Improved DFT-based channel estimation for spatial modulated orthogonal frequency division multiplexing systems

Abstract

In current cellular networks, Spatial Modulation with Orthogonal Frequency Division Multiplexing (SM OFDM) is a newly designed transmission method that substitutes Multiple Input Multiple Output (MIMO) OFDM transmission. In practical frameworks, estimation of the channel is vital for recovering the transmitted data. In a conventional DFT-based channel estimation algorithm, impact of noise is reduced in the time domain. It retains the significant values with some sort of threshold value equal to the Cyclic Prefix (CP) length that differentiates the signal values and noise terms of Channel Impulse Response (CIR), thereby improving the performance in OFDM systems. Here we require information on CIR length in advance. But this results in performance degradation if the path delays are less for a channel. Hence this work proposes improved DFT-based channel estimation where the threshold value is set equal to the estimated CIR length using RMS delay spread approximation in SM OFDM systems. Symbol Error Rate (SER) and Mean Square Error (MSE) are the performance metrics considered in Rayleigh fading channel employing ITU- Vehicular A channel model for 16 QAM SM OFDM systems. Simulation output shows performance improvement of SER and MSE by the implementation of improved DFT-based channel estimation. At an SNR of 40 dB, the proposed method can enhance SER performance by about 2.8 dB and MSE performance about 4dB compared with conventional DFT estimation.