Optimizing physical layer data transmission for minimal signal distortion

Phase-change memory (PCM) and resistive memory (RRAM) are promising alternatives to traditional memory technologies. However, both PCM and RRAM suffer from limited write endurance. Wear-leveling (WL) techniques are essential to extend the lifetime of these memories before experiencing endurance faults. Beyond the additional usage afforded by WL, row-sparing and focused error correction can extend the lifetime further after wear faults appear. Unfortunately, the need for extended WL techniques continues to become more pressing as scaling exacerbates process variation. Similarly, scaling causes challenges such as more severe noise and crosstalk to traditional DRAM.In this paper, we propose novel fault-aware WL schemes to allocate write frequencies according to the strength of the rows and handle the imbalance of writes in columns. We use runtime detection schemes to identify weak rows and protect them prior to wear out. In particular, row-level WL, aka RETROFIT, leverages the spare rows provided for redundancy to be used strategically to guard against early cell wear out. RETROFIT is compatible with error correction schemes that guarantee to mitigate hard faults and error-correcting codes (ECC). Rather than discard retired rows, when any spare row completely replaces a retired row, we retarget the retired row to assist with column sparing. It becomes a group of Page Protecting Pointers (PPPs), which utilizes otherwise discarded error correction potential to further enhance the leveling ability of RETROFIT. To relieve column-level imbalance, we apply idle error correction bits before they are used to reduce average bit flips. The evaluation demonstrates that RETROFIT and enhanced RETROFIT with the PPPs improve lifetime by as much as 0.64× and 5.4× in the average case, respectively, over state-of-the-art row-level method while also reducing area overhead. In the worst-case scenario, these improvements further increase to 2.6× and 16.0×. Combined with the proposed column-level WL, enhanced RETROFIT realizes an overall 1.5× memory lifetime improvement over the perfectly uniform wear-leveling with equal storage overhead.

An error correcting code has teen developed for the serial transmission of digital data over a fiber optic link. It is a 96 bit block code, with single bit error correction and double bit error detection. Each block consists of eight 8b/lOb coded symbols and sixteen error correction bits. The coded data has a maximum run length of 4, minimum transition density of 25%, and a digital sum variation of +1- 4 bits.A chip set implementing the code has been demonstrated in an experimental link running at 1 Gbit/s (1.5Gbaud).INTRODUCTIONCoding is required for fiber optic data links to transform unencoded data into a serial sequence withknown properties. Since the clock must be recovered fromthe serial data, a high transition density andshort run length are desirable. Most wide bandwidth amplifiers are AC-coupled, dictating that the serialdata have minimal DC content.Forward Error Control (FEC) block codes have been widely used in magnetic media and dynamicRAM systems to improve the effective error rate. At the expense of some channel bandwidth (or storagespace), the error rate may be improved by including check codes that allow detection and correction oferrors.The code described here combines an 8b/lOb code with a Single Error Correcting-Double Error De-tecting (SEC-DED) block code while maintaining the desirable serial data characteristics.8b/IOb CODE

Computational approaches to the analysis of collective behavior in social insects increasingly rely on motion paths as an intermediate data layer from which one can infer individual behaviors or social interactions. Honey bees are a popular model for learning and memory. Previous experience has been shown to affect and modulate future social interactions. So far, no lifetime history observations have been reported for all bees of a colony. In a previous work we introduced a recording setup customized to track up to 4,000 marked bees over several weeks. Due to detection and decoding errors of the bee markers, linking the correct correspondences through time is non-trivial. In this contribution we present an in-depth description of the underlying multi-step algorithm which produces motion paths, and also improves the marker decoding accuracy significantly. The proposed solution employs two classifiers to predict the correspondence of two consecutive detections in the first step, and two tracklets in the second. We automatically tracked ~2,000 marked honey bees over 10 weeks with inexpensive recording hardware using markers without any error correction bits. We found that the proposed two-step tracking reduced incorrect ID decodings from initially ~13% to around 2% post-tracking. Alongside this paper, we publish the first trajectory dataset for all bees in a colony, extracted from ~3 million images covering 3 days. We invite researchers to join the collective scientific effort to investigate this intriguing animal system. All components of our system are open-source.

Power budget of direct-detection ultra-dense WDM-Nyquist-SCM PON with low-complexity SSBI mitigation

Digital coherent detection and wavelength division multiplexing are promising technologies for achieving throughputs from 100 Gb/s to Tb/s order in optical satellite communications. The adoption of these technologies leads to the possibility of using high performance, low cost commercial off the shelf (COTS) components for satellite communications. In this paper, as COTS optical devices and a DSP-ASIC suitable for enabling 100 Gb/s/wavelength capacity, a micro integrable tunable laser assembly, a lithium niobate modulator, a micro intradyne coherent receiver (Micro ICR) and a DSP-ASIC for 100 Gb/s fiber optical communications were successively exposed to a total fluence of more than 4.0 × 10¹⁰ p/cm² of 70 MeV protons. This fluence is equivalent to the total non-ionizing dose (TNID) and total ionizing dose (TID) experienced during 10 years in low earth orbit. The components were powered, and the performance of a 100 Gb/s signal was measured during irradiation to assess the single event effect (SEE). All components showed no degradation as a result of TNID/TID effects. On the other hand, the Micro ICR and DSP-ASIC showed SEEs which required power cycling of the components to recover their functionality. The SEE affecting the DSP-ASIC also increased the post forward error correction bit error ratio (BER), but the BER nevertheless remained sufficiently low in practice. These radiation test results show that these COTS components can be good candidates for use in satellite communications.

A 112 Gb/s PAM4 ADC based SERDES receiver is implemented on Intel 10 nm FinFET process. The receiver consists of a low noise analog front end (AFE), a 64-way time interleaved analog to digital converter (ADC) and a clock/data recovery (CDR) loop utilizing a 7GHz digitally controlled oscillator (DCO). The receiver supports long reach, -35 dB at Nyquist, channels with a pre-forward error correction bit error rate (BER) $\lt 1\mathrm{e} -6$ making it compatible with existing and projected Reed-Solomon FEC.

Pluggable transceiver modules become the preferable form factor in data centers connectivity due to the lower power consumption and smaller size, which allows high port density. In such small form factors, the optical components may be packaged separately from the electronic blocks. This introduces a new challenge for transmission of ultra-broadband signal over high-speed coupled printed analog traces, as the coupling impairments are enhanced dramatically. Here, an extended model of coupling impairments of coupled microstrip traces in optical fiber communication systems is proposed. This model consists of an accurate formulation of the frequency dependency, which is shown to be critical in the transmission of ultra-broadband signals. In addition, a comparison is presented to standard practice coupling models that are currently being used for crosstalk cancellation techniques in sub-10G systems or very short traces. Based on the inclusive coupling model, novel coupling compensation techniques are proposed and are compared with the classical ones. Comprehensive analysis, which is verified by Monte-Carlo simulation, reveals that the newly introduces coupling compensation technique significantly outperforms the existing crosstalk compensation solutions, for 25G transmission and beyond, thus, longer coupled traces are supported. The new approach is based on symbol-spaced sampling and reduced-complexity digital signal processing. It is shown in the case of 100G PAM4 transmission, based on commercially available 25G components, the newly proposed compensation technique supports long microstrip traces of more than 15 cm, while maintaining 10 -3 pre-forward error correction (pre-FEC) bit error rate (BER).

The enhanced Mobile Broadband (eMBB) and ultra-Reliable Low Latency Communications (uRLLC) are the two main scenarios of $5^{th}$ generation (5G) mobile communication system networks. There is an obvious difference in service requirements between different scenarios. When multi-scenario services coexist in the 5G networks, exploring optimized resource scheduling and allocation strategies become a critical issue. The 5G New Radio (NR) and numerology technologies have been standardized, which lay the foundation for flexible frame structure and adaptive scheduling. In this paper, we propose the self-adaptive flexible transmission time interval (TTI) scheduling (SAFE-TS) strategy in the eMBB and uRLLC coexistence scenario. Machine learning (ML) is applied to achieve flexible TTI scheduling. Moreover, we design the random forest-based ensemble TTI decision algorithm (RF-ETDA) to accomplish the TTI selection for each service. Compared with the existing ML methods, the proposed algorithm has a performance improvement in selecting TTI, especially for the uRLLC services. Then, the TTI selection results will be the basis of system resource scheduling and allocation. The simulation results prove that the proposed SAFE-TS effectively reduce the delay and packet loss rate of the uRLLC services while guaranteeing the eMBB requirements. Therefore, it is highly recommended that flexible TTI scheduling should be applied in the construction of the 5G networks to achieve superior network performances.

An invasive-attack-resistant physical unclonable function (PUF) with strong reliability is presented. The mismatch of capacitor ratios in real fabrication is sampled by a switched-capacitor (SC) circuit and further amplified by a latch-styled sense amplifier. Transmission lines connected to the sampling capacitors are used to cover and protect the whole chip from outside invasive attacks. The proposed SC PUF is fabricated using HJ standard 0.18 $\mu{\rm m}$ CMOS process. Simulation results show that the invasive-attack-resistant sensitivity is less than 2.56 fF. The failed outside invasive attack efforts verify the anti-invasive-attack property of the PUF unit. Besides, measured results show that the proposed SC PUF circuit with 3 V/1.8 V-VDD achieves an error rate 1x-5.5x lower than other PUF implementations. An error correction block is also presented to achieve a nearly zero error rate. The measured average error rate, bias, average intra-die HD and average inter-die HD of the 3 V/1.8 V-VDD SC PUF with error correction block is 0%, 46.72%, 0%, and 49.84% respectively.

This paper is focused on the mobile networks with deployment of LTE offering Voice over LTE (VoLTE) services and existing WCDMA layer — in which voice call continuity is ensured through the Single Radio Voice Call Continuity (SRVCC) feature. SRVCC feature enables handover of voice call in LTE to WCDMA layer when users enter poor LTE coverage area. In this paper, it is envisaged that quality of call measured through Mean Opinion Score (MOS) is always better in LTE compared with WCDMA layer. Due to better call quality, maximizing the voice call duration in LTE layer is recommended through SRVCC serving cell threshold optimization. In order to properly set the optimal SRVCC threshold setting, radio condition at which uplink block error rate (UL BLER) increasing higher than 10% and the acknowledgement rate (ACK Rate) degrading below 90% is evaluated. The Transmission Time Interval (TTI) bundling feature improving the uplink VoLTE coverage is also taken into consideration. This paper proposes serving cell SRVCC threshold RSRP = −108dBm for networks without TTI bundling and RSRP = −117dBm for networks with TTI bundling deployment.

Exchange of information in fast, reliable and secure way with access anywhere is a must for a modern society. This leads to emerging traffic in Data Centers imposing tight requirements for short-reach optical links [1]. Research activities around next generation GbE interfaces requires wavelength division multiplexing (WDM) for short-reach optical links. Therefore C-band for intensity modulation direct detection (IMDD) short-reach optical communications is even more attractive [2]. In this talk we review experimental intensity modulation and direct detection demonstrations for short reach optical links. Furthermore, we evaluate 330-Gbps line rate with 128 quadrature amplitude modulation (QAM) discrete multitone (DMT) [3] and 300 Gbps line rate with 100 Gbaud 8-level pulse amplitude modulation (PAM8) [4] IMDD transmission after 400 meters below pre-forward error correction (FEC) bit error rate (BER) with a single packaged externally modulated laser (EML), a DAC and a packaged InP photo-detector (PD) in C-band [5].

A 224-Gb/s pulse amplitude modulation 4-level (PAM4) ADC-based SerDes receiver (RX) is implemented in a 5-nm FinFET process. The RX consists of a low-noise hybrid analog front-end (AFE) that incorporates both inductive peaking and source degeneration, a 64-way time-interleaved ADC, digital equalization consisting of an up to 30 tap feed-forward equalizer (FFE), optional decision-feedback equalizer (DFE), and a clock-data-recovery (CDR) loop utilizing a 14-GHz digitally controlled oscillator (DCO). The RX was characterized for medium-to-long reach channels (20.6-, 27-, 31.4-, and 38-dB insertion loss at Nyquist) with a corresponding pre-forward error correction (FEC) bit error rates of 1.9E-11, 5.2E-10, 1.2E-8, and 6E-7, respectively. The analog power consumption of the RX is 1.41 pJ/b.

A digital signal processing (DSP) scheme based on Volterra equalizer (VE) combined with adaptive noise-whitening post-filter and maximum likelihood sequence detection (MLSD) is proposed to mitigate nonlinear impairments in vertical-cavity surface-emitting lasers (VCSEL) multimode fiber (MMF) system. Successfully transmission of 108 Gb/s, 100 Gb/s and 60 Gb/s 4-ary pulse amplitude modulation (PAM4) signal over 5 m, 160 m and 460 m OM3-MMF is demonstrated below the 7% overhead hard-decision forward error correction (HD-FEC) bit error rate (BER) threshold by using a 20-GHz class VCSEL at 850 nm. Linear pre-equalization is applied to mitigate severe bandwidth limitation of the system. Our experimental results show that the scheme can well mitigate modulation nonlinearity induced by VCSEL and fiber nonlinearity induced by MMF. The BER decreases about two order of magnitude compared to linear equalizer after 100 m OM3-MMF transmission for 100 Gb/s PAM4 signal.

In this work, we innovatively equalize optical intensity-modulated and directly detected (IMDD) four-level pulse amplitude modulation (PAM-4) signals using a complex-valued decision feedback equalizer (CDFE). Through mapping adjacent symbols of PAM-4 signals onto the complex domain, the influence of strongest inter-symbol interference (ISI) can be alleviated during the decision process in a decision feedback equalizer (DFE), effectively combating burst-error propagation when signals are noisy. Moreover, signal-adaptive manipulations of DFE parameters in both the time and the amplitude domain are performed by using an ultra-stable timing recovery and level-adaptive decision. Performance evaluations are made on vertical cavity surface emitting laser (VCSEL) modulated and multimode fiber (MMF) transmitted 100-Gbit/s optical PAM-4 signals. Based on experimental results of the short-reach optical communication, the proposed DFE outperforms the traditional DFE with a 0.5-dB system power budget gain at the 7% overhead (OH) forward error correction (FEC) bit error rate (BER) threshold.

Emerging high-density non-volatile random access memories (NVRAMs) can significantly enhance server main memory by providing both higher memory density and fast persistent memory. An unique design requirement for server main memory is strong reliability because uncorrectable errors can cause a system crash or permanent data loss. Traditional dynamic random access memory (DRAM) subsystems have used chipkill-correct to provide this reliability, while storage systems provide similar protection using very long ECC words (VLEWs). This paper presents an efficient chipkill-correct scheme for persistent memory based on high-density NVRAMs. For efficiency, the scheme decouples error correction at boot time from error correction at runtime. At boot time, when bit error rates are higher, the scheme uses VLEWs to efficiently ensure reliable data survival for a week to a year without refresh by correcting a large number of bit errors at low storage cost. At runtime, when bit error rates are lower, it reuses each memory block's chip failure protection bits to opportunistically correct bit errors at high performance. The proposal incurs a total storage cost of 27%. Compared to a bit error correction scheme, the proposal adds chip failure protection at no additional storage cost and at 2% average performance overhead.