De, Digveer (2017-07). Iterative Demodulation and Decoding for LDPC Coded Generalized Frequency Division Multiplexing. Master's Thesis. Thesis uri icon


  • Currently, there is a standardization process underway to design the fifth generation of wireless systems or 5G wireless systems. The ambitious targets set forth for 5G wireless systems call for novel approaches in all layers of the network. At the physical layer (PHY), Orthogonal Frequency Division Multiplexing (OFDM) has become a de facto standard for wireless systems such as 4G cellular and IEEE 802.11 (Wi-Fi) systems. However, the large peak to average power ratio of OFDM signals makes OFDM an unattractive candidate for some services envisioned in 5G systems, particularly in the uplink. Recently, Generalized Frequency Division Multiplexing (GFDM), which is a member of the non-orthogonal multiple access technologies has been proposed as the modulation scheme for 5G wireless systems. GFDM has some advantages over OFDM, such as looser requirements on synchronization, a lower PAPR requirement,as well as a lower out-of-band spectral leakage. However, in GFDM the sub-channels are not orthogonal which results in inter-carrier interference and, hence, an increased uncoded bit error rate. While iterative receivers have been proposed for improving the bit error rate performance of uncoded GFDM, there are very few works that have studied the performance of coded GFDM systems. In this thesis, we investigate the performance of coded systems with GFDM. Using earlier results on soft interference cancellation based turbo equalization and turbo multi-user detection, we design an iterative receiver for GFDM with low density parity check codes. We show that the receiver is able to successfully combat the non-orthogonality of sub-channels in GFDM and provide performance similar to that of coded OFDM systems at an increased receiver complexity.

publication date

  • July 2017