Introduction to OFDM. Characteristics of OFDM (Orthogonal Frequency Division Multiplexing)

Similar documents
ECE5984 Orthogonal Frequency Division Multiplexing and Related Technologies Fall Mohamed Essam Khedr. Channel Estimation

Multi-carrier Modulation and OFDM

Receiver Designs for the Radio Channel

3.6 Intersymbol interference. 1 Your site here

Lecture 13. Introduction to OFDM

Multipath can be described in two domains: time and frequency

Generalized Frequency Division Multiplexing: Analysis of an Alternative Multi-Carrier Technique for Next Generation Cellular Systems

Generalized Frequency Division Multiplexing: Analysis of An Alternative Multi-Carrier Technique for Next Generation Cellular Systems

Orthogonal frequency division multiplexing (OFDM)

OFDM system: Discrete model Spectral efficiency Characteristics. OFDM based multiple access schemes. OFDM sensitivity to synchronization errors

Broadband OFDM-FDMA System for the Uplink of a Wireless LAN

Chapter 5 OFDM. Office Hours: BKD Tuesday 14:00-16:00 Thursday 9:30-11:30

Orthogonal Frequency Division Multiplexing (OFDM)

Practical issue: Group definition. TSTE17 System Design, CDIO. Quadrature Amplitude Modulation (QAM) Components of a digital communication system

Using Delta-Sigma Modulators in Visible Light OFDM Systems

MITIGATING CARRIER FREQUENCY OFFSET USING NULL SUBCARRIERS

Part 3. Multiple Access Methods. p. 1 ELEC6040 Mobile Radio Communications, Dept. of E.E.E., HKU

EC 551 Telecommunication System Engineering. Mohamed Khedr

WLAN a Spec. (Physical Layer) 2005/04/ /4/28. WLAN Group 1

CHAPTER 3 ADAPTIVE MODULATION TECHNIQUE WITH CFO CORRECTION FOR OFDM SYSTEMS

Basic idea: divide spectrum into several 528 MHz bands.

Sinusoidal signal. Arbitrary signal. Periodic rectangular pulse. Sampling function. Sampled sinusoidal signal. Sampled arbitrary signal

Sampling and Multirate Techniques for Complex and Bandpass Signals

Outline. Wireless Networks (PHY): Design for Diversity. Admin. Outline. Page 1. Recap: Impact of Channel on Decisions. [hg(t) + w(t)]g(t)dt.

An Equalization Technique for Orthogonal Frequency-Division Multiplexing Systems in Time-Variant Multipath Channels

Performance Evaluation of OFDM System with Rayleigh, Rician and AWGN Channels

Experimenting with Orthogonal Frequency-Division Multiplexing OFDM Modulation

S PG Course in Radio Communications. Orthogonal Frequency Division Multiplexing Yu, Chia-Hao. Yu, Chia-Hao 7.2.

Effect of Oscillator Phase Noise and Processing Delay in Full-Duplex OFDM Repeaters

BER Analysis for MC-CDMA

Orthogonal Frequency Domain Multiplexing

Cyclostationarity-Based Spectrum Sensing for Wideband Cognitive Radio

Fundamentals of OFDM Communication Technology

Consumers are looking to wireless

The Communications Channel (Ch.11):

( ) D. An information signal x( t) = 5cos( 1000πt) LSSB modulates a carrier with amplitude A c

ELEC 546 Lecture #9. Orthogonal Frequency Division Multiplexing (OFDM): Basic OFDM System

Detection and direction-finding of spread spectrum signals using correlation and narrowband interference rejection

Mobile & Wireless Networking. Lecture 2: Wireless Transmission (2/2)

ESTIMATION OF FREQUENCY SELECTIVITY FOR OFDM BASED NEW GENERATION WIRELESS COMMUNICATION SYSTEMS

Fading & OFDM Implementation Details EECS 562

CARRIER FREQUENCY OFFSET ESTIMATION ALGORITHMS IN ORTHOGONAL FREQUENCY DIVISION MULTIPLEXING SYSTEMS

Wireless Networks (PHY)

Principles and Experiments of Communications

Technical Aspects of LTE Part I: OFDM

Noise. Interference Noise

ICT 5305 Mobile Communications. Lecture - 3 April Dr. Hossen Asiful Mustafa

Analysis of Interference & BER with Simulation Concept for MC-CDMA

Traditional Analog Modulation Techniques

ECEN 5014, Spring 2013 Special Topics: Active Microwave Circuits and MMICs Zoya Popovic, University of Colorado, Boulder

OFDM (Orthogonal Frequency Division Multiplexing) SIMULATION USING MATLAB Neha Pathak MTech Scholar, Shri am Institute of Technology

Self-interference Handling in OFDM Based Wireless Communication Systems

Page 1. Overview : Wireless Networks Lecture 9: OFDM, WiMAX, LTE

Nonlinear FM Waveform Design to Reduction of sidelobe level in Autocorrelation Function

Introduction to OFDM Systems

Underwater communication implementation with OFDM

S.D.M COLLEGE OF ENGINEERING AND TECHNOLOGY

OFDM and FFT. Cairo University Faculty of Engineering Department of Electronics and Electrical Communications Dr. Karim Ossama Abbas Fall 2010

Performance of Orthogonal Frequency Division Multiplexing System Based on Mobile Velocity and Subcarrier

Chapter 7 Multiple Division Techniques for Traffic Channels

Signals and Systems II

Optimizing Reception Performance of new UWB Pulse shape over Multipath Channel using MMSE Adaptive Algorithm

OFDMA and MIMO Notes

Multi-Carrier Systems

Performance Analysis of OFDM for Different Digital Modulation Schemes using Matlab Simulation

Multiple access techniques

Ultra Wide Band Communications

Bit Loading and Peak Average Power Reduction Techniques for Adaptive Orthogonal Frequency Division Multiplexing Systems

OFDM and Downlink Physical Layer Design

ICI Mitigation for Mobile OFDM with Application to DVB-H

IEEE P Wireless Personal Area Networks

Implementation and Comparative analysis of Orthogonal Frequency Division Multiplexing (OFDM) Signaling Rashmi Choudhary

PEAK TO AVERAGE POWER RATIO REDUCTION USING BANDWIDTH EFFICIENCY INCREASING METHOD IN OFDM SYSTEM

Unified Out-of-Band Emission Reduction with Linear Complexity for OFDM

Chapter 7. Multiple Division Techniques

ALi Linear n-stage t ShiftRegister output tsequence

Principles of Baseband Digital Data Transmission

II. OFDM TRANSRECEIVER SYSTEMS

Wireless Information Transmission System Lab. Interference 2006/3/9 王森弘. Institute of Communications Engineering. National Sun Yat-sen University

Wireless Medium Access Control and CDMA-based Communication Lesson 16 Orthogonal Frequency Division Medium Access (OFDM)

Chapter 6: Introduction to Digital Communication

Channel Estimation in Wireless OFDM Systems

Chapter 3. System Theory and Technologies. 3.1 Physical Layer. ... How to transport digital symbols...?

Performance analysis of OFDM with QPSK using AWGN and Rayleigh Fading Channel

Researches in Broadband Single Carrier Multiple Access Techniques

A Physical Sine-to-Square Converter Noise Model

COMPARISON OF CHANNEL ESTIMATION AND EQUALIZATION TECHNIQUES FOR OFDM SYSTEMS

Fund. of Digital Communications Ch. 3: Digital Modulation

Communication Theory

Evaluation of BER and PAPR by using Different Modulation Schemes in OFDM System

Reduction of PAR and out-of-band egress. EIT 140, tom<at>eit.lth.se

Outline / Wireless Networks and Applications Lecture 7: Physical Layer OFDM. Frequency-Selective Radio Channel. How Do We Increase Rates?

ENHANCING BER PERFORMANCE FOR OFDM

6.976 High Speed Communication Circuits and Systems Lecture 16 Noise in Integer-N Frequency Synthesizers

An image rejection re-configurable multi-carrier 3G base-station transmitter

1. Motivation. 2. Periodic non-gaussian noise

SHIV SHAKTI International Journal of in Multidisciplinary and Academic Research (SSIJMAR) Vol. 3, No. 4, August-September (ISSN )

A Unified Perspective of Different Multicarrier CDMA Schemes

Chapter 4 Investigation of OFDM Synchronization Techniques

FPGA implementation of Generalized Frequency Division Multiplexing transmitter using NI LabVIEW and NI PXI platform

Transcription:

Introduction to OFDM Characteristics o OFDM (Orthogonal Frequency Division Multiplexing Parallel data transmission with very long symbol duration - Robust under multi-path channels Transormation o a requency-selective channel into N requency-lat channels Cyclic preix (CP is used to eiciently eliminate inter-symbol intererence (ISI. High spectral eiciency o the orthogonal subcarriers Eicient implementation o the transmitter-receiver pair using Inverse Fast Fourier transorm (IFFT and Fast Fourier transorm (FFT High pea-to-average power ratio (PAPR and sensitivity to carrier requency oset (CFO D/A Conversion - 6 / 54

Introduction to OFDM (cont. General system structure o OFDM transmitter Data Sequence Modulation and Coding S/P Inverse DFT Add Cyclic Preix/ Postix Transmit signal to channel RF Device Pulse Shaping D/A P/S Windowing D/A Conversion - 7 / 54

Review o Non-IDFT-Based Signal Model and its PSD Non-IDFT-based signal model or OFDM system Common continuous-time signal model in many literatures or a single OFDM symbol in complex baseband would be Y( t = = X e jπt 0 t < where X is the data symbol transmitted by the th subcarrier whose carrier requency is. The length o a single OFDM symbol and the subcarrier requencies are related as = which leads to the orthogonality o the signals e observation interval with time duration T 0. T 0 or =,,, in the For simplicity, assume that X s are uncorrelated, proper complex, zero-mean random variables with unit variance., T 0 j π t D/A Conversion - 8 / 54

Review o Non-IDFT-Based Signal Model and its PSD (cont. Non-IDFT-based signal model or OFDM system (cont. For the transmission o more than one OFDM symbol, the signal model is modiied to which X [m] is the th subcarrier data symbol or time-index m, /T b is the OFDM bloc transmission rate and w C (t is the windowing unction. T b is the OFDM bloc transmission time and selected as larger than or equal to T + 0 τ max where Yˆ ( t = τ max = X m= [ m] w C ( t mt is the maximum delay spread o the channel. The shape o the roll-o o the windowing unction w C (t is usually designed to reduce the inter-carrier intererence (ICI under carrier requency oset (CFO. e jπt I no CFO, not only IBI but also ICI can be completely removed. b D/A Conversion - 9 / 54

Review o Non-IDFT-Based Signal Model and its PSD (cont. Example o non-idft based continuous-time signal structure with windowing Yˆ ( t = = X m= [ m] w C ( t mt b e jπt Windowing unction (overlapping with adjacent OFDM bloc Guard interval τ max One bloc period T b One OFDM symbol period T 0 t D/A Conversion - 30 / 54

Review o Non-IDFT-Based Signal Model and its PSD (cont. Power spectral density (PSD o non-idft-based signal model The PSD o this signal model is given by S = Yˆ Yˆ ( WC ( Tb = where W C ( is the continuous-time Fourier transorm (CTFT o the windowing unction w C (t. Note that X [m] s are uncorrelated, proper complex, zero-mean random variables with unit variance. I a wideband analog ilter is additionally used, i some subcarriers are null-out, and i dierent transmit power is allocated or each subcarrier, signal model can be rewritten to a more general orm shown in the next page. D/A Conversion - 3 / 54

D/A Conversion - / 54 3 Review o Non-IDFT-Based Signal Model and its PSD (cont. Power spectral density (PSD o non-idft-based signal model (cont. General orm o non-idft-based signal model - p(t is the complex envelope o the bandpass transmit ilter. - is the complex weighting actor or the th subcarrier. ( is zero. The th subcarrier is called a null subcarrier. - X [m] is the th subcarrier data symbol or time-index m. - w C (t is the continuous-time windowing unction. - /T b is the OFDM bloc transmission rate. The PSD o this signal model is given by = = = t j m b C e mt t w m X t p t Z ( ] [ ( ( π α. ( ( ( = = C b ZZ W T P S α

Review o Non-IDFT-Based Signal Model and its PSD (cont. Power spectral density (PSD o non-idft-based signal model (cont. Parameter α =, = 3 Rectangular windowing (a : (b : T b =T 0 T b =.5T 0 Window duration= T b (a : Fourier transorm o the rectangular window unction is a sinc unction. PSD is almost lat except at the edges. (At the edge, the PSD decays approximately at the rate /. = (b : WC ( is a periodic unction with period / T. PSD is not lat in the mid-bands. 0 D/A Conversion - 33 / 54

Review o Non-IDFT-Based Signal Model and its PSD (cont. OFDM signal model in IEEE 80..a Baseband signal rame : r - t SIGNAL = 6µs, t Data = 0µ s Subrame : r PACET ( t = w ( t = r PREAMBLE - w T, SUBFRAME (t : Continuous-time windowing unction with duration t TR - C : Data, pilots, or training symbols ( t + r N ST / - =/ : Subcarrier requency spacing (=/T 0 in our models T FFT SUBFRAME T,SUBFRAME - T GUARD : Guard time to create the cyclic preix SIGNAL ( t t SIGNAL ( t t For the short training seq. (=0 s, long training seq. (=T GI, data OFDM symbols (=T GI Windowing unction may have the time duration extended over than one OFDM period, T FFT, and the transition time overlapped via an adjacent OFDM symbol windowing. (Also in our signal model ( t C = NST / e jπ ( t + r TGUARD DATA DATA D/A Conversion - 34 / 54

Review o Non-IDFT-Based Signal Model and its PSD (cont. OFDM signal model in IEEE 80..a (cont. Timing-related parameter Parameter Value N SD : Number o data subcarriers N SP : Number o pilots subcarriers N ST : Number o total subcarriers : Subcarrier requency spacing T FFT : IFFT/FFT period T PREAMBLE : PLCP Preamble duration T SIGNAL : Duration o the SIGNAL BPS-OFDM symbol T GI : GI duration T GI : Training symbol GI duration T SYM : Symbol interval T SHORT : Short training sequence duration T LONG : Long training sequence duraion 48 4 5 (N SD +N SP 0.35 MHz (=0 MHz/64 3. s ( / 6 s (T SHORT + T LONG 4.0 s (T GI + T FFT 0.8 s (T FFT /4.6 s (T FFT / 4 s (T GI + T FFT 8 s (0T FFT /4 8 s (T GI + T FFT D/A Conversion - 35 / 54

Review o Non-IDFT-Based Signal Model and its PSD (cont. OFDM signal model in IEEE 80.6.e and WiBro OFDM transmit signal NST / s( t = Re C = NST 0 / j π,0 t T - C : Complex number o data symbol (QAM - = F / N : Subcarrier requency spacing s FFT - N FFT : Smallest power by greater than number o used subcarriers - F s : Smallest power by greater than number o used subcarriers - T g : Guard time to create the cyclic preix e ( t t g s D/A Conversion - 36 / 54

IDFT-Based Signal Model and its PSD. IDFT-based discrete-time signal model Problem o the non-idft signal model : Its straightorward implementation requires local oscillators that are tuned to dierent subcarrier requencies { } where κ = : α 0, i.e., is the number o non-null subcarriers. j π t Idea : Using the times sampling o the signal e in the interval at the rate /T 0 or =,,, Orthogonal vectors that are proportional to the column vectors o the - point IDFT matrix The most common discrete-time signal model or a single vector o OFDM symbols is κ 0 t < T 0 κ Y = S X j where S is the -point IDFT matrix whose (, lth entry is given by s [ l] = e, X is the vector consisting o data symbols { X } and Y is the vector o OFDM = symbol. κ πl D/A Conversion - 37 / 54

D/A Conversion - / 54 38 IDFT-Based Signal Model and its PSD (cont.. IDFT-based discrete-time signal model (cont. This procedure can be rewritten as Operation characteristics - The data symbols are orthogonally multiplexed. - The energy o dierent data symbol is centered at dierent subcarrier requencies. + + + = = j j j j j j j j j e e e X e e e X e e e X S X Y π π π π π π π π π Μ Λ Μ Μ { } X =

IDFT-Based Signal Model and its PSD (cont.. IDFT-based discrete-time signal model (cont. For a transmission over a continuous-time channel, the OFDM symbol is D/Aconverted then up-converted using a local oscillator whose requency is tuned to the center requency o the transmitted bandpass signal. perormed eiciently without multiple local oscillators In many case, IDFT multiplication can be implemented by a computationally eicient inverse Fast Fourier Transorm (IFFT instead o IDFT, then multiplication requires only O ( NlogN operations reduced rom O( N Not or a single OFDM symbol but or a sequence o OFDM symbols, what is the signal model considering all o analog wideband ilter, D/A converter, digital interpolation ilter, cyclic preix (or postix, windowing sequence? Needs some modiications to IDFT-based continuous-time signal model D/A Conversion - 39 / 54

IDFT-Based Signal Model and its PSD (cont.. IDFT-based continuous-time signal model Revised system structure or IDFT-based transmitter bloc diagram or OFDM Data Sequence Modulation and Coding S/P Subcarrier weighting IDFT Add Cyclic Preix/ Postix Null subcarrier insertion Transmit signal to channel RF Device (Upconversion Complex baseband signal Z(t Analog iltering D/A P/S Windowing P/S - IDFT matrix point : Digital interpolation - Length o cyclic preix : L+ M, cyclic postix : M L ( channel length, M : # o overlapped entries o the vectors ater P/S - Length o windowing sequence : + L+ M D/A Conversion - 40 / 54

IDFT-Based Signal Model and its PSD (cont.. IDFT-based continuous-time signal model (cont. D/A converter : would be unctions such as the sample-and-hold - I the D/A input is cloced in every T c [sec], then D/A converter can be viewed as a linear modulator employing a rectangular pulse with duration T c. The D/A converter combined with the analog ilter can be viewed as a linear modulator employing a pulse, say p(t with the pulse transmission rate /T c (also when D/A uses a digital interpolation that operates oversampling. Windowing sequence : Usually designed to properly weight the irst M and the last M symbols, The proper shape o the roll-o o the windowing sequence reduces the inter-channel intererence (ICI. Null subcarriers : Not to use an analog ilter with a sharp transition rom passband to stopband, some carriers whose the index s are around being nulled out. - The useul subcarriers will be less than, but still -point IDFT is perormed with data symbols. D/A Conversion - 4 / 54

IDFT-Based Signal Model and its PSD (cont.. IDFT-based continuous-time signal model (cont. General orm o the complex baseband equivalent or IDFT-based continuoustime signal model is, α Z( t = = m= [ m] s - is the subcarrier weighting actor. - X [m] is the mth symbol or the th subcarrier. - T b is the OFDM symol bloc period given by T = ( + L + M. - T c is the transmit pulse time duration. (OFDM subcarrier spacing is /T 0 [Hz], so T c =T 0 - s (t is the th subcarrier waveorm. α X ( t mt b. b T c D/A Conversion - 4 / 54

IDFT-Based Signal Model and its PSD (cont.. IDFT-based continuous-time signal model (cont. s (t is deined by considering the windowing sequence w D [l], IDFT matrix element s [l], and transmit pulse p(t. So, s (t is given by s + L + M ( t = α l= [ l] p( t lt where α [ l] = wd[ l] s[ l] or =,,3,...,, and l =,,3,..., + L + M. πl j (s [l] is rom IDFT matrix (, lth element and is deined as s [ l] = e, but now up to l = + L + M p(t is the equivalent transmit pulse including all the eects o the D/A converter, digital interpolation ilter, and analog wideband ilter. This modiied signal model loos similar to a symbol synchronous DS-CDMA signal model where s (t serves as the signature waveorm and a [l] serves as the signature sequence, both the th users. c D/A Conversion - 43 / 54

D/A Conversion - / 54 44 IDFT-Based Signal Model and its PSD (cont. Power spectral density (PSD o IDFT-based continuous signal model Fourier transorm o the th subcarrier waveorm s (t Using above result, PSD o this signal model is obtained as where W D ( is the discrete-time Fourier transorm o the windowing sequence deined as 0 0 ( ( ( ( = + + = D ZZ T W T M L P S α { } M L l w D l ] [ + + =. ] [ ( + + = = M L l l T j D D c e l w W π = = + + = 0 0 exp T W P( T j ] [l w P( ( S D M L D π

IDFT-Based Signal Model and its PSD (cont. Power spectral density (PSD o IDFT-based continuous signal model (cont. Non-IDFT-based : S ZZ P( ( = α T b = W C ( IDFT-based : S ZZ P( ( = ( + L + M T α WD( 0 = T0 Signiicant dierence rom the non-idft-based model is that the Fourier transorm + L+ M o the windowing sequence { w D [ l] } l= is used instead o that o the continuous windowing unction w c (t [Note] Two properties o W D ( ( W D ( is periodic with a period /T c ( W D ( T0 is periodic in with a period under the assumption T c =T0 D/A Conversion - 45 / 54

IDFT-Based Signal Model and its PSD (cont. Example o PSD o IDFT-based continuous signal model Parameter Rectangular windowing sequence or (a : =0, L+ M =0 (b : =0, L+ M = [Note] W D ( sin( π ( + L+ = sin( πt c M T c Increase in (+L+M Decrease in the width o the main lobe and increase in the amount o in-band energy contained in the interval T0 T0 Periodic with period /T c D/A Conversion - 46 / 54

IDFT-Based Signal Model and its PSD (cont. New insight into PSD o OFDM signals Deinition I the aggregate transmit pulse p(t has a positive excess bandwidth, then, due to the periodicity o W D (, a logical subcarrier can have the signal energy + l concentrated at = or some integers l, that is, a logical subcarrier can have more than one physical subcarriers. T 0 Nominal subcarrier : Physical subcarrier with the greatest energy, and Ghost subcarriers : Other physical subcarriers except the nominal subcarrier For some OFDM logical subcarriers, we may observe two physical subcarriers, the nominal subcarriers and ghost carrier, using IDFT-based continuous signal model. D/A Conversion - 47 / 54

IDFT-Based Signal Model and its PSD (cont. Example : Energy spectral density o the subcarrier waveorms p(t is the SRRC pulse with =0.5, =8, L=0, M=0, no null subcarrier The subcarriers with indexes =3,4, and 5 have the signiicant portion o the energy split into two physical subcarriers located /T c [Hz] apart. D/A Conversion - 48 / 54

IDFT-Based Signal Model and its PSD (cont. Example : Virtual roll-o eect by using proper null-subcarrier Very ast roll-o in the PSD may be needed, but designing such a ilter is costly. Because we need a high-rate D/A converter, high order wideband transmit ilter Eicient solution : By nulling subcarriers with indexes PSD S ZZ ( o complex baseband signal Z(t Parameter - =8, L=0, M=0, 9 null-subcarriers - Sample-and-hold in D/A converter (a : No wideband analog iltering (b : Use o an analog ilter to mae p(t the SRRC pulse with =0. Possible to mae a transition band very small D/A Conversion - 49 / 54

IDFT-Based Signal Model and its PSD (cont. D/A Conversion - 50 / 54

IDFT-Based Signal Model and its PSD (cont. In WiBro standard, the signal model is given by NST s( t = Re = NST 0 / C / e jπ ( t t g,0 t T s A rectangular windowing unction without any window overlapping is used. Non-IDFT-based signal model is used. To use IFFT in the transmitter, we need the ollowing or an equivalent : A rectangular windowing sequence without any window overlapping A sample-and-hold D/A converter with inverse-sinc shaped transmit ilter in the requency band or non-null subcarriers D/A Conversion - 5 / 54

IDFT-Based Signal Model and its PSD (cont. W D ( = sin( π ( + L Tc sin( πt sin( π ( + L T πt c c c or large +L and or 0 D/A Conversion - 5 / 54

IDFT-Based Signal Model and its PSD (cont. The transmit ilter must invert the sinc square envelope to equally distribute the power among non-null subcarriers. The transition rom the passband to stopband must lie over the nullsubcarrier requency band. D/A Conversion - 53 / 54

Conclusions D/A conversion is a necessary procedure in modern digital audio systems and in digital communication systems. Y ( P( X ( T c = d Upsampling combined with interpolation is oten used in commercial DACs and in modern radio transceiver. Y ( X ( N d = d IFFT-based OFDM signaling also requires D/A conversion. S P( ZZ ( = α WD( ( + L+ M T 0 = T0 To ease the burden o designing a good interpolation ilter, null subcarriers are used. The interpolation ilter must have the transition rom passband to stopband in the null subcarriers requency band. D/A Conversion - 54 / 54