PAPR REDUCTION OF MIMO-OFDM SYSTEM USING MODIFIED CLIPPING AND TURBO CODING

Abstract

Orthogonal frequency division multiplexing (OFDM) has been the most widely used multicarrier in recent mobile communication system due to its multiple ad vantages like high spectral efficiency, high data rate transmission over a multipath fading channel, simple implementation, and low receiver complexity. Currently, multiple-input multiple-output (MIMO) is combined with OFDM to boost spectral efficiency, work in a frequency selective environment, and enhance link reliability. Besides those crucial advantages, MIMO-OFDM system suffers from a high peak to-average power ratio (PAPR) that originates from the superimposition of the number of sub-carriers with random amplitudes and phases. This high PAPR re sults in high power consumption by the power amplifier and high complexity in the design of analog to digital converter. In this thesis work, modified clipping based PAPR reduction technique has been designed for MIMO-OFDM system and it is integrated with turbo channel coding scheme. In modified clipping method the MIMO-OFDM signal in each transmitter antenna has been clipped into different threshold values using a settle quantiza tion levels, step size value and number of clipped high peak envelope signals as parameters. A powerful turbo channel coding enables to reduce the clipping and channel fading distortion with the aids of iterative decoding using space-time block code (STBC) bit-wise soft detection values as input. This allows to achieve better bit error rate (BER) performance. The PAPR and BER of the turbo code based modified clipping STBC MIMO-OFDM have been verified using Matlab software for the selected numerical values of modified clipping parameters. In addition, the designed modified clipping performance has been compared to the conventional clipping performance using PAPR and BER metrics. The simulation result demonstrated that the modified clipping having compa rable PAPR performance to that of the conventional clipping has better BER performance. Using modified clipping method having 25% of clipped high peak envelopes, 5 quantization levels, and 0.05 step size with 1/3 rate turbo code the PAPR performance can improve by 69.09%, 71.05%, and 72.65% for a sequential

sub-carrier length of 128, 256, and 512 over its respective original PAPR perfor mances. Under those modified clipping parameters value with different turbo code iteration the remarkable bit energy to noise power spectral density ratio Eb/N0 gain has been obtained at a given BER value