Complexity-Performance Trade-Offs in a Single Carrier Transmission with Iterative Equalization

In this paper, we study the performance of our single carrier receiver model using turbo coding and regular low density parity check (LDPC) coding. The single carrier receiver utilizes filter bank based frequency domain equalizer, which is then followed by a low complexity, iterative noise predictive decision feedback equalizer (NP-DFE). Furthermore, the average computation rate as a function of E_b/N₀ ratio is observed for different coding methods, thus characterizing the performance vs. complexity trade-offs in the considered receiver structure in terms of the channel coding. Clear performance benefits from the NP-DFE loop can be achieved and the averaged complexity increase in our system is tolerable in the E_b/N₀ range where frame error rates below 10^-1 are achieved. In addition, by fixing the maximum decoding complexity to be the same between turbo and LDPC codes, the performance of these codes is similar but the averaged complexity of the LDPC code saturates to a clearly lower level.