ofdm dft dft inverse dft an n-point dft (or inverse dft) requires a total of n 2 complex...

4
OFDM DFT DFT Inverse DFT An N-point DFT (or inverse DFT) requires a total of N 2 complex multiplications This transform can be implemented very efficiently by the fast Fourier transform (FFT) Serial-to- parallel converter Parallel-to- serial converter Signal m apper Signal dem apper ID FT DFT P/S S/P Guard interval / cyclic prefix insertion C P rem oval channel 1 , , 1 , 0 , 2 exp ] [ ) ( 1 0 N k N nk j n x k X N n 1 , , 1 , 0 , 2 exp ] [ ] [ 1 0 N n N nk j k X n x N n

Upload: egbert-carpenter

Post on 14-Dec-2015

217 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: OFDM DFT  DFT  Inverse DFT  An N-point DFT (or inverse DFT) requires a total of N 2 complex multiplications  This transform can be implemented very

OFDM

DFT DFT

Inverse DFT

An N-point DFT (or inverse DFT) requires a total of N2 complex multiplications This transform can be implemented very efficiently by the fast Fourier

transform (FFT)

Serial-to-parallel

converter

Parallel-to-serial converter

Signal mapper

Signal demapper

IDFT

DFT

P/S

S/P

Guard interval /

cyclic prefix insertion

CP removal

channel

1,,1,0,2exp][)(1

0

NkN

nkjnxkX

N

n

1,,1,0,2exp][][1

0

NnN

nkjkXnx

N

n

Page 2: OFDM DFT  DFT  Inverse DFT  An N-point DFT (or inverse DFT) requires a total of N 2 complex multiplications  This transform can be implemented very

OFDM OFDM implementation

An OFDM signal consists of a sum of subcarriers that are modulated by using QAM, and is implemented using the inverse DFT (discrete Fourier transform)

di : the i-th complex QAM symbol N : the number of subcarriers

1,,1,0,2exp)(1

0

NnN

injdns

N

ii

Serialto

parallelQAM data

dN-1 … d2 d1 d0

dN-1

d1

d0

X

X

X

Exp(j*2*pi*0*n/N)

Exp(j*2*pi*1*n/N)

Exp(j*2*pi*(N-1)*n/N)

+s(N-1) … s(2) s(1) s(0)

Page 3: OFDM DFT  DFT  Inverse DFT  An N-point DFT (or inverse DFT) requires a total of N 2 complex multiplications  This transform can be implemented very

OFDM FFT

The FFT drastically reduces the amount of calculations by exploiting the regularity of the operations in the DFT

Using the radix-2 algorithm, an N-point FFT (or inverse FFT) requires only (N/2)log2N complex multiplications 256 multiplications for DFT versus 32 for FFT (for a 16-point transform) – a

reduction by a factor of 8 In VDSL system, 4096-point transform is used – a reduction by a factor of

about 683

Page 4: OFDM DFT  DFT  Inverse DFT  An N-point DFT (or inverse DFT) requires a total of N 2 complex multiplications  This transform can be implemented very

Homework 16-point DFT & 16-point inverse DFT programming 16-point FFT & 16-point inverse FFT programming

(Tip) Definition for complex variable should be needed

Complex variable (real + imaginary, A=a+ib) Complex multiplication (A*B=(a+ib)*(c+id)=(ac-bd)+i*(ad+bc)) Complex addition (A+B=(a+ib)+(c+id)=(a+c)+i*(b+d)) Complex subtraction (A-B=(a-c)+i*(b-d))

Block diagram

마감 : 12 월 12 일 수요일 낮 12 시까지 ( 교수 이메일로 보낼것 ) 제출자료

source program (c-code) with comments .exe file (file_name = student_id.exe) for demonstration

Random generator

16 complex symbols Serial-to-

parallel

16-point inverse

DFT

16-point DFT

parallel-to-serial

printf(these values) printf(these values) printf(these values)

Random generator

16 complex symbols Serial-to-

parallel16-point

inverse FFT16-point

FFTparallel-to-

serial

printf(these values) printf(these values) printf(these values)