MIMO-DFE(Multiple-Input-Multiple-Output Decision Feedback Equalizer) based receiver architectures are researched recently to detect signals in BLAST(Bell laboratories LAyered Space-Time) over frequency-selective chann...MIMO-DFE(Multiple-Input-Multiple-Output Decision Feedback Equalizer) based receiver architectures are researched recently to detect signals in BLAST(Bell laboratories LAyered Space-Time) over frequency-selective channels. Due to their recursive structure, these receivers may suffer from error propagation which results in an overall mean square error degradation. An MIMO-DFE based BLAST receiver with limited error propagation to combat frequencyselective channel is proposed, which employs both norm constraint on feedback filter taps and soft decision device. Simulation results show that the proposed receiver outperforms conventional ones in various frequency selective channels.展开更多
This paper evaluates the effect of decision feedback equalizer( DFE) error propagation for400 Gb/s Ethernet( 400 GbE) electrical link in order to propose some effective methods to improve bit error rate( BER). First,a...This paper evaluates the effect of decision feedback equalizer( DFE) error propagation for400 Gb/s Ethernet( 400 GbE) electrical link in order to propose some effective methods to improve bit error rate( BER). First,an analytical model for DFE burst error length distribution is proposed and simulated based on a NRZ electrical link in which a 5-tap DFE combined with a multiple-tap feed forward equalizer( FFE) is included. Then,a detailed derivation for BER considering DFE error propagation is given based on the distribution of burst error run length and the BER performance with and without forward error correction( FEC) is simulated too. After that,this paper investigates several MUX-based FEC interleaving methods including their complexity and latency in order to improve BER further. At last,three FEC interleaving schemes are compared not only in interleaving gain,but also in hardware complexities and latencies. Simulation results show that pre-interleave bit muxing can obtain good tradeoff between BER and complexity for 400 Gb E electrical link.展开更多
针对正交时频空(Orthogonal Time Frequency Space, OTFS)调制系统中均衡器性能不佳及线性滤波器复杂度较高等问题,提出了一种LU(Lower-Upper)分解与迭代最小均方误差(Iterative Minimum Mean Square Error, IMMSE)均衡器结合的OTFS系...针对正交时频空(Orthogonal Time Frequency Space, OTFS)调制系统中均衡器性能不佳及线性滤波器复杂度较高等问题,提出了一种LU(Lower-Upper)分解与迭代最小均方误差(Iterative Minimum Mean Square Error, IMMSE)均衡器结合的OTFS系统信号检测算法(LU-IMMSE)。该算法依据时延多普勒域稀疏信道矩阵的特征,采用一种低复杂度的LU分解方法,以避免MMSE均衡器求解矩阵逆的过程,在保证均衡器性能的前提下降低了均衡器复杂度。在OTFS系统中引入一种IMMSE均衡器,通过不断迭代更新发送符号均值和方差这些先验信息来逼近MMSE均衡器最优估计值。LU-IMMSE算法通过调节迭代次数可以有效降低误比特率。在比特信噪比为8 dB时,5次迭代后的LU-IMMSE均衡器误比特率相比传统的MMSE均衡器降低了约11 dB。随着迭代次数的增大,较传统IMMSE算法降低了计算复杂度。在最大时延系数为4、符号数为16的情况下,与直接求逆相比,所提出的低复杂度LU分解方法降低了约91.72%的矩阵求逆计算复杂度。展开更多
We propose a multi-tap look-up-table(LUT)-based error-tracking decision feedback equalizer with partial unrolling(ET-DFEPU)algorithm to improve log-likelihood ratio(LLR)quality with error propagation.We introduce an e...We propose a multi-tap look-up-table(LUT)-based error-tracking decision feedback equalizer with partial unrolling(ET-DFEPU)algorithm to improve log-likelihood ratio(LLR)quality with error propagation.We introduce an error-tracking model to obtain the probabilities of different error states,which are input into the decoder to obtain more accurate LLRs.We also use LUTs to reduce the computation complexity of ET-DFE-PU.Moreover,we adopt a low-complexity partially unrolled architecture to relax the feedback timing constraint.We conduct experiments to transmit 170-Gb/s 4-ary pulse amplitude modulation signal in intensity modulation and direct-detection system at C-band.The results show that the proposed method can achieve 3-d B receiver sensitivity improvement at the same post-forward error correction bit error ratio compared with the conventional DFE.展开更多
The reasonable prior information between the parameters in the adjustment processing can significantly improve the precision of the parameter solution. Based on the principle of equality constraints, we establish the ...The reasonable prior information between the parameters in the adjustment processing can significantly improve the precision of the parameter solution. Based on the principle of equality constraints, we establish the mixed additive and multiplicative random error model with equality constraints and derive the weighted least squares iterative solution of the model. In addition, aiming at the ill-posed problem of the coefficient matrix, we also propose the ridge estimation iterative solution of ill-posed mixed additive and multiplicative random error model with equality constraints based on the principle of ridge estimation method and derive the U-curve method to determine the ridge parameter. The experimental results show that the weighted least squares iterative solution can obtain more reasonable parameter estimation and precision information than existing solutions, verifying the feasibility of applying the equality constraints to the mixed additive and multiplicative random error model. Furthermore, the ridge estimation iterative solution can obtain more accurate parameter estimation and precision information than the weighted least squares iterative solution.展开更多
文摘MIMO-DFE(Multiple-Input-Multiple-Output Decision Feedback Equalizer) based receiver architectures are researched recently to detect signals in BLAST(Bell laboratories LAyered Space-Time) over frequency-selective channels. Due to their recursive structure, these receivers may suffer from error propagation which results in an overall mean square error degradation. An MIMO-DFE based BLAST receiver with limited error propagation to combat frequencyselective channel is proposed, which employs both norm constraint on feedback filter taps and soft decision device. Simulation results show that the proposed receiver outperforms conventional ones in various frequency selective channels.
基金Supported by the National Natural Science Foundation of China(No.61471119)
文摘This paper evaluates the effect of decision feedback equalizer( DFE) error propagation for400 Gb/s Ethernet( 400 GbE) electrical link in order to propose some effective methods to improve bit error rate( BER). First,an analytical model for DFE burst error length distribution is proposed and simulated based on a NRZ electrical link in which a 5-tap DFE combined with a multiple-tap feed forward equalizer( FFE) is included. Then,a detailed derivation for BER considering DFE error propagation is given based on the distribution of burst error run length and the BER performance with and without forward error correction( FEC) is simulated too. After that,this paper investigates several MUX-based FEC interleaving methods including their complexity and latency in order to improve BER further. At last,three FEC interleaving schemes are compared not only in interleaving gain,but also in hardware complexities and latencies. Simulation results show that pre-interleave bit muxing can obtain good tradeoff between BER and complexity for 400 Gb E electrical link.
文摘针对正交时频空(Orthogonal Time Frequency Space, OTFS)调制系统中均衡器性能不佳及线性滤波器复杂度较高等问题,提出了一种LU(Lower-Upper)分解与迭代最小均方误差(Iterative Minimum Mean Square Error, IMMSE)均衡器结合的OTFS系统信号检测算法(LU-IMMSE)。该算法依据时延多普勒域稀疏信道矩阵的特征,采用一种低复杂度的LU分解方法,以避免MMSE均衡器求解矩阵逆的过程,在保证均衡器性能的前提下降低了均衡器复杂度。在OTFS系统中引入一种IMMSE均衡器,通过不断迭代更新发送符号均值和方差这些先验信息来逼近MMSE均衡器最优估计值。LU-IMMSE算法通过调节迭代次数可以有效降低误比特率。在比特信噪比为8 dB时,5次迭代后的LU-IMMSE均衡器误比特率相比传统的MMSE均衡器降低了约11 dB。随着迭代次数的增大,较传统IMMSE算法降低了计算复杂度。在最大时延系数为4、符号数为16的情况下,与直接求逆相比,所提出的低复杂度LU分解方法降低了约91.72%的矩阵求逆计算复杂度。
基金supported by the National Natural Science Foundation of China(Nos.62301128 and U22A2086)the STCSM(No.SKLSFO2021-01)+1 种基金the Open Fund of IPOC(BUPT)(No.IPOC2020A011)the Fundamental Research Funds for the Central Universities(Nos.ZYGX2020ZB043 and ZYGX2019J008)。
文摘We propose a multi-tap look-up-table(LUT)-based error-tracking decision feedback equalizer with partial unrolling(ET-DFEPU)algorithm to improve log-likelihood ratio(LLR)quality with error propagation.We introduce an error-tracking model to obtain the probabilities of different error states,which are input into the decoder to obtain more accurate LLRs.We also use LUTs to reduce the computation complexity of ET-DFE-PU.Moreover,we adopt a low-complexity partially unrolled architecture to relax the feedback timing constraint.We conduct experiments to transmit 170-Gb/s 4-ary pulse amplitude modulation signal in intensity modulation and direct-detection system at C-band.The results show that the proposed method can achieve 3-d B receiver sensitivity improvement at the same post-forward error correction bit error ratio compared with the conventional DFE.
基金supported by the National Natural Science Foundation of China,Grant Nos.42174011,41874001 and 41664001Innovation Found Designated for Graduate Students of ECUT,Grant No.DHYC-202020。
文摘The reasonable prior information between the parameters in the adjustment processing can significantly improve the precision of the parameter solution. Based on the principle of equality constraints, we establish the mixed additive and multiplicative random error model with equality constraints and derive the weighted least squares iterative solution of the model. In addition, aiming at the ill-posed problem of the coefficient matrix, we also propose the ridge estimation iterative solution of ill-posed mixed additive and multiplicative random error model with equality constraints based on the principle of ridge estimation method and derive the U-curve method to determine the ridge parameter. The experimental results show that the weighted least squares iterative solution can obtain more reasonable parameter estimation and precision information than existing solutions, verifying the feasibility of applying the equality constraints to the mixed additive and multiplicative random error model. Furthermore, the ridge estimation iterative solution can obtain more accurate parameter estimation and precision information than the weighted least squares iterative solution.
