logo

SCIENCE CHINA Information Sciences, Volume 61, Issue 2: 022304(2018) https://doi.org/10.1007/s11432-016-9053-3

Partial intersection sphere decoding with weighted voting for sparse rotated V-OFDM systems

More info
  • ReceivedSep 18, 2016
  • AcceptedFeb 15, 2017
  • PublishedAug 24, 2017

Abstract

Vector orthogonal frequency division multiplexing (V-OFDM) is a general system that builds a bridge between OFDM and single-carrier frequency domain equalization in terms of intersymbol interference level and receiver complexity. In this paper, we focus on a rotated V-OFDM system over a broadband sparse channel owing to its large time delay spread and fewer nonzero taps. In order to collect the multipath diversity, a simple rotation matrix is designed, independent of the number of subchannels at the transmitter. For the rotated V-OFDM receiver, a partial intersection sphere decoding with weighted voting method is proposed by exploiting the sparse nature of the multipath channel. The proposed receiver chooses the transmitted vector from the set with the maximum likelihood estimation generated using the partial intersection sphere decoding method. For an extreme case, such as when a candidate set is empty, which usually occurs at a low signal-to-noise ratio (SNR), an efficient weighted voting system is used to estimate the transmitted vectors symbol-by-symbol. Simulation results indicate that the proposed receiver improves the symbol error rate performance with reduced complexity, especially for low SNR scenarios.


Acknowledgment

This work was supported by National Natural Science Foundation of China (Grant No. 61401194), China Scholarship Council (Grant No. 201406190083), and Program B for Outstanding Ph.D. Candidate of Nanjing University (Grant No. 201501B013).


References

[1] Cimini L J. Analysis and simulation of a digital mobile channel using orthogonal frequency division multiplexing. IEEE Trans Commun, 1985, 33: 665--675. Google Scholar

[2] Astely D, Dahlman E, Furuskar A, et al. LTE: the evolution of mobile broadband. IEEE Commun Mag, 2009, 47: 44--51. Google Scholar

[3] Hwang T, Yang C, Wu G, et al. OFDM and its wireless applications: a survey. IEEE Trans Veh Tech, 2009, 58: 1673--1694. Google Scholar

[4] Daoud O. Performance improvement of wavelet packet transform over fast Fourier transform in multiple-input multiple-output orthogonal frequency division multiplexing systems. IET Commun, 2012, 6: 765--773. Google Scholar

[5] Zhang T X, Xia X-G, Kong L J. CP-based MIMO OFDM radar IRCI free range reconstruction using real orthogonal designs. Sci China Inf Sci, 2017, 60: 022301. Google Scholar

[6] Wang N, Blostein S D. Comparison of CP-based single carrier and OFDM with power allocation. IEEE Trans Commun, 2005, 53: 391--394. Google Scholar

[7] Tajer A, Nosratinia A. Diversity order in ISI channels with single-carrier frequency-domain equalizers. IEEE Trans Wirel Commun, 2010, 9: 1022--1032. Google Scholar

[8] Souto N, Dinis R, Silva J C. Impact of channel estimation errors on SC-FDE systems. IEEE Trans Commun, 2014, 62: 1530--1540. Google Scholar

[9] Falconer D, Ariyavisitakul S L, Benyamin-Seeyar A, et al. Frequency domain equalization for single-carrier broadband wireless systems. IEEE Commun Mag, 2002, 40: 58--66. Google Scholar

[10] Pancaldi F, Vitetta G, Kalbasi R, et al. Single-carrier frequency-domain equalization. IEEE Signal Process Mag, 2008, 25: 37--56. Google Scholar

[11] Xia X-G. Precoded and vector OFDM robust to channel spectral nulls and with reduced cyclic prefix length in single transmit antenna systems. IEEE Trans Commun, 2001, 25: 1363--1374. Google Scholar

[12] Zhang H, Xia X-G, Zhang Q, et al. Precoded OFDM with adaptive vector channel allocation for scalable video transmission over frequency selective fading channels. IEEE Trans Mobile Comput, 2002, 1: 132--141. Google Scholar

[13] Zhang H, Xia X-G, Cimini L J, et al. Synchronization techniques and guard-band-configuration scheme for single-antenna vector-OFDM systems. IEEE Trans Wirel Commun, 2005, 4: 2454--2464. Google Scholar

[14] Zhang H, Xia X-G. Iterative decoding and demodulation for single-antenna vector OFDM systems. IEEE Trans Veh Tech, 2005, 55: 1447--1454. Google Scholar

[15] Zhou W, Fan L, Chen H. Vector orthogonal frequency division multiplexing system over fast fading channels. IET Commun, 2014, 8: 2322--2335. Google Scholar

[16] Ngebani I, Li Y, Xia X-G, et al. Analysis and compensation of phase noise in vector OFDM systems. IEEE Trans Signal Process, 2014, 62: 6143--6157. Google Scholar

[17] Ngebani I, Li Y, Xia X-G. EM-based phase noise estimation in vector OFDM systems using linear MMSE receivers. IEEE Trans Veh Tech, 2016, 65: 110--122. Google Scholar

[18] Han C, Hashimoto T, Suehiro N. Constellation-rotated vector OFDM and its performance analysis over Rayleigh fading channels. IEEE Trans Commun, 2010, 58: 828--838. Google Scholar

[19] Han C, Hashimoto T. Tight PEP lower bound for constellation-rotated vector-OFDM under carrier frequency offset and fast fading. IEEE Trans Commun, 2014, 62: 1931--1943. Google Scholar

[20] Cheng P, Tao M, Xiao Y, et al. V-OFDM: on performance limits over multi-path Rayleigh fading channels. IEEE Trans Commun, 2011, 59: 1878--1892. Google Scholar

[21] Li Y, Ngebani I, Xia X-G. On performance of vector OFDM with linear receivers. IEEE Trans Signal Process, 2012, 60: 5268--5280. Google Scholar

[22] Kannu A P, Schniter P. On communication over unknown sparse frequency-selective block-fading channels. IEEE Trans Inf Theory, 2011, 57: 6619--6632. Google Scholar

[23] Salous S, Degli Esposti V, Fuschini F, et al. Millimeter-wave propagation. IEEE Trans Inf Theory, 2016, 58: 115--127. Google Scholar

[24] Wang G, Liao H, Wang H. Systematic and optimal cyclotomic lattices and diagonal space-time block code designs. IEEE Trans Inf Theory, 2004, 50: 3348--3360. Google Scholar

[25] Agrell E, Eriksson A, Vardy A, et al. Closest point search in lattices. IEEE Trans Inf Theory, 2002, 48: 2201--2214. Google Scholar

[26] Studer C, Burg A, Bolcskei. Soft-output sphere decoding: algorithms and VLSI implementation. IEEE J Sel Areas Commun, 2008, 26: 290--300. Google Scholar

[27] Hassibi B, Vikalo H. On the sphere decoding algorithm I. Expected complexity. IEEE Trans Signal Process, 2005, 53: 2806--2818. Google Scholar

[28] Vikalo H, Hassibi B. On the sphere decoding algorithm II. Generalizations, second-order statistics, and applications to communications. IEEE Trans Signal Process, 2005, 53: 2819--2834. Google Scholar

[29] Jalden J, Ottersten B. On the complexity of sphere decoding in digital communications. IEEE Trans Signal Process, 2005, 53: 1474--1484. Google Scholar

[30] Seethaler D, Jalden J, Studer C, et al. On the complexity distribution of sphere decoding. IEEE Trans Inf Theory, 2011, 57: 5754--5768. Google Scholar

  • Figure 1

    The block diagram of a rotated V-OFDM modulation system.

  • Figure 2

    (Color online) Example of the updating process of the partial intersection sphere decoding method.

  • Figure 3

    (Color online) Example of the weighted voting system for strong noise scenarios.

  • Figure 4

    (Color online) Comparison of probability mass function of $\kappa$ with $K=16$ and $M=8,16,32,64$, respectively.

  • Figure 5

    (Color online) Comparison of probability mass function of $\kappa$ with $M=16$ and $K=2,4,8,16$, respectively.

  • Figure 6

    Time flow of PIS-WV decoding for rotated V-OFDM system. The $l$th subchannel is taken as an example.

  • Figure 7

    (Color online) Comparison of V-OFDM system with different rotation angles.

  • Figure 8

    (Color online) Diversity orders of PIS-WV decoding for power delay profile.

  • Figure 9

    (Color online) Symbol error rates of different receivers over sparse rotated V-OFDM system.

  • Figure 10

    (Color online) Decoding complexities of different receivers over sparse rotated V-OFDM system.

  •   

    Algorithm 1 Partial intersection sphere decoding with weighted voting

    Require:$\boldsymbol~Y,~\boldsymbol~h,~L,~M,~P,~r$

    Output:$\widehat{\boldsymbol~X}$

    Initialization: $\mathcal~U^0\gets\varnothing,~\mathcal~X^0\gets\varnothing,~\boldsymbol\beta\gets\boldsymbol0$;

    for $l\gets0,1,\ldots,L-1$

    for $m\gets0,1,\ldots,M-1$

    $Y_{lM+m}$ is the $m$th entry of column vector $\boldsymbol~Y_l$;

    $\boldsymbol{\mathcal{H}}_l^m$ is the $1\times\kappa$ vector aligned as the $m$th row and the $(m-i_0)\boldsymbolod~M$th, $(m-i_1)\boldsymbolod~M$th$,\ldots,(m-i_{\kappa-1})\boldsymbolod~M$th columns of $\boldsymbol{\mathcal{H}}_l$;

    Generate a set of $\kappa$-dimensional symbol sequences: $\mathcal~S^m\gets\{\boldsymbol~S\Big|\boldsymbol~S\in\mathbb~X^{\kappa},~\big|Y_{lM+m}-\boldsymbol{\mathcal{H}}_l^m\boldsymbol~S\big|\leqslant~r\}$;

    Generate a set of current coordinates of nonzero entries: $\mathcal~V^m\gets\left\{v\big|\forall~i\in\mathcal~I,~v\gets(m-i)\boldsymbolod~M\right\}$;

    Generate a intersection of the existing and current nonzero coordinates sets: $\mathcal~W^m\gets\mathcal~U^m\bigcap\mathcal~V^m$;

    Construct an injective mapping of coordinates: $f:~\ell\rightarrow(m-i_\ell)\boldsymbolod~M,~\ell\in\{0,1,\dots,\kappa-1\}$;

    for all $\boldsymbol~S\in\mathcal~S^m$

    Generate a set of symbol sequences for a given current $\boldsymbol~S\in\mathcal~S^m$:$\mathcal~X_{\boldsymbol~S}^{m+1}\gets\big\{\boldsymbol~X^{m+1}\big|\boldsymbol~X^m\in\mathcal~X^m,~\forall~w\in\mathcal~W^m,~X_w^{m+1}=S_{f^{-1}(w)};~\forall~\nu\in\complement_{\mathcal~V^m}\mathcal~W^m,~X_\nu^{m+1}\gets~S_{f^{-1}(\nu)}\big\}$;

    The candidate receives one vote: $\beta_{lM+v}^m(S_{f^{-1}(v)})\gets\beta_{lM+v}^m(S_{f^{-1}(v)})+1,~v\in\mathcal~V^m$;

    end for

    Generate a set of entire symbol sequences for all $\boldsymbol~S\in\mathcal~S^m$: $\mathcal~X^{m+1}\gets\bigcup_{\boldsymbol~S\in\mathcal~S^m}\mathcal~X_{\boldsymbol~S}^{m+1}$;

    Update the existing nonzero coordinates set: $\mathcal~U^{m+1}\gets\mathcal~U^m\bigcup\mathcal~V^m$;

    end for

    if $\mathcal~X^M=\varnothing$ then

    Weighted voting decision symbol-by-symbol: $\widehat~X_{lM+k}\gets\mathop{\arg\max}_{X\in\mathbb~X}\sum_{m=0}^{M-1}\frac{\beta_{lM+k}^m(X)}{\left|\mathcal~S^m\right|},~k\in\{0,1,\ldots,\linebreak~M-1\}$;

    else

    Maximum likelihood estimation from candidate set: $\widehat{\boldsymbol~X}_l\gets\mathop{\arg\min}_{\boldsymbol~X_l^M\in\mathcal~X^M}\big\|\boldsymbol~Y_l-\boldsymbol{\mathcal{H}}_l\boldsymbol~X_l^M\big\|_2$;

    end if

    end for

    return $\widehat{\boldsymbol~X}$.

Copyright 2019 Science China Press Co., Ltd. 《中国科学》杂志社有限责任公司 版权所有

京ICP备18024590号-1