Near Capacity RCQD Constellations for PAPR Reduction of OFDM Systems

Abstract

We investigate an optimized blind SeLected Mapping (SLM) algorithm to reduce the Peak-to-Average Power Ratio (PAPR) for Orthogonal Frequency Division Multiplexing (OFDM) systems with Signal Space Diversity (SSD). Several phase sequences based on two Rotated and Cyclically Q-Delayed (RCQD) constellations are used at the transmitter to considerably reduce the PAPR. The pair of RCQD constellations is selected so as to achieve the highest average mutual information for both the Coded Modulation (CM) and the Bit-Interleaved Coded Modulation (BICM) schemes. These constellations also allow, along with Hadamarad sequences, good blind detection performances with no spectral spoilage. Simulations totally confirm our theoretical results.

Low-Complexity Blind PAPR Reduction for OFDM Systems With Rotated Constellations

“…At the transmitter side, note that no multiplication by any phase sequence is required for our proposal; this reduces the complexity burden by 4N (D − 1) RMs and 2N (D − 1) RSs compared to the SLM algorithm in [5], [6]. Furthermore, at the receiver side, the total computational complexity of the proposed MMSE-based blind detector (see equations ( 15)-( 17)) is 38N D RMs, 15DN RSs, (2log 2 M − 2)DN RCs, and DN RIs.…”

“…Besides, when we jointly use the TR and the proposed algorithms with D = 8, we obtain similar PAPR reduction performance to the optimal PTS with much lower computational complexity and higher spectral efficiency. Finally, when compared to the blind SLM method [5], [6], both methods have approximately the same PAPR reduction performance; however, our proposal for UP-RCQD constellations reduces considerably the computational complexity both at the transmitter and the receiver sides; in particular, in addition to the computational complexity reduction obtained for the LLR computation thanks to the use of the UP-RCQD constellation [9], the computational complexity is reduced, with the chosen system parameters, by 97% (see Table I and Fig. 6).…”

“…Indeed, thanks to an inherent increased modulation diversity (Signal Space Diversity (SSD)), RCQD signals allow a better system performance over fading channels compared to conventional QAM signals [2]. To construct the two-dimensional (2D) RCQD signal, one first needs to rotate the conventional signals with a proper rotation angle and then to interleave the Q component, so as to ensure that the I and Q components of a given symbol experience independent Tarak Arbi, Zi Ye and Benoit Geller are with Department U2IS of ENSTA-Paris, IP Paris, 828 Boulevard des Marechaux, 91120, Palaiseau, France (email: tarak.arbi,zi.ye,benoit.geller@ensta-paris.fr); fading channels [2]- [6]. At the receiver side, each component brings all the information carried by the original 2D symbol, which reduces the effect of fading.…”

“…To specifically reduce the PAPR of OFDM symbols when RCQD signals are used, the DVB-T2 standard recommends to only use a single PAPR reduction method (the TR method); indeed, contrarily to the case of conventional QAM signals, the ACE method and the TR method cannot be deployed together to further reduce the PAPR [1]. Therefore, [5], [6] propose a blind SLM technique with two RCQD constellations; the couple of rotation angles involved in this technique is obtained by jointly optimizing the theoretical system performance and the blind decoding performance. However, this method does not consider neither the high demapping complexity of RCQD signals, nor the high complexity introduced by the blind decoding.…”

Window-based UFMC technique for 5G systems

“…The measured EVM rms measurements using either the average constellation power or the peak constellation power method as Calculated using the following equations [16,17].…”