Advanced telecommunications for wireless systems

Advanced telecommunications for wireless systems

Advanced telecommunications for wireless systems Investigating OFDM by MathCAD Timo Korhonen, Communications Laboratory, TKK

Motto If you tell me I forget If you show me I will remember If you involve me I can understand - a Chinese proverb

Topics The objective of workshop OFDM module is to get familiar with OFDM physical level by using MathCAD for system studies. Topics:

OFDM Signal in time and frequency domain Channel model and associated effects to OFDM Windowing

Cyclic prefix Peak-to-average power ratio (PAPR) OFDM transceiver Water-pouring principle System modeling: Constellation diagram, error rate System impairments

References for exercises Bahai, Ahmad R. S: Multi-Carrier Digital Communications : Theory and Applications of OFDM Hara, Shinsuke: Multicarrier Techniques for 4G

Mobile Communications Prasad, Ramjee: OFDM for Wireless Communications Systems Xiong, Fuqin: Digital Modulation Techniques. Norwood, MA, USA Exercise: Using MathCAD Plot the sinc-function Create a script to create and draw a rectangle waveform.

Demonstrate usage of FFT by drawing a sinwave and its spectra. Determine Fourier-series coefficients of a sinusoidal wave and plot the wave using these coefficients Prepare a list of problems/solutions encountered in your tasks.

Rect Spectra of a sinus Fourier transformation of a sinusoidal

Introduction Background Objectives: High capacity and variable bit rate information transmission with high bandwidth

efficiency Limitations of radio environment, also Impulse / narrow band noise Traditional single carrier mobile communication systems do not perform well if delay spread is large. (Channel coding and adaptive

equalization can be still improve system performance) OFDM Each sub-carrier is modulated at a very low symbol rate, making the symbols much longer

than the channel impulse response. Discrete Fourier transform (DFT) applied for multi-carrier modulation. The DFT exhibits the desired orthogonality and can be implemented efficiently through the fast fourier transform (FFT) algorithm.

Basic principles The orthogonality of the carriers means that each carrier has an integer number of cycles over a symbol period. Reception by integrate-and-dump-receiver

Compact spectral utilization (with a high number of carriers spectra approaches rectangularshape) OFDM systems are attractive for the way they handle ISI and ICI, which is usually introduced by frequency selective multipath fading in a wireless environment. (ICI in FDM)

Drawbacks of OFDM The large dynamic range of the signal, also known as the peak-to-average-power ratio (PAPR). Sensitivity to phase noise, timing and

frequency offsets (reception) Efficiency gains reduced by guard interval. Can be compensated by multiuser receiver techniques (increased receiver complexity)

Examples of OFDM-systems OFDM is used (among others) in the following systems:

IEEE 802.11a&g (WLAN) systems IEEE 802.16a (WiMAX) systems ADSL (DMT = Discrete MultiTone) systems DAB (Digital Audio Broadcasting)

DVB-T (Digital Video Broadcasting) OFDM is spectral efficient, but not power efficient (due to linearity requirements of power amplifier= the PAPR-problem). OFDM is primarily a modulation method; OFDMA is

the corresponding multiple access scheme. OFDM Signal Multiplexing techniques

OFDM signal in time domain OFDM TX signal = Sequence s t g k t kTS

k of OFDM symbols gk(t) consisting of serially converted complex data symbols

The k:th OFDM symbol (in complex LPE form) is N 2 n g k t an ,k exp j 2 t

TS n N 2 n 0 k 1 TS t kTS

where N = number of subcarriers, TG + TS = symbol period with the guard interval, and an,k is the complex data symbol modulating the n:th subcarrier during the k:th symbol period.

In summary, the OFDM TX signal is serially converted IFFT of complex data symbols an,k Orthogonality of subcarriers ( k 1)T

s (t )s (t )dt 0 Definition: 1

2 kT Orthogonality over the FFT interval: TFFT

TS 2 m n cos 2 mt TS cos 2 nt TS dt m n 0

0 Phase shift in any subcarrier - orthogonality over the FFT interval should still be retained: TFFT

cos 2 mt T S 0

cos 2 nt TS dt 0 m n Exercise: Orthogonality Create a MathCAD script to investigate

orthogonality of two square waves #1 Create the rect-function #2 Create a square wave using #1 #3 Create a square wave with a time offset #4 Add the waves and integrate

Exercise: Orthogonality of OFDM signals Create and plot an OFDM signal in time domain and investigate when your subcarriers are orthogonal #1 Create a function to generate OFDM symbol with

multiple subcarriers #2 Create a function to plot comparison of two subcarriers orthogonality (parameter is the frequency difference between carriers) Note: also phase continuity required in OFDM

symbol boarders #3 Inspect the condition for orthogonality and phase continuity

OFDM Spectra OFDM in frequency domain TG TFFT

Square-windowed sinusoid in time domain => "sinc" shaped subchannel spectrum in frequency domain sinc fTFFT sin fTFFT

fTFFT See also A.13 in Xiong, Fuqin. Digital Modulation Techniques. Norwood, MA, USA: Artech House,

Incorporated, 2006. p 916. id=10160973&ppg=932 Spectra for multiple carrier Single subchannel

OFDM spectrum Subcarrier spacing = 1/TFFT Spectral nulls at

other subcarrier frequencies Next carrier goes here!

Exercise: Analytical spectra Draw the spectra of OFDM signal by starting its frequency domain presentation (the sinc-function). Plot the spectra also in log-scale

#1 Plot three delayed sinc(x) functions in the range x = -12 such that you can note they phase align correctly to describe the OFDM spectra #2 Plot in the range from f = -20 to 20 Hz an OFDM spectra consisting of 13 carriers around

f=0 in linear and log-scale Ofdm Exercise: Spectra modified Investigate a single OFDM carrier burst

and its spectra by using the following script: How the spectra is changed if the Carrier frequency is higher Symbol length is altered

OFDM Spectra by MathCAD for a single carrier ofdm spectra by rect windowed Spectral shaping by windowing

Exercise: Windowed spectra The next MathCAD script demonstrates effect of windowing in a single carrier. How the steepness of the windowing is adjusted? Why function win(x,q) is delayed by ?

Comment the script burst windowing and ofdm Modeling OFDM Transmission

Transceiver Some processing is done on the source data, such as coding for correcting errors, interleaving and mapping of bits onto symbols. An example of mapping used is multilevel QAM. The symbols are modulated onto orthogonal

sub-carriers. This is done by using IFFT Orthogonality is maintained during channel transmission. This is achieved by adding a cyclic prefix to the OFDM frame to be sent. The cyclic prefix consists of the L last samples of the frame, which are copied and placed in the

beginning of the frame. It must be longer than the channel impulse response. OFDM and FFT ~ Aalborg-34-lecture13.pdf

Exercise: Constellation diagram of OFDM system Steps #1 create a matrix with complex 4-level QAM constellation points

#2 create a random serial data stream by using outcome of #1. Plot them to a constellation diagram. #3 create complex AWGN channel noise. Calculate the SNR in the receiver. #4 form and plot the received complex noisy time domain waveform by IFFT (icfft-function)

#5 detect outcome of #4 by FFT and plot the resulting constellation diagram Exercise : Constellation diagram of OFDM Ofdm

Channel Combating multipath channel Multipath prop. destroys orthogonality Requires adaptive receiver channel sensing

required (channel sounding by pilot tones or using cyclic extension) Remedies Cyclic extension (decreases sensitivity) Coding

One can deal also without cyclic extension (multiuser detection, equalizer techniques) More sensitive receiver in general More complex receiver - more power consumed Pilot allocation example

To be able to equalize the frequency response of a frequency selective channel, pilot subcarriers must be inserted at certain frequencies: Time Pilot subcarriers at some,

selected frequencies Between pilot subcarriers, some form of interpolation is necessary!

Subcarrier of an OFDM symbol Frequency Pilot allocation example cont.

- A set of pilot frequencies The Shannon sampling theorem must be satisfied, otherwise error-free interpolation is not possible: D f 1 2Tm Time

Tm maximum delay spread Df Frequency

Channel Path Loss Shadow Fading Multipath: Flat fading Doppler spread

Delay spread Interference OFDM: Inter-symbol interference (ISI) flat fading, sampling theorem must be fulfilled

Inter-carrier interference (ICI) multipath propagation (guard interval) Multipath radio channel *

*Spike distance depends on impulse response Multipath channel model Received signal

Pr: Received mean power Gs: Shadow fading Log r ~ Fast fading Exercise: Modeling channel

Create a MathCAD script to create artificial impulse and frequency response of a multipath channel (fast fading) #1 Create an array of complex AWGN #2 Filter output of #1 by exp(-5k/M) where M is the number of data points

#3 Plot the time domain magnitude of #2 Is this a Rayleigh or Rice fading channel? How to make it the other one than Rayleigh/ Rice #4 Plot #3 in frequency domain

Comment how realistic this simulation is? Rayleigh or Rice fading channel? impulse response radio Frequency response

Frequency response shown by swapping left-hand side of the fft Exercise: Rayleigh distribution #1 create a Rayleigh distributed set of random numbers (envelope of complex Gaussian rv.)

#2 plot the pdf of #1 (use the histogramfunction) #3 add the theoretical pdf to #2 Rayleigh distribution - Note that true pdf area equals unity, how could you adjust the above for this?

- Add comparison to the theoretical Rayleigh distribution! Path loss Shadow fading

N(0,s2): Log normal distribution Doppler spread DELAY SPREAD IN TIME DOMAIN

Small delay spread Large delay spread Delay spread in frequency domain

Exercise: Variable channel Discuss a model of a channel with flat/ frequency selective characteristics and report its effect to the received modulated wave Amplitude and phase spectra What happens to the received frequency components


Flat fading Frequency selective fading Time invariant / time variant channel Doppler effected channel Exercise: ODFM in a multipath

channel #1 Create an impulse response of 256 samples with nonzero values at h2=16, h10=4+9j, and h25= 10+3j and plot its magnitude spectra #2 Create OFDM symbol for three subcarriers with 1,2 and 3 cycles carrying bits 1,-1 and 1

#3 Launch the signal of #2 to the channel of #1 and plot the OFDM signal before and after the channel to same picture. #4 Detect (Integrate an dump) the bits after and before the channel and compare. See the generated ICI also by detecting the 4:n carrier!

The channel OFDM in a multipath Single carrier transmission and

rms delay spread Determine by using MathCADs linterp-function the maximum rate for delay spread of 70 us!

Rate and delay Cyclic prefix Example: a two-path channel

Home exercise: Orthogonality and multipath channel Demonstrate by MathCAD that the orthogonality of OFDM signal can be maintained in a multipath channel when guard interval is applied #1 modify syms2-signal to include a cyclic prefix

#2 introduce multipath delays not exceeding the duration of cyclic prefix (apply the rot-function) #3 determine integrate and dump detected bits for #2 and especially for carriers that are not used to find that no signal is leaking into other subcarriers -> ICI is avoided!

cyclic Detection Peak-power problem

Envelope Power Statistics Exercise Create an OFDM signal in time domain and determine experimentally its PAPR

Experiment with different bit-patterns to show that the PAPR is a function of bit pattern of the symbol #1 create 64 pcs BPSK LPE bits #2 define a function to create OFDM symbol with the specified number of carriers (with 256 samples)

carrying bits of #1 #3 check that the carriers are generated correctly by a plot #4 determine PAPR for a set of 64 OFDM subcarriers. Compare with different bit patterns (eg. evaluate #1 again by pressing F9)

OFDM Transfer characteristics of power amplifier

Exercise: Non-linear distortion Demonstrate build-up of harmonics for a sinusoidal wave due to non-linearity of a power amplifier #1 Create a sinusoidal wave, 256 samples and 8 cycles

#2 Create a clipping function that cuts a defined section of waves amplitudes #3 Apply #2 to #1 and plot the result #4 compare #1 to #3 in frequency domain logscale with different levels of clipping PAPR suppression

Selective mapping (coding) Cons: Table look-up required at the receiver Signal distortion techniques Clipping, peak windowing, peak cancellation Cons: Symbols with a higher PAPR suffer a

higher symbol error probability Prasad, Ramjee. OFDM for Wireless Communications Systems. Norwood, MA, USA: Artech House, Incorporated, 2004. p 150.

id=10081973&ppg=166 Selective mapping Clipping Cancel the peaks by simply limiting the

amplitude to a desired level Self-interference Out-of band radiation Side effects be reduced by applying different clipping windows

Peak cancellation Comparing clipping and peak cancellation

a) undistorted b) peak cancellation c) clipping Effect of peak cancellation on packet error rate (PER)

OFDM transceiver System error rate In AWGN channel OFDM system performance same as for single-carrier

In fading multipath a better performance can be achieved Adjusts to delay spread Allocates justified number of bits/subcarrier DESIGN

EXAMPLE Exercise: BPSK error rate How would you modify this function to simulate unipolar system?

Bit allocation for carriers Each carrier is sensed (channel estimation) to find out the respective subchannel SNR at the point of reception (or channel response) Based on information theory, only a certain,

maximum amount of data be allocated for a channel with the specified BT and SNR OFDM bit allocation policies strive to determine optimum number of levels for each subcarrier to (i) maximize rate or (ii) minimize power for the specified error rate

Water-pouring principle Assume we know the received energy for each subchannel (symbol), noise power/Hz and the required BER Assume that the required BER/subchannel is the

same for each subchannel (applies when relatively high channel SNR) Water-pouring principle strives to determine the applicable number of levels (or bit rate) for subcarriers to obtain the desired transmission

Home exercise 2: Water pouring principle Follow the previous script and Explain how it works by own words Comment the result with respect of information theory

System impairments Exercise ICI

Create a MathCAD script to investigate frequency offset produced ICI Summary

Recently Viewed Presentations

  • Land Planning Concepts

    Land Planning Concepts

    Steps of HIA Assessment, which is a two step process that first describes the baseline health status and then assesses potential impacts Scoping, to establish the scope of health effects that will be included in the HIA, the populations affected,...
  • Bridge BetweenSWDA & CWA - New York Department of State

    Bridge BetweenSWDA & CWA - New York Department of State

    Nutrient-Related Concerns. Health. University of Toledo. Aquatic Life. Recreation. This next slide is intended to illustrate New York States general approach to the development of nutrient criteria, which is basically to derive criteria based on best use.
  • BizTalk Integration Summit 2013 - Microsoft Developer Network

    BizTalk Integration Summit 2013 - Microsoft Developer Network

    Clinical Enterprise Integration using BizTalk - Getting BizTalk to Process 25M+ Messages a Day. Alan Scott, Enterprise Integration Architect HCA
  • Vocabulary: fraction improper terminating decimal  simplest numerator recurring

    Vocabulary: fraction improper terminating decimal simplest numerator recurring

    Convert between fractions, decimals and percentages including improper fractions and mixed numbers. Complete the four operations with fractions and mixed numbers. Calculate fractions and decimals of quantities. Round off decimals. Express one quantity as a fraction of another. Order fractions,...


    Dobavljač papira Gramatura papira [g/m2] Trgovački naziv papira RB TEHNOLOŠKI PROCES Knjigoveška proizvodnja LIJEPLJENA MEKO UVEZANA KNJIGA 0,8 mm Deb. nanosa ljepila 130 st. Celzijusa Radna temperatura termalnog (hot-melt) ljepila Planatol HM 6010 Kt.oznaka ljepila Pony 5 Tip stroja Muller...
  • 7.3 Structures and Organelles -

    7.3 Structures and Organelles -

    7.3 Structures and Organelles. Objectives: 4(A) Compare and contrast prokaryotic and eukaryotic cells. 4(B) Investigate and explain cellular processes, including homeostasis, energy conversions, transport of molecules, and synthesis of new molecules.
  • Virginia Department of Education Module Two Preparing to

    Virginia Department of Education Module Two Preparing to

    Walking, Driving, or Riding. Vehicles. Bicycles, Cars and SUVs, Mopeds Tractor-Trailer Trucks. Roadways . From the smallest dirt road to the complex multilane expressways. The Highway Transportation System (HTS) Three Components of the HTS. QUESTIONS AND ANSWERS: People. Question: How...
  • Observations How we look at things in science

    Observations How we look at things in science

    The Latin word "Dei" means "God." The coin was made by deeply religious people. The date 1722 is printed on one side of the coin. The coin was made in 1722. The face on the coin is a representation of...