allocated data subcarriers in order of increase frequency offset index. The first symbol out of the data constellation mapping should be modulated onto the allocated subcarrier with the lowest frequency index.
3.2.2 Preamble structure
All preambles are structured as either one or two OFDM symbol is. The OFDM symbols are defined by the values of composing subcarriers. Each of those OFDM symbols contains a cyclic prefix (CP), which length is the same as the CP for data OFDM symbols. The first preamble consists of two consecutive OFDM symbols. The time domain waveform of the first symbol consists of four repetitions of 64-sample fragment, preceeded by a cyclic prefix (section A). The second OFDM symbol in time domain structure composed of two repetitions of a 128-sample fragment, preceded by a cyclic prefix (section B). The time domain structure are shown in Figure 3.2.2- 1
Figure 3.2.2- 1 Downlink and network entry preamble structure
The frequency domain sequences for all full bandwidth preambles are derived form the sequence:
PALL={ 1-j, 1-j, -1-j, 1+j, 1-j, 1-j, -1+j, 1-j, 1-j, 1-j, 1+j, -1-j, 1+j, 1+j, -1-j, 1+j, -1-j, -1-j, 1-j, -1+j, 1-j, 1-j, -1-j, 1+j, 1-j, 1-j, -1+j, 1-j, 1-j, 1-j, 1+j, -1-j, 1+j, 1+j, -1-j, 1+j, -1-j, -1-j, 1-j, -1+j, 1-j, 1-j, -1-j, 1+j, 1-j, 1-j, -1+j, 1-j, 1-j, 1-j, 1+j, -1-j, 1+j, 1+j, -1-j, 1+j, -1-j, -1-j, 1-j, -1+j, 1+j, 1+j, 1-j, -1+j, 1+j, 1+j, -1-j, 1+j, 1+j, 1+j, -1+j, 1-j, -1+j, -1+j, 1-j, -1+j, 1-j, 1-j, 1+j, -1-j, -1-j, -1-j, -1+j, 1-j, -1-j, -1-j, 1+j, -1-j, -1-j, -1-j, 1-j, -1+j, 1-j, 1-j, -1+j, 1-j, -1+j, -1+j, -1-j, 1+j, 0, -1-j, 1+j, -1+j, -1+j, -1-j, 1+j, 1+j, 1+j, -1-j, 1+j, 1-j, 1-j, 1-j, -1+j, -1+j, -1+j, -1+j, 1-j, -1-j, -1-j, -1+j, 1-j, 1+j, 1+j, -1+j, 1-j, 1-j, 1-j, -1+j, 1-j, -1-j, -1-j, -1-j, 1+j, 1+j, 1+j, 1+j, -1-j, -1+j, -1+j, 1+j, -1-j, 1-j, 1-j, 1+j, -1-j, -1-j, -1-j, 1+j, -1-j, -1+j, -1+j, -1+j, 1-j, 1-j, 1-j, 1-j, -1+j, 1+j, 1+j, -1-j, 1+j, -1+j, -1+j, -1-j, 1+j, 1+j, 1+j, -1-j, 1+j, 1-j, 1-j, 1-j, -1+j, -1+j, -1+j, -1+j, 1-j, 1-j, -1-j, 1-j, -1+j, -1-j, -1-j, 1-j, -1+j, -1+j, -1+j, 1-j, -1+j, 1+j, 1+j, 1+j, -1-j, -1-j, -1-j, -1-j, 1+j, 1-j, 1-j }
The frequency domain sequences for the 4×64 sequence,
P
4×64, is defined by Equation (3.2.2-1), and 2×128 sequences,P
EVEN, is defined by Equation (3.2.2-2), where factor 2 equates the Root-Mean-Square power with that of the data section.And the additional factor of 2 is related to the 3-dB boost.
In licensed bands, the duplexing method must be either TDD or FDD. In license-exempt bands, the duplexing method should be TDD. The OFDM PHY supports a frame-based transmission. A frame consists of a downlink subframe and an uplink subframe. A downlink subframe consists of only one downlink PHY PDU. An uplink subframe consists of contention intervals scheduled for initial ranging and bandwidth request purposes and one or more multiple uplink PHY PDUs, each transmitted from a different subscriber station.
A downlink PHY PDU starts with a long preamble, which is used for PHY synchronization. The preamble is followed by a FCH burst. The FCH burst is one OFDM symbol long and is transmitted using BPSK rate 1/2. The FCH contains downlink frame prefix to specify burst profile and length of one or several downlink bursts immediately following the FCH. The FCH is followed by one or more multiple downlink bursts, each transmitted with different burst profile. Each downlink burst consists of an integer number of OFDM symbols.
With the OFDM PHY, a PHY burst, either a downlink PHY burst or an uplink PHY burst, consists of an integer number of OFDM symbols, carrying MAC messages.
To from an integer number of OFDM symbols, unused bytes in the burst payload maybe padded by the bytes 0xFF.
In each time division duplex or duplexing (TDD) frame, see Figure 3.2.3- 1, the TTG and RTG must be inserted between the downlink and uplink subframe and at the end of each frame, respectively, to allow the base station (BS) to turn around. The downlink and uplink frame structure of FDD of an OFDM system are shown in Figure 3.2.3- 2 .
Figure 3.2.3- 1 Example of OFDM frame structure with TDD
Figure 3.2.3- 2 Example of OFDM frame structure with FDD
Chapter 4 Frame Synchronization Techniques
FDM is a kind of multi-carrier modulation on multi-channel. By adding guard interval (GI) between OFDM symbols, OFDM WLAN system can be against multipath delay spread. Signals can be detected up on the reception of just one training sequence of two OFDM symbols. Finding the symbol timing for OFDM means finding an estimate of where the symbol starts.
There have been several methods of synchronization for OFDM in recent years [4]-[6], [17] and [22]. Some new schemes has been proposed in [18]-[21], they use the convolution characteristic of cyclic prefix to overcome the defects of the fluctuation of the estimated start position.
In this chapter, we will introduce Schmidl & Cox algorithm [10] and Feng Lu, et al.’s adaptive symbol timing (AST) algorithm [11], which try to overcome the fault of
O
Finally, we will introduce the pseudo-multipath iteration algorithm (PMIA) [12], which provide a new concept of pseudo-multipath.
4.1 System Architecture
Figure 4.1- 1 is the system architecture of OFDM system. It shows the action between Transceiver (Tx) and receiver (Rx). Transceiver deliver signals to receivers through radio channel. As for the channel model, we use SUI channel model, see section 4.1.1.
Signals from transceivers may distort because of channel fading or delay spread. We must create some mechanism to recover the correct timing so as to let receivers receive the correct signals form transceivers.
Because of not knowing the correct timing for signals, we must detect the exact timing of signals of the receivers for synchronization. If the timing of these signals can not synchronized, the signal may either produce phase rotation or cause errors on the received signals. Therefore, synchronization at the receiver is one important step that must be performed.
4.1.1 SUI Channel Model
As for the channel model, we would like to construct SUI channel, taking SUI parameters in reference document 802.16.3c01-29r4 [7]. Because the system is designed for NLOS environment, we use Jakes model accompanied with SUI channel parameters to build this fading channel [8], [9].
Figure 4.1- 1 802.16-2004 System Structure
The SUI model defines three types of terrains, given below:
Category A: maximum path loss, hilly terrain, moderate-to-heavy tree density.
Category B: Intermediate path loss, flat terrain, light tree density.
Category C: minimum path loss, flat terrain, light tree density
These categories are shown in Table 4.1.1- 1 and Table 4.1.1- 2. Figure 4.1.1- 1 shows examples for each taps magnitude. We list all of the SUI channel parameters in Table 4.1.1- 3 to Table 4.1.1- 8 that will be used in our simulations below.
Table 4.1.1- 1 SUI Channel model (1)
Terrain Type SUI Channel
A SUI-5, SUI-6
B SUI-3, SUI-4
C SUI-1, SUI-2
Table 4.1.1- 2 SUI Channel model (2)
Doppler Low delay spread Moderate delay spread High delay spread
Low
SUI-3 - SUI-5
High
- SUI-4 SUI-6
Figure 4.1.1- 1 An example of tap fading of SUI-3 Channel
Table 4.1.1- 3 The parameters of SUI-1 Channel
SUI-1 Channel (Terrain Type: C)
Tap 1 Tap 2 Tap 3 Units
Delay
0 0.4 0.9
μ secPower
0 -15 -20 dB
Doppler
0.4 0.4 0.4
HzNormalization Factor
F
ommi= -0.1771 dBTable 4.1.1- 4 The parameters of SUI-2 Channel
SUI-2 Channel (Terrain Type: C)
Tap 1 Tap 2 Tap 3 Units
Delay
0 0.4 1.1
μ secPower
0 -12 -15 dB
Doppler
0.2 0.2 0.2
HzNormalization Factor
F
ommi= -0.3930 dBTable 4.1.1- 5 The parameters of SUI-3 Channel
SUI-3 Channel (Terrain Type: B)
Tap 1 Tap 2 Tap 3 Units
Delay
0 0.4 0.9
μ secPower
0 -5 -10
dBDoppler
0.4 0.4 0.4
HzNormalization Factor
F
ommi= -1.5113 dBTable 4.1.1- 6 The parameters of SUI-4 Channel
SUI-4 Channel (Terrain Type: B)
Tap 1 Tap 2 Tap 3 Units
Delay
0 1.5 4
μ secPower
0 -4 -8 dB
Doppler
0.2 0.15 0.25
HzNormalization Factor
F
ommi= -1.9218 dBTable 4.1.1- 7 The parameters of SUI-5 Channel
SUI-5 Channel (Terrain Type: B)
Tap 1 Tap 2 Tap 3 Units
Delay
0 4 10
μ secPower
0 -5 -10
dBDoppler
2 2 2.5
HzNormalization Factor
F
ommi= -1.5513 dBTable 4.1.1- 8 The parameters of SUI-6 Channel
SUI-6 Channel (Terrain Type: B)
Tap 1 Tap 2 Tap 3 Units
Delay
0 14 20
μ secPower
0 -10 -14 dB
Doppler
0.4 0.4 0.4
HzNormalization Factor
F
ommi= -0.5683 dBIn accordance with IEEE Std. 802.16e in reference [3], we modify the SUI channel model. We only adjust the parameters with Doppler frequency [13]. In 2.5 Ghz licensed band, 7 Mhz bandwidth and 120 km/hr to calculate the Doppler frequency, and then modify the tap magnitude of each SUI channel parameters. We can see the detail equations and its computing procedure in the section 5.2.
4.1.2 Simulation Parameters
Table 4.1.2- 1 lists the parameters and its value that we used in our fixed and mobile simulations [12].
Table 4.1.2- 1 System Simulation Parameters
Parameter Value
Number of subcarrier (N)
256 (include 200 used subcarriers)
Pilot Number (P)
8
Bandwidth (BW)
7 Mhz / (10 Mhz for licensed-exempt band)
Carrier spacing (Δ
f
)31.25 Khz
Sampling Rate (
f
s)8 Mhz / (11 Mhz for licensed-exempt band)
Symbol Rate
25 Khz
Mapping Modulation
16-QAM
Useful Time
32μsec (256 samples)
Cyclic Prefix
1/4
OFDM symbol time
40μsec (320 samples)
Frequency offset (Normalized)
0.25
4.2 Some Frame Synchronization Algorithms
Symbol timing error will affect the amplitude of the received signal and carrier phase. It also introduces ISI. In order to correctly demodulate the signal, we must find the start point of OFDM symbol before FFT demodulation.
4.2.1 Schmidl & Cox Algorithm
In 1997, Schmidl & Cox [10] introduced a technique employing the preamble to estimate the start point of the received OFDM signal. However, the outcome produces a plateau in metric function. This plateau leads to some uncertainty in the estimated start position. We list the timing metric functions of Schmidl & Cox method for frame timing estimation below:
Figure 4.2.1- 1 shows the block diagram of Schmidl & Cox algorithm.
Schmidl & Cox algorithm has a plateau region, the result of single frame
simulation as shown in Figure 4.2.1- 2. We still can find a maximum value in this region. The position of maximum value is used as the start point of the OFDM symbol.Figure 4.2.1- 1 Block diagram of Schmidl & Cox algorithm
4.2.2. Adaptive Symbol Timing Algorithm (ASTA)
Recently, Feng Lu et al. introduced a new method called adaptive symbol timing (AST) algorithm [11], which used the autocorrelation of the received signal and its delayed version to get frame timing synchronization. The AST algorithm needs a preamble whose length is equal to two OFDM symbols (a section A preamble and a section B preamble). Based on Schmidl & Cox algorithm, by fixing the length of cyclic prefix and the length of correlation window, we can find the first peak at the end of the section A preamble. Then, we can find another peak at the end of the section B preamble. Then, we can find the only peak by the mean value of the metric function
M(k) and the metric function M
(k
−320)and divided by two, see Equation (4.2.2-4).All timing metric functions are shown in Equations (4.2.2-1), (4.2.2-2), (4.2.2-3) and (4.2.2-4). Finally, we can get the peak value. The position of the peak value is defined as the start point of the symbol.
Note that, N is the length of correlation window. Figure 4.2.2- 1 shows the block diagram of AST algorithm.
After ASTA operation, we will not find the plateau as that in using Schmidl & Cox algorithm anymore and get a unique peak in curve of metric function MM(k) [18]. In this way, we have a much better precision in the estimate of start position. The results of single frame simulation of the metric functions MM(k) and M(k) are shown in Figure 4.2.2- 2.
Figure 4.2.2- 1 Block diagram of AST algorithm
Figure 4.2.2- 2 AST Algorithm
In a wireless environment, signals may have fading and multipath delay spread when passing the channels. Figure 4.2.2- 3 is the probability distribution of estimated start positions of 20000 frames. However, the more precise ratio in the estimation by using AST algorithm would just be 61.8% at time error index 0 [12], as shown in Figure 4.2.2- 3.
Figure 4.2.2- 3 The estimate probability distribution in SUI-3 of AST Algorithm
4.2.3 Pseudo-Multipath Iteration Algorithm (PMIA)
In 2005, C. C. Wu [12] in his thesis introduced a concept of pseudo multipath. The scheme is based on the AST algorithm. First, the algorithm defined a pseudo path with preamble which is two-OFDM symbol long (a section A preamble and a section B preamble). We can get a unique peak at the end of the preamble. Then, we exploit the peak index for the start position of the pseudo path that we want to insert in. We would use the original signal added with a preamble multiplied a factor α, and moves forward
same way, we must iterate several times to get the final outcome of the PMIA algorithm. After several times of iteration, the peak’s position of the last iteration will then become our estimated start position. The PMIA algorithm can further increase the precision of the estimation of start position.
In PMIA, three variables must be set: 1) the strength of pseudo path, 2) the timing of pseudo path inserted and 3) times of iteration. Because there are three variables that should be optimized, it is difficult to get the trade off among these variables. The PMIA needs a lot of computation complexity for the iterative operation, and need more hardware for the computation.
Although PMIA has some disadvantages but it still has advantages. The PMIA is not only more precise than AST algorithm but also able to fine-tune the system performance when we select a smaller value of α. The timing of inserting the pseudo path in PMIA is shown in Figure 4.2.3- 1.
From the computer simulation results within the thesis [12], we can see that the optimized value of these parameters in fixed SUI-3 environment is: α=0.1, λ=6 and 4 times of iteration. We will show our computer simulation results using the same optimized value [12] as listed above and compare the system performance with PMIA and PMA in fixed SUI-3 environment as shown in the section 5.1.1.
Figure 4.2.3- 1 The timing of pseudo path inserted for AST algorithm
Chapter 5 Modified Pseudo-Multipath
Algorithm
n this chapter, we will present a new method by putting a section B preamble in the receiver when we use the pseudo-path for the frame timing synchronization. This modified pseudo multipath algorithm (PMA) can increase the accuracy of frame timing estimation without adding more hardware. We apply this method in TDD, downlink OFDM system which fits IEEE 802.16-2004/e Std. environment [2], [3].
5.1 Modified Pseudo-Multipath Algorithm (PMA)
To increase the precision of estimation of the start position, we apply the concept of pseudo-multipath [12]. The pseudo multipath algorithm uses a section B preamble in
To increase the precision of estimation of the start position, we apply the concept of pseudo-multipath [12]. The pseudo multipath algorithm uses a section B preamble in