Analysis to P eak - to -A verage P ower R atio in OFDM Systems

Analysis to Peak-to-Average Power Ratio in OFDM Systems

指導老師 : 黃文傑 博士研究生 : 吳濟廷

2003.7.30

OUTLINE

What is PAPR Why PAPR happens Simulations How to solve the PAPR problem Conclusion

What is PAPR

Definition

2

2

max x tPAPR N

E x t

1 1

0 0

1

0

2 2

exp( 2 ) exp( 2 )N N

n nn n

N

nn

n

x t X j n ft X j n ft

X N

X N

2 22 x t N

E x t NN N

Why PAPR happens(1/2)

Large peaks cause saturation in power amplifiers

Why PAPR happens(2/2)

Hard to extend the linear working area of amplifier

PAPR depends on the number of subcarriers

Simulations(1/3)

Transmit block using QPSK

Simulations(2/3)

QPSK Discrete-time Scatter Plot Scope

Simulations(3/3)

Spectrum Scope after IFFT

How to solve the PAPR problem(1/6) Signal distortion

Clipping, Peak windowing, Peak cancellation Coding

Error correction, Use lower PAPR signals Scrambling

How to solve the PAPR problem(2/6) Peak windowing

Multiplied by certain window function Rectangular, Kasier, Hanning, Hamming …..

Let’s discuss … Different window with same window length Same window with different window length

How to solve the PAPR problem(3/6) Different window with same window length (time domain)

rectangular kaisar hamminghanning

How to solve the PAPR problem(4/6) Different window with same window length (frequency domain)

How to solve the PAPR problem(5/6) Same window with different window length-hamming window (time domain)

N=7 N=11

N=13 N=15

How to solve the PAPR problem(6/6) Same window with different window length-hamming window (frequency

domain)

Conclusion

Advantages Simple to implement Independent of number of carriers Large reduction in PAPR

Disadvantages No longer orthogonal Difficult to find the relation between windowing

parameter and PAPR