Search

Faster-than-Nyquist (FTN) transmission is a promising technique to increase the spectral efficiency at a fixed constellation size. However, traditional timing synchronization algorithms mostly rely on the eye diagram opening after matched-filtering (e.g., Gardner and zero-crossing). Such approaches are, thus, unsuitable in the presence of the FTN-induced intersymbol interference. In this letter, we first show that FTN signals exhibit cyclostationarity at the $l$ th order ($l > 2$) according to their bandwidth and symbol period. Then, we derive non-data-aided timing offset estimators from the cyclic temporal cumulant function. Besides its relevance for strong FTN scenarios, the proposed solution is also robust to frequency and phase offsets. [ABSTRACT FROM AUTHOR]

As optical transceivers evolve toward 200G per lane, the comparatively low bandwidth and sampling speed of the commercially available high-speed digital-to-analog converters (DACs) and analog-to-digital converters (ADCs) increasingly become the dominating obstacle toward success. On the other hand, DACs and ADCs are essential to make use of digital signal processing for impairment compensation and error correction; altogether driving systems’ engineers to a hard design dilemma. With the purpose of alleviating this compromise, we propose a transceiver structure—encompassing both hardware and software—enabling the generation, transmission, and detection of strongly over-filtered data signals whose baudrate exceeds the sampling rates of (potentially) both the DAC and ADC. We call this approach sub-baudrate sampling, a technique which does not increase the processing complexity at the transmitter side and does not even require the information about the channel state for pre-distortion or pre-coding purposes. We experimentally assess the performance of sub-baudrate sampled signals in two blocks of experiments: first the proof of concept in back-to-back configuration, including up to 112-Gbaud on–off keying at 0.785 samples per symbol in ∼25-GHz aggregate bandwidth. Then the C-band transmission demonstration, showing up to 100-Gbaud quaternary pulse-amplitude modulation (0.92 samples per symbol) and 125-Gbaud on–off keying (0.736 samples per symbol) over 1 and 80-km of single-mode fiber, respectively, with more than 3-dB receiver sensitivity margin at the 7% overhead hard-decision bit error rate limit. [ABSTRACT FROM AUTHOR]

In this paper, a pulse-overlapping super-Nyquist (Pol-SN) wavelength-division multiplexing (WDM) scheme is proposed to improve spectral efficiency of coherent optical transmission. In this scheme, two tributaries of polarization-division multiplexing quadrature phase-shift keying (PDM-QPSK) signals at the transmitter, carried by the same wavelength, are grouped together to allow for pulse overlapping in the time domain. Then, several wavelength channels are packed tightly in the frequency domain to form a super-Nyquist WDM system. At the receiver, three partial-response detection schemes are investigated and compared: 1) constant modulus algorithm (CMA) + duobinary shaping + maximum likelihood sequence estimation (MLSE); 2) duobinary shaping + multimodulus algorithm (MMA) + MLSE; and 3) channel shortening filter. Among these schemes, the easy-to-implement duobinary shaping +MMA + MLSE is selected by virtue of its best performance and lowest complexity. A 224-Gb/s Pol-SN PDM-QPSK system is investigated through numerical simulations. The simulation results show that the channel spacing of the 224 Gb/s Pol-SN PDM-QPSK system can be reduced to 20 GHz (11.2 bit/s/Hz SE) with 7% overhead hard-decision forward-error correction (HD-FEC, 3.8 × 10−3). At 26-GHz channel spacing, Pol-SN PDM-QPSK signals show 5.5-dB OSNR improvement, as compared with the signals of 224-Gb/s super-Nyquist PDM-16QAM. The simulation results also show that 224-Gb/s Pol-SN PDM-QPSK system, with 20-GHz spacing, can transmit up to 1000-km SSMF with 7% overhead HD-FEC. [ABSTRACT FROM AUTHOR]

In this letter, we investigate the lattice staggered multicarrier faster-than-Nyquist (MFTN) signaling. Specifically, we consider the time–frequency packing and optimal hexagonal lattice over additive white Gaussian noise channels. First, an efficient implementation of the lattice staggered MFTN based on the fast Fourier transform algorithm is proposed, and we show that the modulation and demodulation complexity could be substantially reduced. Furthermore, we consider, at the receiver, a low-complexity symbol-by-symbol detector. Our practical spectral efficiency and bit-error-rate performance investigation demonstrate that the MFTN with optimal hexagonal lattice outperforms the conventional rectangular lattice. [ABSTRACT FROM PUBLISHER]

In this letter, a novel maximizing signal-to-interference ratio (max-SIR) time–frequency packing scheme for multicarrier faster-than-Nyquist signaling is proposed. The proposed scheme is based on the low complexity symbol-by-symbol detector. Specifically, we optimize the time–frequency spacing under the given signaling efficiency to maximize the SIR, i.e., the ratio between the energy of useful signal and the intersymbol interference and intercarrier interference introduced by time–frequency packing. We show that the proposed max-SIR time–frequency packing outperforms other time–frequency packing schemes for both the symbol-by-symbol detection and coupled with successive interference cancellation. [ABSTRACT FROM PUBLISHER]

We consider channels affected by intersymbol interference with reduced-complexity, mutual information optimized, channel-shortening detection. For such settings, we optimize the transmit filter, taking into consideration the reduced receiver complexity constraint. As figure of merit, we consider the achievable information rate of the entire system and with functional analysis, we establish a general form of the optimal transmit filter, which can then be optimized by standard numerical methods. As a corollary to our main result, we obtain some insight of the behavior of the standard waterfilling algorithm for intersymbol interference channels. With only some minor changes, the general form we derive can be applied to multiple-input multiple-output channels with intersymbol interference. To illuminate the practical use of our results, we provide applications of our theoretical results by deriving the optimal shaping pulse of a linear modulation transmitted over a bandlimited additive white Gaussian noise channel which has possible applications in the faster-than-Nyquist/time packing technique. [ABSTRACT FROM PUBLISHER]

We propose new M-algorithm BCJR (M-BCJR) algorithms for low-complexity turbo equalization and apply them to severe intersymbol interference (ISI) introduced by faster than Nyquist signaling. These reduced-search detectors are evaluated in simple detection over the ISI channel and in iterative decoding of coded FTN transmissions. In the second case, accurate log likelihood ratios are essential and we introduce a 3-recursion M-BCJR that provides this. Focusing signal energy by a minimum phase conversion before the M-BCJR is also essential; we propose an improvement to this older idea. The new M-BCJRs are compared to reduced-trellis VA and BCJR benchmarks. The FTN signals carry 4—8 bits/Hz-s in a fixed spectrum, with severe ISI models as long as 32 taps. The combination of coded FTN and the reduced-complexity BCJR is an attractive narrowband coding method. [ABSTRACT FROM PUBLISHER]

The performance of turbo-like codes highly depends on their frame size and thus, the bit error rate (BER) performance of turbo-like codes can be improved by increasing the frame size. Unfortunately, increasing the frame size induces some drawbacks such as the increase of transmission and decoding latencies. On the other hand, faster than Nyquist (FTN) transmitter transmits modulated symbols at a rate higher than the Nyquist rate and thus, inter-symbol interference (ISI) is introduced. In this letter, we propose an FTN system with multiple turbo-like codes. In the proposed scheme, an information frame is divided into subframes and the subframes are encoded independently by multiple turbo-like encoders before being presented at an FTN modulator. Due to the ISI induced within the FTN modulator, the code frames generated by the multiple turbo-like encoders are combined and act as a larger code frame. Compared with conventional coded FTN schemes, the proposed scheme not only improves the BER performance but also reduces the decoding latency. [ABSTRACT FROM PUBLISHER]