Massive MIMO systems are the key technology for evolution of 4G, 5G standards in telecommunication environment. One of the major limitations in OFDM based massive multiple-input multiple-output (MIMO) downlink system is peak to average power ratio (PAPR). Transmitting symbol vectors to different set of users, the main idea is estimate the low PAPR OFDM modulated signal with reduced multi user interference. Many techniques were used to mitigate the PAPR problem, but they consume more computational time, particularly in Massive MIMO systems. The proposed ETG (Expectation maximization Truncated Gaussian mixture Generalized approximate passing) employs truncated Gaussian mixture prior to get low PAPR signal. To understand the prior signal, expectation step helps in identifying hidden variables; maximization step helps in identifying deterministic parameters. Generalized approximate passing is applied to mitigate the computational complexity. Numerical simulated results in comparison with existing techniques suggests that desired level of PAPR is achieved with less computation time with minute degradation in symbol error rate(SER).By choosing proper normalization we can achieve same SER with reduced PAPR.
EFFICIENT ANTENNA SELECTION STRATEGY FOR A MASSIVE MIMO DOWNLINK SYSTEM
