In this work, a hardware efficientalgorithm for sequential in-phase (I) and quadrature (Q) imbalance (IQ imbalance) and carrier frequency offset (CFO) compensation under chromatic dispersion (CD) and phase noise (PN) environment is proposed. Two identical orthogonal frequency division multiplexing (OFDM) symbols, which are namely training sequences (TSs), are used to acquire CFO and IQ imbalance coefficients.The CFO is obtained by calculating phase differences between the two TSs. To achieve the image frequency interference factors which are caused by an IQ imbalance effect, each symbol of the TS is modulated on only a half of all the subcarriers while the remaining subcarriers are modulated with zeros. By doing this, the IQ imbalance coefficientsare directly estimated without recursive calculation requirements. This brings a low complexity to implementation in hardware. The performances of the modeling system are evaluated by a numerical simulation method where the error vector magnitude (EVM), the bit error ratio (BER), and the mean square error (MSE) quantities are used as performance indicators. The numerical simulation results are showed that the performance of the modeling system is enormously improved even when highly dispersive channels and phase noise are considered.
offset (CFO) and phase noise (PN), respectively. This paper considers all of those offsets; however, only amplitude and phase imbalance in the presence of CFO are focused under chromatic dispersion and PN. In practical, the imperfection of modulator and demodulator equipment generates in-phase (I) and quadrature (Q) imbalance (IQ imbalance). However, the impact can be characterized into two main issues. One is the phase miss matched where the phase imbalance of the transmitter comes from the Mach-Zehnder modulator (MZM) when I and Q components do not have a phase difference of 90° for transmitting information signal. Second, the amplitude imbalance comes from imperfect of digital-to-analog conversion (DAC) or similar effects. Several methods and techniques had been proposed by many institutes. For example, a hybrid frequency-time domain compensation has been proposed by (3-4) for compensation of IQ imbalance channel effects. Transmitting two identical training symbol (TS) sequences are used to learn the IQ imbalance coefficients. However, this is very complex to implement in hardware at large fast Furrier transforms (FFT) and the TS can only estimate for IQ imbalance and not for CFO. Chung and et al. (5) had been presented a method for compensating quadrature imbalance by using the Gram-Schmidt orthogonalization procedure (GSOP). This shows the disadvantage that the minimum of 1000 received sampling signals are needed for averaging in I and Q, which leads to long delays and long calculations. Recently, Nguyen and et al. (6) had been proposed IQ imbalance compensation by estimating signal to noise ratio (SNR) of the received signal where the received phase is adjusted. If the tuning phase is matched, IQ imbalance is cancelled. Consequently, the SNR will be reached to peak. However, the range of the phase tuning is limited to only−π/4:π/4In this paper, an efficient low- complexity algorithm to estimate and compensate for IQ imbalance and CFO by using two training sequences (TS) is proposed. The TS is used to learn the coefficients of those two impacts. In addition, since the image frequency interference due to the IQ imbalance is appeared on the opposite frequency on each other; therefore, to determine the IQ imbalance coefficient and CFO, each one modulates on only one half of the subcarriers. The rest is modulated by zeros. Obviously, the proposed method is calculated faster than the GSOP and can be estimated for both IQ imbalance and CFO because the data which is used to estimate for that is only a half of FFT size. The TS is followed by many OFDM symbols, thereby completing the OFDM frame as shown in Figure 1.