Enhanced peak to average power ratio (PAPR) algorithm in long term evolution (LTE) standard

Abstract

Long Term Evolution (LTE), is an OFDM based system considered the 4G standard that supports high mobility up to 350km/h and high bandwidths. However, LTE suffers from high peak to average power ratio (PAPR) which critically affects the system performance by increasing the bit error rate and decreasing the spectral efficiency. Many solutions have been introduced to the LTE standard to overcome this issue. A hardware solution has been suggested utilizes high power amplifiers with very wide linearity area covering the high peak area of the OFDM symbol, but, the solution is expensive rendering it not affordable by end users. Thus, PAPR reduction algorithms come into place to resolve the issue by manipulating the OFDM symbols blocks produced by MIMO-OFDM systems. The result is an increase in the efficiency of the whole system by reducing PAPR. However, a lot of these algorithms suffer increasing computational complexity to achieve high PAPR reduction performance which affects the overall circuitry complexity. A modified algorithm has been proposed in this research that aims to achieve high PAPR reduction performance while keeping the computational complexity level as low as possible. A modified Cross Antenna Rotation and Inversion (CARI) is introduced in this work, where the normal CARI operations are applied to the first two blocks in the space domain which maintains the circuitry complexity lower than the original CARI, followed by a rotation in the space and frequency domains, hence, utilizing the degree of freedom inherent in both OFDM-MIMO space and frequency domains. The proposed PAPR reduction scheme has been evaluated using Matlab simulator. For simulation approach, the performance metric considered is the Complementary Cumulative Distribution Function CCDF averaged our 105 OFDM symbols. The results have shown that the proposed scheme has better performance by gain of 0.18dB than original CARI at CCDF=10-3 and the sub-blocks M=8. The bench marking against Individual SLM the proposed scheme has shown better performance by gain of 0.5dB than Individual SLM at CCDF=10-3 and the sub-blocks M=8, and finally with concurrent SLM the proposed scheme has better performance by gain of 0.8dB Individual SLM at CCDF=10-3 and the sub-blocks M=8.