Creating and sharing knowledge for telecommunications

A new hybrid multicarrier transmission technique with iterative frequency domain detection

Simões, T. G. ; Silva, A. P. S. ; Gomes, M. ; Silva, V. ; Dinis, R.

Physical Communication-Elsevier Vol. 27, Nº -, pp. 7 - 16, April, 2018.

ISSN (print): 1874-4907
ISSN (online): 1874-4907

Journal Impact Factor: 0,802 (in 2015)

Digital Object Identifier: 10.1016/j.phycom.2017.12.014

The growing progress in wireless communication services led to a demand in high data rates, spectral efficiency and flexibility requirements. The Block-Windowed Burst Orthogonal Frequency Division Multiplexing (BWB-OFDM) technique has been recently proposed to face these demands. This technique employs smoother, non-rectangular windows, allowing a power spectral density similar to the filtered OFDM approach, thus achieving high spectral efficiency; also, it packs together several OFDM symbols, with the addition of a sole zero-padding to accommodate the multipath channel's propagation delay, thereby improving power efficiency.
However, BWB-OFDM has the same drawbacks of OFDM when transmitting over hostile channel conditions, namely the performance degradation due to deep fades associated to severe frequency-selective channels.

This paper proposes a new Time Interleaved BWB-OFDM (TIBWB-OFDM) technique that performs interleaving on the time-samples of each BWB-OFDM block, creating a kind of diversity effect at the frequency domain, granting a much better resilience against deep inband fades, while keeping all the mentioned advantages of BWB-OFDM at the cost of no added complexity.
Also, by regarding TIBWB-OFDM as a hybrid technique combining single-carrier and multicarrier characteristics, this paper also proposes the use of non-linear frequency domain equalizers based on the Iterative Block Frequency Domain Equalization (IB-DFE) concept for TIBWB-OFDM detection. It is shown that noteworthy improvements can be achieved in bit error rate (BER) performance compared to conventional OFDM schemes when employing typical zero-forcing (ZF) and minimum mean-square error (MMSE) linear equalizers.