• 沒有找到結果。

CHAPTER 1. INTRODUCTION

1.3 O RGANIZATION OF T HIS T HESIS

FEC QAM

de-mapping P/S DFT

Guard-interval remove

A/D RX

front-end

P/S

P/S

Channel Noise

Interferences Frequency domain Time domain

Figure 1.5 Bock diagram of a simple OFDM system

1.3 Organization of This Thesis

This thesis is organized as follows. In Chapter 2, the simulation platform and detail specifications of the IEEE 802.11a WLAN, single-band LDPC-COFDM UWB system [17]

and the multi-band OFDM-based Ultra-Wideband [8] system will be introduced. Algorithms of the proposed frequency synchronizer for different requirements will be described in Chapter 3 and Chapter 4 respectively. The simulation result and performance analysis will be discussed in Chapter 5. Chapter 6 will introduce the design methodology, hardware architecture, and the chip summary of the proposed design. Conclusion and future work will be given in Chapter 7.

Chapter 2.

System Platform

In this chapter, we introduce the three system platforms for design analysis and performance simulation. The first one is IEEE 802.11a physical layer (PHY) [3]; the second one is single-band LDPC-COFDM Ultra-Wideband (UWB) system [17]; the other one is IEEE 802.15.3a UWB with multi-band OFDM modulation proposed by Texas Instrument (TI) [4]. The detail block diagram, system specification and preamble format will be described as follows.

2.1 Introduction to IEEE 802.11a System

2.1.1 IEEE 802.11a basic

IEEE 802.11a is an OFDM-based indoor WLAN system. The block diagram of the baseband transceiver can be illustrated in Figure 2.1. The system platform includes a COFDM modem and an indoor radio channel model. The COFDM modem comprises a 64-point DFT-based QAM-OFDM modem and a forward-error correction (FEC) coding. The 64 subcarriers contain 48 data carriers and 4 pilot carriers, the others 16 carriers called as null band are set to zero. The OFDM symbol time TS is 3.2µs, the bandwidth of the subcarriers is 1/TS = 312.5 KHz and total bandwidth is N/TS = 20 MHz. The indoor radio channel model comprises a Rayleigh fading channel and AWGN. The supported data rate is from 6Mbits/s to 54 Mbits/s with coding rate equals 1/2, 2/3 and 3/4. The system parameters can be listed in Table 2.1.

Data out

Detection Guard-interval FFTFFT

reduction

de-mapping Viterbi De-scramblerDe-scrambler

decoder

encoder InterleaverInterleaver Constellation mapping

Figure 2.1 System platform of IEEE 802.11a PHY

Table 2.1 System parameters of IEEE 802.11a PHY

Constellation mapping method BPSK, QPSK, 16QAM, 64QAM Date rate (Mbits/s) 6, 9, 12, 18, 24, 32, 48, 54 FEC coding rate (R) 1/2, 2/3, 3/4

FFT size (N) 64

Number of used subcarriers (NST) 52 Number of data carriers (NSP) 48 Number of pilot carriers (NSD) 4

Bandwidth (MHz) 20

Subcarrier bandwidth (KHz) 312.5 (20 MHz/64) IFFT/FFT period (TFFT) 3.2us

GI duration (TGI) 0.8us (TFFT/4) PLCP preamble duration 16us (TSHORT + TLONG)

The IEEE 802.11a provides 8 kinds of data rates up to 54 Mb/s by using QAM modulation and convolutional code. The system adopts packet transmission and the PHY protocol data unit (PPDU) frame of IEEE 802.11a shown in Figure 2.2. The PPDU frame format includes the physical layer convergence procedure (PLCP) Preamble, Header and Data fields. The PLCP preamble is a training sequence which is used to perform the synchronization. And the SIGNAL field is always the BPSK modulation and 1/2 coding rate FEC coding. Following the SIGNAL field is the Data field which is used to transmit the general information.

PLCP Preamble 12 Symbols

SIGNAL 1 OFDM Symbol

DATA

Variable Number of OFDM Symbols For Synchronization BPSK, 1/2 coding rate Normal data transmission

Figure 2.2 PPDU frame format

2.1.2 PLCP preamble

Figure 2.3(a) shows the structure of the IEEE 802.11a PLCP preamble. The PLCP preamble is composed of 10 identical short symbols and 2 identical long symbols. The short symbol occupies 0.8µsec while the long symbol occupies 3.2µsec. The total length of the PLCP preamble is 320 samples, each short training symbol contains 16 samples and each long training symbol contains 64 samples. The short symbols serve to do frame detection, automatic gain control, and coarse timing and frequency offset estimation. The two long symbols can be used to do channel estimation and fine CFO estimation. The generation pattern of the short preamble is shown in the left of Figure 2.3(b). These is only one data every four subcarriers. The data carrier spacing is four times larger than the carrier spacing of normal OFDM symbols. The second partition of the PLCP is the long training sequence. It is generated by the pattern in the left of Figure 2.3(c). The right is the sequence of a long

training symbol. The long training sequence contains two repeat OFDM symbols and a guard interval. The GI is two times longer than normal OFDM symbols’.

t1 t2 t3 t4 t5 t6 t7 t8 t9 t10 GI2 T1 T2 Short preamble 10*0.8 = 8us Long preamble 2*0.8 + 2*3.2 = 8us

16us

Signal Detect, AGC, Diversity Selection

Coarse CFO Estimation, Timing Sync.

Channel and Fine CFO Estimation

(a) PLCP preamble format

0 10 20 30 40 50 60 70

Frequency domain carriers Time domain sequence 0

(b) Short preamble

0 10 20 30 40 50 60 70

(c) Long preamble

Figure 2.3 PLCP preamble in IEEE 802.11a

2.1.3 Transmit Center Frequency Tolerance

In the specification of IEEE 802.11a, the transmitter frequency offset is asked to be small than ±20ppm. If the receiver can achieve the same requirement, the relative frequency between the transmitter and the receiver shall be ±40ppm. The working frequency is about 5.3GHz, the frequency tolerance is ±40ppm. The maximum frequency offset is ±212KHz.

2.2 Introduction to Ultra WideBand System

2.2.1 IEEE 802.15.3a UWB basic

OFDM-based wireless ultra-wideband (UWB) technology has received attention from both the academia and the industry. The main reason for the increased attention is the Federal Communications Commission (FCC) allocated 7,500MHz of spectrum (from 3.1GHz to 10.6GHz) for use by UWB devices. It helped to create new standardization, like IEEE 802.15.3a which focuses on developing high-speed wireless communication systems for personal area network. Another reason is because this technology promises to deliver data rates up to 480Mb/s at a distance of 2 meters in realistic multi-path environments.

We have established two kinds of UWB systems in platform; one is single-band LDPC-COFDM UWB system [17] which is used low-density parity check (LDPC) FEC codec, the other one is multi-band OFDM-based UWB system [4] which is used convolutional encoder and Viterbi decoder. The block diagram of the UWB PHY is shown in Figure 2.4 which is similar to the IEEE 802.11a WLAN system. The key differences between these two systems can be listed as follow,

OFDM symbols are interleaved across both frequency and time. An example of the time-frequency interleaving (TFI) can be shown in Figure 2.5.

Scrambler

QPSKQPSK SpreadingSpreading IFFTIFFT

Clipping

Clipping ShapingShaping Preamble insertion

RF effects MultipathMultipath

Timing

Figure 2.4 System platform of UWB PHY

The supported data rate is up to 480Mbits/s, which is almost ten times of the data rate in IEEE 802.11a systems. A 128-point FFT is applied and only PSK (BPSK, QPSK) is used in the UWB system.

The bandwidth is up to 528MHz, which is 26.4 times wider than IEEE 802.11a.

Channel 1

9.5 ns Guard Interval for TX/RX Switching Time

60.6 ns Cyclic Prefix

Period = 937.5 ns 312.5 ns

In order to achieve better system performance and higher decoding speed, (600,450) LDPC code is exploited as the kernel of error correcting mechanism in our simulation platform. The parallelism of LDPC decoding makes it easy to decode 480Mb/s data stream or even higher to

multiple Gb/s. Because of fixed 3/4 FEC coding rate with different spreading gain, we have 120Mb/s, 240Mb/s and 480Mb/s three data-rates. The coding performance can be near Shannon limit when using iterative decoding algorithm. We summarize system parameters about UWB platform that we used in Table 2.2.

Table 2.2 System parameters of the LDPC-COFDM UWB PHY

Date rate (Mb/s) 120 240 480

Constellation QPSK QPSK QPSK

FFT size 128 128 128

Coding rate 3/4 3/4 3/4

Spreading gain 4 2 1

Data carrier per OFDM symbol 100 100 100

Baseband bandwidth (MHz) 528 528 528

OFDM symbol duration (ns) 312.5 312.5 312.5

The detail specifications of the multi-band OFDM UWB PHY can be listed in Table 2.3.

Because of different FEC coding rate with different spreading gain, we have 8 kinds of data-rates. Figure 2.6 shows the format for the PLCP frame including the PLCP preamble, PLCP header and data-field. The PLCP preamble shall be added prior to the PCLP header to aid receiver algorithms related to synchronization, carrier-offset recovery, and channel estimation. The PLCP header is always sent at an information data rate of 53.3 Mb/s. The remainder of the PLCP frame is sent at the desired information data rate of 53.3, 80, 110, 160, 200, 320, 400 or 480 Mb/s.

Table 2.3 System parameters of the MB-OFDM UWB PHY Constellation mapping method BPSK, QPSK

Date rate (Mbits/s) 55, 80, 110, 160, 200, 320, 480 FEC coding rate (R) 1/2, 3/4, 5/8, 11/32

FFT size (N) 128

Bandwidth (MHz) 528

Subcarrier bandwidth (MHz) 4.125 (528 MHz/128)

Data bytes per packet 1024

IFFT/FFT period (TFFT) 242.42ns Cyclic prefix duration (TCP) 60.61ns (TFFT/4) Guard interval duration (TGI) 9.47ns

PLCP preamble duration 9.375us (TSHORT + TLONG)

PLCP Preamble PLCP Header 53.3 Mb/s

Data field

53.3, 80, 110, 160, 200, 320, 400 , 480 Mb/s Synchronization

Carrier-offset recovery

Channel estimation TX spec. record Normal data transmission

Figure 2.6 PLCP frame format

2.2.2 PLCP preamble

The signal format of the UWB PLCP preamble is shown in Figure 2.7 and it is different from IEEE 802.11a because of the time interleaving. The preamble contains 30 OFDM symbols, 21 are packet synchronization sequence, 3 are frame synchronization sequences and 6 are channel estimation sequences. The packet synchronization portion of the preamble can be used for packet detection and acquisition and coarse carrier frequency estimation. The

frame synchronization portion of the preamble can be used to synchronize the receiver algorithm within the preamble, and it also provides one sequence period per band with an inverted polarity with respect to the packet synchronization portion of the preamble. Finally, the channel estimation portion of the preamble, denoted as {CE0, CE1, … CE5}, shall be constructed by successively appending 6 periods of an OFDM training sequence.

Each OFDM symbol of the UWB contains 32-point pre-GI, 128-point FFT symbol and 5-point post-GI. In IEEE 802.11a system, GI is the cyclic prefix of each OFDM symbols, which is used for the concern of multipath spreading. In UWB system, cyclic prefix (CP) is for multipath concern and the GI is particularly referred to the time between band switching.

PS20 21 OFDM symbol*312.5ns

9.375us

PS0 PS1 FS0 FS1 FS2 CE0 CE1 CE5

3 OFDM symbol*312.5ns 6 OFDM symbol*312.5ns

Packet Sync. Sequence Frame Sync. Sequence Channel Estimation Sequence

• • • • • • • • •

128-point FFT Symbol 32-point

Pre-GI

5-point Post-GI 242.42ns

60.61ns 9.47ns

312.5ns

(a) OFDM symbol format

(b) Training structure Figure 2.7 UWB PLCP preamble

2.2.3 Transmit Center Frequency Tolerance

The MB-OFDM UWB PHY operates in the 3.1~ 10.6 GHz frequency and the relationship between center frequency and band number are given by the following equation:

Band center frequency = 2904 + 528 * nb, nb = 1…14 (MHz). This definition provides a unique numbering system for all channels that have a spacing of 528 MHz and lie within the band 3.1~10.6 GHz. Based on this, five band groups are defined, consisting of four groups of three bands each and one group of two bands. Band group 1 is used for Mode 1 devices

(mandatory mode). The remaining band groups are reserved for future use. The frequency of operation form Mode 1 devices is shown in Figure 2.8, and the band allocation is summarized in Table 2.4. In the UWB system, maximum ±20ppm CFO is expected to exist in both transmitter and receiver. So the requested CFO estimation range should be within ±40ppm (TX+RX) of 3.1~10.6 GHz RF frequency. That will be equal ±124KHz~±424KHz.

3432

f

MHz

3960 MHz

4488 MHz Band #1 Band #2 Band #3

Figure 2.8 Operation frequency for Mode 1 device

Table 2.4 OFDM UWB PHY band allocation Band Group BAND_ID Lower frequency

(MHz)

Center frequency (MHz)

Upper frequency (MHz)

1 3168 3432 3696 2 3696 3960 4224 1

3 4224 4488 4752 4 4752 5016 5280 5 5280 5544 5808 2

6 5808 6072 6336 7 6336 6600 6864 8 6864 7128 7392 3

9 7392 7656 7920

10 7920 8184 8448

11 8448 8712 8976

4

12 8976 9240 9504

13 9504 9768 10032

5 14 10032 10296 10560

2.3 The Indoor Wireless Channel Model

In order to simulate the data transmission in the real environment, an indoor wireless channel model is established, which includes a time-variant multipath fading [18]-[19], CFO, SCO, and AWGN. The detailed are introduced individually below

2.3.1 Multipath Fading Channel Model

In wireless communication transmission systems, transmitted signal arrives at receiver through several paths with different time delay and power decay, which is called multipath interference. Figure 2.9 shows the concept of the indoor multipath interference and its channel impulse response and frequency response. The received signal can be modeled as

+

=

N

N

N x t

t x t

y( ) ( ) β ( τ ) (2-1)

Transmitter

Receiver Multipaths

time

Multipaths

pulse

time freq

Figure 2.9 Concept of indoor multipath channel

Due to this effect, the Inter-Symbol Interference (ISI) and frequency-selective fading occur when the maximum delay spread is larger than the symbol period or the channel coherent bandwidth is smaller than the data bandwidth. Although the multipath scales tones, it also remains the orthogonal property because of cyclic prefix technique of each OFDM symbol.

The applied multipath fading channel is established according to the IEEE specification. It consists of 13 independent taps, which has Rayleigh distributed magnitude, exponentially decayed power and random uniformly distributed phase [18]. Figure 2.10shows the effect of ISI and frequency-selective fading, and the channel impulse response (CIR) and the channel frequency response (CFR) with RMS delay equals 50ns are shown in Figure 2.11.

Transmitted data

Figure 2.10 (a) ISI effect (b) frequency-selective fading channel

(a) (b)

RMS Delay Spread [100ns] Subcarrier Index [312.5KHz]

CIR Magnitude CFR Magnitude

Figure 2.11 (a) CIR (b) CFR example of the multipath fading channel

2.3.2 Carrier Frequency Offset Model

The sensitivity to carrier frequency offset (CFO) is one of the main drawbacks of the OFDM system. One of the reasons causing the CFO in wireless communication is the RF circuit mismatch between the transmitter and the receiver. When the transmitter carriers the data x with a frequency t f1 (2.2) and the receiver gets the data with another frequency f2 (2.3), the received data y contains the original data and a sinusoidal signal with a frequency t

2

Another reason caused the CFO is the Doppler Effect. From the Doppler equation (2.4), the received signal will not equal to the transmitted one if the relative speed between the transmitter and the receiver is not zero. There will be some frequency offset f existed. δ Because of the v is the velocity of light, the f is usually negligible as compared with the δ CFO caused by the circuit mismatch.

t

No matter the causes of the CFO, the behavior of the CFO in the spectrum domain is shown in the Figure 2.12. The total frequency offset is f1-f2+fδ and it will be expressed as f in the thesis. Besides, the orthogonal property between each subcarrier is based on the perfect sampling on some specific frequencies in the spectrum domain. When we transmit the data without CFO, the received data will be recovered perfectly because of the influences from others subcarriers’ are all zero, which shown in Figure 2.14(a). The equation (2.6) and (2.7) are the N points IDFT and DFT equation. The x and n X in the (2.6) are the data of time k domain and frequency domain, respectively.

{ }

1 1 01 2 1

Figure 2.12 The behavior of the CFO in spectrum domain

Applying equation (2.3) to discrete time, if there is no CFO within transmission, the received data y is equal to the transmitted datan x . The received data n Y in the frequency k domain are the same with the transmitted dataX . k

∑ ∑

However, the data can not be received perfectly in the CFO environment. In the time domain, the received data suffering from CFO is shown as equation (2.9), where Ts is the sampling time and the 1/NTs is the subcarrier bandwidth of an OFDM symbol. We normalize the frequency offset f to the subcarrier bandwidth and ∈ is the relative frequency offset.

N

From the publish of Moose [14], the CFO caused linear phase shift in time domain will convert to ICI in the frequency domain after passing the DFT, which shown in Figure 2.13 and Figure 2.14(b) respectively, and the relative equation is shown in (2.10).

[ ] [ ]

Figure 2.13 Linear phase shift

Ideal With CFO

frequency frequency

(a) (b)

Perfect sampling point imperfect sampling point

Figure 2.14 The received data (a) without CFO (b) with CFO (ICI)

In order to simulate the real CFO environment, we find a phase noise model such as Figure 2.15. According to this phase noise model, we built it in our platform. Figure 2.16 shows the example of phase noise which normal distributed with 40ppm (400 KHz) mean-CFO. In general, the mean CFO is time-invariant. However, in order to consider about more complete simulation condition, we define the mean CFO could be time-variant. We defined three kinds of CFO environments, the first one is time-invariant (TIV), and the second one is slow-variant (SV) which changed 80ppm within 50ms, and the final one is fast-variant (FV) which changed 80ppm within 5ms. Figure 2.17 shows the three CFO definitions of TIV, SV and FV.

These definitions are helpful for explanation of following design.

-100

1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05

Frequency(Hz)

PSD (dBc/Hz)

Figure 2.15 Phase noise model

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Figure 2.16 Example of phase noise

-40ppm

1 s

Figure 2.17 Mean CFO definition

2.3.3 Sampling Clock Offset Model

Sampling Clock Offset (SCO) is the sampling clock rate mismatch between the digital to analog converter (DAC) in transmitter and the analog to digital converter (ADC) in receiver.

In the platform, the model of clock offset is built using the concept of interpolation. The input digital signals and the shifted sinc wave can interpolate the value between two sampling points. The received signal after ADC can be derived as equation (2.11).

) SCO, even if the initial sampling point is optimized, the following sampling points will slowly shift with time. This shift in time-domain becomes a phase rotation in frequency-domain. Figure 2.18 shows the time-domain oversampled received data and the frequency-domain linear phase shift caused by SCO.

(a) (b) (b) the frequency-domain linear phase shift

2.3.4 AWGN Model

The AWGN channel model is established by the random generator in Matlab. The output random signal is normally distributed with zero mean and variance equal to 1. The complex AWGN noise can be modeled as

2 1 10

1 20

SNR PS

l randn j

l randn t

w

⋅ +

=[ ( , ) ( , )]

)

( (2.12)

Where PS is the data signal power, SNR is the signal to noise power ratio, and l is the data signal length.

Chapter 3.

A Frequency Synchronizer Design for OFDM WLAN Systems

As for the complement to periodic and non-constant-power preamble in IEEE 802.11a, low-complexity frequency estimators are of interest. Such estimator is relied on the phase information of auto correlations. In this chapter, we chose fewer samples for auto correlation based on the average power of preamble before FFT process. We also derive the performance and show how the correlation samples should be properly chosen with acceptable SNR loss.

3.1 Carrier Frequency Offset Synchronization

For packet-based OFDM wireless systems, the burst synchronization is needed. We generally used data-aided methods which inserted the special synchronization information to estimate the CFO. The data-aided CFO estimation can achieve the system requirement in a very short time period.

3.1.1 The algorithms of CFO estimation and compensation

In the 1994, Paul H. Moose [14] proposed a method to estimate frequency offset from the demodulated data signals in the receiver. This method involves repetition of a data symbol and compares the phases of each of the carriers between the successive symbols. Since the modulation phase values are not changed, the phase shift of each of the carriers between successive repeated symbols is due to the frequency offset.

received training symbol r that interfered by CFO n ∈ is shown in (3.11), and –K…K are

The kth elements of the N point DFT of the first and second N points of (3.11) are shown in (3.12) and (3.13) respectively.

1

Including the AWGN as below,

1

Both the ICI and the signal of the first and second observations are altered in exactly the same way, by a phase shift proportional to frequency offset. Therefore, the offset ∈ will be estimated using observations (3.15) is shown in (3.16).

⎭⎬

This algorithm, however, can only correctly distinguish the phase rotation in the range [-π, π], the estimation limitation is shown in equation (3.17). In (3.17) and (3.18), the NTs

means the time interval between the two repeat symbols. Minimizing the interval time to the OFDM symbol time can make the maximum CFO estimation range to be half of the subchannel bandwidth. According to the (3.18), the method to get a larger CFO estimation range is to shorten the training symbol time. The idea is also suggested in the [14].

π π

π∈= 2 NTsf

2 (3.17)

NTs

f 2

1

(3.18)

After finishing the acquisition of CFO, we counteract the frequency offset that we estimated to compensate the following complex data signal as equation (3.19).

After finishing the acquisition of CFO, we counteract the frequency offset that we estimated to compensate the following complex data signal as equation (3.19).

相關文件