Understanding Peak to Average Power Ratio (PAPR)

This article describes the basics of Peak to Average Power Ratio (PAPR), its importance in wireless systems, and PAPR reduction techniques.

Description

In modulation schemes such as OFDM, the transmitter’s time-domain signal can exhibit a high PAPR. This leads to various distortions in the transmitter chain and degradation of system performance (i.e., BER/PER). It also necessitates the use of a highly linear power amplifier, increasing the system’s cost.

PAPR is the ratio of the peak power to the average power of the time-domain complex baseband signal that is to be transmitted. We will explore PAPR reduction techniques below.

PAPR Reduction Techniques

The following sections describe various PAPR reduction techniques: Clipping, Peak Windowing, Scrambling, Block Coding, Selected Mapping (SLM), and Partial Transmit Sequences (PTS).

Clipping

The simplest method involves clipping the transmitted signal. This can be modeled as the multiplication or convolution of a signal with high PAPR with a window function. Undesired widening of the signal due to clipping is limited using filter techniques. Various window functions, such as cosine, Kaiser, and Hamming, are available.

Pros: Simple

Cons: Causes interference/distortion/out-band emission, which degrades system performance. Filtering after clipping can reduce distortion/emission but may also cause some peak re-growth.

Peak Windowing

Here, the large signal peak of the signal is multiplied with a Gaussian-shaped window. A suitable window function is selected for multiplication from Cosine, Kaiser, and Hamming windows.

Pros: Simple

Cons: Both BER and out-of-band radiation are increased.

Scrambling

This technique utilizes scrambling to alter the polarities of the subcarriers, removing the correlation among the subcarriers irrespective of the user codes used. PAPR will be greatly reduced as it spreads the information over a larger band.

Pros: About 5 dB of PAPR reduction can be achieved using this technique in certain cases.

Cons: Gives poor performance when the number of subcarriers increases.

Block Coding

Find code words with minimum PAPR from a given set of code words. Map the input data blocks to these selected code words. All block codes provide a low PAPR, typically below 3 dB for a small number of carriers.

Pros: Lowest PAPR

Cons: Low data rate and significant overhead

Block Coding Sequences most commonly used are Walsh, Gold, Orthogonal Gold, and Zadoff-Chu sequences. Zadoff Chu is used in LTE.

Selected Mapping (SLM)

The concept here is that any single data vector of the transmitter signal can have multiple representations. Out of these, the lowest PAPR time-domain vector is selected for transmission.

Pros: Lowest PAPR

Cons: Complexity issue, as the SLM scheme needs multiple IFFT operations.

Partial Transmit Sequence (PTS)

This method is similar to SLM but divides the frequency vector into smaller blocks before applying the phase transformations.

Pros: Lowest PAPR and little redundancy.

Cons: Increased system complexity.