On BER assessment of conventional- and wavelet-OFDM over AWGN channel

Abstract

Orthogonal wavelets enclose more potential of dropping the ISI- and ICI-sway amid the carriers due to multipath propagation over the free-space fading channels. Moreover, wavelet-OFDM is more bandwidth efficient as compared to conventional OFDM and is less affected by Doppler shift. Accordingly, a BER estimation of discrete Fourier transform (DFT) based OFDM system and discrete wavelet transform (DWT) based OFDM system is carried out in this work. The assessment is demonstrated by means of diverse modulation schemes explicitly BPSK, QPSK, 8-PSK, QAM, 8-QAM, 16-QAM for both conventional- and wavelet-OFDM wireless system over AWGN channels. Higher orders of PSK schemes involve more SNR requirement to target an acceptable BER of 10⁻⁵ over AWGN channels in both the cases. Furthermore, the efficiency of W-OFDM increases with decreasing the required value of BER.

Introduction

Necessitating of high speed communication turns us to broadband communication. In the current and future mobile communications systems, data transmission at high bit rates is crucial for many services like high quality audio/video services and mobile integrated-service-digital-network. When the data is transmitted at high bit rates over mobile radio channels, the channel impulse response can widen over many symbol periods and leads to inter-symbol interference (ISI). As the channel width grows, it is tricky to model the channel. Therefore, the splitting of a channel into smaller sub-channels is recommendable to realize OFDM, a multi-carrier transmission technique. One of the main reasons to exercise OFDM is to increase the robustness against frequency selective fading or narrowband interference. In a single carrier system, a single fade or interferer can cause the whole link to fail, but in case of multi-carrier system, only a small proportion of the subcarriers will be affected. Furthermore, orthogonal based wavelet transformation [1], [3], [4], [6], [7] is an additional method to study the signals in the joint time–frequency domains and enjoy better orthogonality. Orthogonal wavelets are capable of reducing the influence of inter symbol interference (ISI) and inter carrier interference (ICI) which are caused by loss of orthogonality between the carriers due to multipath propagation. In C-OFDM, ISI- and ICI-interference reduced by implying cyclic prefix (CP). On contrary, in W-OFDM, there is no need of CP and provides at least 20% or more bandwidth efficiency [2]. W-OFDM is less affected by Doppler shift as well and provides better orthogonality. Additionally, it offers better phase linearity and remarkable out-of-band rejection.

In preliminary stage, OFDM was projected for use in implementing multi-carrier modulation [8], [9]. Further, with the emergence and gradual completeness of wavelet analysis, discrete wavelet transform shows an ability to replace the FFT process in OFDM system. By varying the wavelet filter and the wavelet filter network, waveforms with selectable time/frequency partitioning for multi-user applications can be designed to achieve better multi-user capacity or to provide strong performance in difficult channel conditions [9]. Wavelet transform is a tool for studying signals in the joint time–frequency domains. Thus, it is capable of providing the time and frequency information at the same time, hence giving a time–frequency representation of the signal [10]. In time variable channels, Doppler shift occurs and affects the orthogonality of OFDM [11]. But W-OFDM is less affected by this shift and provide main side lobe of much lower magnitude than those of Fourier transform. An important property of W-OFDM is perfect reconstruction with reduced complexity, to reconstruct the transmitted signal at the receiver precisely [12]. This work addresses a recital contrast of C-OFDM and W-OFDM through diverse modulation schemes explicitly BPSK, QPSK, 8-PSK, QAM, 8-QAM, 16-QAM over AWGN channels. This paper is organized as: in Section 1, a brief introduction of the previous work demonstrated on C-OFDM and W-OFDM is discussed. Section 2 presents model description using Matlab simulink and result discussion, followed by the conclusion drawn in Section 3 on the basis of our observations.