Article Preview
TopIntroduction
Modern communication systems are expected to support highest data rates over transmission channels that are selective in time and frequency. Cyclic Prefix Orthogonal Frequency Division Multiplexing (CP-OFDM) is considered one of the most important schemes of multi carrier systems that is adopted for the 4-th generation (4G) systems and used in the form of WiMAX and LTE standards. CP-OFDM may not represent an optimal solution in the future systems due to some drawbacks such as: low spectral efficiency due to the Cyclic Prefix (CP) (Qinwei & Schmeink, 2015), high sensitivity to frequency offset and high out-of-band emission due to poor frequency localization and high side-lobes level for the rectangular prototype filter (Vahlin & Holte, 1996). One of the suggested solutions that will be considered in 5G is Filter Bank Multicarrier with Offset Quadrature Amplitude Modulation (FBMC/OQAM) (Xu, Zhang, Lu, Wang, Guidotti, & Chang, 2016). It offers many advantages that allow overcoming CP-OFDM drawbacks, it does not need any CP and enables better spectrum usage and mobility support since FBMC allows the use of well localized time-frequency prototype filters such as PHYDYAS prototype filter (Siohan & Roche, 2000), and this ensures that every subcarrier overlaps only with adjacent subcarriers, the well stopband attenuation of these filters can reduce the out of band emission and sensitivity to frequency(Siohan, Siclet, & Lacaille, 2002).
The problem in FBMC systems is the complexity of the equalizer techniques, which considered an open issue and high complexity compared with these in CP-OFDM since CP is not used to reduce ISI, so one tap equalizer per subchannel is not sufficient for high frequency selective channels (Marijanovic, Schwarz, & Rupp, 2016). For equalization problem an efficient linear Minimum Mean Square Error (MMSE) per-subcarrier equalizer has been proposed in (Viholainen, Bellanger, & Huchard, 2009), which takes into account intercarrier interference (ICI) coming from adjacent subchannels. The performance of this equalizer has been enhanced in (Ikhlef & Louveaux, 2009) by adding second stage which takes into account the decisions on the received symbols at the output of the first stage to remove the term corresponding to ICI from the received signal for each subchannel.
There is some research to improve the performance of FBMC/OQAM systems as in (Aoude, Vallet, & Nedic, 2012), which an iterative interference cancellation has been proposed for the bit-interleaved coded OFDM/OQAM system. This approach used a Soft Input–Soft Output (SISO) decoder which is put in Turbo-like mode. The iterative algorithms are based on soft successive interference cancellation using message passing on bipartite graphs. However, the soft interference cancellation is performed by removing the causality constraint which leads to a large delay in practical implementations. In (Koslowski & Jondral, 2014) the steps towards a turbo equalizer enabled FBMC receiver have been outlined. However, no practical implementation has been proposed.
All the existed solutions are not sufficient and cannot provide an effective solution to the problem of performance degradation in FBMC systems in high frequency selective channels, and as we will see in this paper, this problem is mainly coming from the adjacent subchannels and symbols interference, so the iterative equalizing is the candidate solution to reduce these interferences.
As a result, we can say that the weak point in FBMC systems is an equalizer, so we tend to study this problem in our research by using turbo equalize. Turbo equalizer is existed and used in OFDM receiver, where every two successive symbols are separated by CP, so the turbo processing can be applied on each symbol separately. the same processing cannot be directly used with FBMC, due to each symbol expands into many adjacent symbols and the interference in each subchannel comes from the adjacent subchannels, so real-time turbo processing for FBMC is studied in this paper, and the design of turbo method is reconstructed to be relevance to FBMC/OQAM systems.