Search

This letter proposes a high throughput bit-flipping (BF) algorithm for regular LDPC codes with dynamic thresholds. In the iterative decoding, the flipping threshold is dynamically updated according to the previous flipping results. Compared to the exiting hard-information-based BF algorithms, the proposed algorithm can get better decoding performance. At the same time, in order to accelerate the convergence speed, we limit the number of occurrences of high thresholds which effectively reduces the invalid decoding iterations. The proposed algorithm has better decoding performance and lower average number of iterations which is suitable for the flash storage systems with low latency and high throughput requirements.

This article presents generative adversarial network processing unit (GANPU), an energy-efficient multiple deep neural network (DNN) training processor for GANs. It enables on-device training of GANs on performance- and battery-limited mobile devices, without sending user-specific data to servers, fully evading privacy concerns. Training GANs require a massive amount of computation, and therefore, it is difficult to accelerate in a resource-constrained platform. Besides, networks and layers in GANs show dramatically changing operational characteristics, making it difficult to optimize the processor’s core and bandwidth allocation. For higher throughput and energy efficiency, this article proposed three key features. An adaptive spatiotemporal workload multiplexing is proposed to maintain high utilization in accelerating multiple DNNs in a single GAN model. To take advantage of ReLU sparsity during both inference and training, dual-sparsity exploitation architecture is proposed to skip redundant computations due to input and output feature zeros. Moreover, an exponent-only ReLU speculation (EORS) algorithm is proposed along with its lightweight processing element (PE) architecture, to estimate the location of output feature zeros during the inference with minimal hardware overhead. Fabricated in a 65-nm process, the GANPU achieved the energy efficiency of 75.68 TFLOPS/W for 16-bit floating-point computation, which is 4.85 $\times $ higher than the state of the art. As a result, GANPU enables on-device training of GANs with high energy efficiency.

Exploring the sharing opportunities among multiple stream queries is crucial for high-performance stream processing. Modern stream processing necessitates accelerating multiple queries by utilizing heterogeneous coprocessors, such as GPUs, and this has shown to be an effective method. Emerging CPU-GPU integrated architectures 6integrate CPU and GPU on the same chip and eliminate PCI-e bandwidth bottleneck. Such a novel architecture provides new opportunities for improving multi-query performance in stream processing but has not been fully explored by existing systems. We introduce a stream processing engine, called FineStream, for efficient multi-query window-based stream processing on CPU-GPU integrated architectures. FineStream's key contribution is a novel fine-grained workload scheduling mechanism between CPU and GPU to take advantage of both architectures. Particularly, FineStream is able to efficiently handle multiple queries in both static and dynamic streams. Our experimental results show that 1) on integrated architectures, FineStream achieves an average 52 percent throughput improvement and 36 percent lower latency over the state-of-the-art stream processing engine; 2) compared to the coarse-grained strategy of applying different devices for multiple queries, FineStream achieves 32 percent throughput improvement; 3) compared to the stream processing engine on the discrete architecture, FineStream on the integrated architecture achieves 10.4× price-throughput ratio, 1.8× energy efficiency, and can enjoy lower latency benefits.

As one of the three main application scenarios of 5 G cellular system, massive Machine-Type Communications (mMTC) has been regarded as the key solution to facilitate the IoT paradigm. One major bottleneck for accommodating mMTC is the severe congestion at the cellular random access channel when plenty of Machine-Type Devices (MTDs) send access requests concurrently while the preamble resources are limited. To remedy this issue, limiting the number of retransmissions and dropping access requests after the limit is reached can be an effective approach. Yet, the effect of the preamble retransmission limit $K$ on the optimal access performance of mMTC in cellular networks remains largely unexploited, which motivates the study in this paper. Specifically, in this paper, we start by characterizing the network steady-state points based on the limiting probability of successful transmission of access requests. We then obtain explicit expressions of the access throughput and the mean access delay of successfully-transmitted access requests as functions of $K$ and the number of preambles $M$ . The maximum access throughput and the corresponding optimal backoff window size are further derived. It is shown that the maximum access throughput is independent of $K$ , while the mean access delay can be significantly reduced with a small $K$ , yet, at the expense of increased request dropping ratio. In addition, to improve both the throughput and delay performance, the analysis shows that more preambles should be allocated but the performance gain becomes marginal when $M$ is large. Therewith, an algorithm is proposed for determining the least number of preambles $M^\ast$ that maximizes the access throughput and the preamble resource utilization ratio. Numerical results show that a smaller preamble retransmission limit $K$ can further reduce $M^\ast$ .

In this letter, we consider outer-loop link adaptation, wherein we pursue to find the optimum modulation and coding scheme (MCS) that provides the maximum throughput without directly estimating the channel state. We cast this problem in a form of multi-armed bandit, an online decision making policy based on sequential observations. To efficiently solve a formulated problem, we propose a novel two-stage Thompson sampling. The proposed method is built based on the observation that the optimum MCS level appears in a certain group that satisfies feasibility conditions. Exploiting this feature, we find the optimum MCS level via two stages. In the first stage, we identify a group that has high probability of including the optimum MCS. In the second stage, we only focus on the MCS levels within the identified group in the first stage, and investigate the optimum MCS. By doing this, the search space is significantly reduced, which leads to the performance improvement. Simulation results show that the proposed method outperforms the existing state-of-art algorithm.

In industrial wireless sensor networks (IWSNs), a lot of energy is wasted in the form of electromagnetic radiations. It can be effectively utilized with energy harvesting (EH), which absorbs part of the energy in the transmission signal but reduces the throughput and reliability of IWSNs. In this article, we study the throughput optimization of IWSNs with EH from the interference radio-frequency (RF) signal considering the reliability constraint of the industrial information transmission. Under the premise of limited energy supply of EH relays, the throughput maximization of IWSNs is formulated as a nonconvex optimization problem. In order to transform the nonconvex problem to a convex optimization problem, the successive convex approximation (SCA) approach is adopted. Furthermore, a power allocation algorithm is designed to maximize the total transmission rate of the network. Simulation results demonstrate that the proposed algorithm can maximize the throughput under the primise of SINR reliability.

Radio resource management (RRM) is critical in 5G mobile communications due to its ubiquity on every radio device and its low latency constraints. The rapidly evolving RRM algorithms with low latency requirements combined with the dense and massive 5G base station deployment ask for an on-the-edge RRM acceleration system with a tradeoff between flexibility, efficiency, and cost-making application-specific instruction-set processors (ASIPs) an optimal choice. In this work, we start from a baseline, simple RISC-V core and introduce instruction extensions coupled with software optimizations for maximizing the throughput of a selected set of recently proposed RRM algorithms based on models using multilayer perceptrons (MLPs) and recurrent neural networks (RNNs). Furthermore, we scale from a single-ASIP to a multi-ASIP acceleration system to further improve RRM throughput. For the single-ASIP system, we demonstrate an energy efficiency of 218 GMAC/s/W and a throughput of 566 MMAC/s corresponding to an improvement of $10\times $ and $10.6\times $ , respectively, over the single-core system with a baseline RV32IMC core. For the multi-ASIP system, we analyze the parallel speedup dependency on the input and output feature map (FM) size for fully connected and LSTM layers, achieving up to $10.2\times $ speedup with 16 cores over a single extended RI5CY core for single LSTM layers and a speedup of $13.8\times $ for a single fully connected layer. On the full RRM benchmark suite, we achieve an average overall speedup of $16.4\times $ , $25.2\times $ , $31.9\times $ , and $38.8\times $ on two, four, eight, and 16 cores, respectively, compared to our single-core RV32IMC baseline implementation.

In this letter, we consider a problem of throughput optimum channel switching for channel access by a secondary user in multichannel Cognitive Radio Networks in which there is a delay involved in switching from one channel to another. We consider an imperfect sensing model that takes into account false alarm and detection probabilities. We show that the problem can be posed in the framework of Partially Observable Markov Decision Process (POMDP), and obtain an average throughput optimal policy for which we provide structural results. Also, we propose a low complex algorithm ASTUTE for channel access based on one stage reward. We compare the performance of the optimum policy with ASTUTE . Our results show that the average throughput performance of ASTUTE is very close to that of the optimal solution.

Intersatellite communication (ISC), also known as intersatellite links (ISLs), comes with many benefits to distributed space systems (DSSs) including advanced capabilities, such as relative navigation among small satellites and a reduction in the need for the ground segment. Radio frequency (RF) has been the de-facto method for ISC in DSS missions. However, as new science missions evolve with advanced payloads and complex architectures, RF may not be able to meet the ISC requirements of high bandwidth and throughput without violating the size, mass, power, and cost (SMaP-C) constraints of small satellites. Laser communication has the potential of meeting the requirements for these more complex missions but may violate the constraints imposed on small satellites due to its high-precision pointing and tracking requirements. This work presents the progress made on the development of LED-based visible light ISC (VL-ISC). Key aspects of the system design are presented with regards to the physical and data link layers. The simulation of the nonorthogonal multiple-access scheme for VL-ISC in a cluster of small satellites indicates an overall system throughput sum rate of about 2 Mbit/s (for system bandwidth of 0.5 MHz) among others. The roadmap to the development of a future VL-ISC standard is discussed.

In view of the fact that the current peanut test was still artificial, it was difficult to meet the actual demand of peanut test, this paper established a peanut high-throughput automatic test syst...

Production systems literature usually attributes throughput losses to two reasons: unreliable equipment and random part processing time (also referred to as machine cycle time). In practice, howeve...

The new extended Wi-Fi band of 5.4 GHz band is less congested compared to the traditional 2.4 GHz band. The ability to handle large traffic of this 5.4 GHz band captured the interest of most researchers and hence there is a need to developed efficient antennas at 5.4 GHz frequency. This manuscript presents a novel compact V-cut patch antenna for the 5.4 GHz Wi-Fi application. The proposed design is a (2 $$\times$$ 1) MIMO antenna with a partial ground structure. The patch structure and partial ground were optimized using simulation software. The simulated structure is optimized using an optimizer. The optimized structure is fabricated on the FR-4 substrate and tested for performance metrics. The proposed MIMO antenna was found out to be a wideband antenna with an overall bandwidth of 1500 MHz. The measured gain was 3.86 and the Return loss (S11) value was $$-23$$ dB which closely resembles a simulated value of −27 dB. The proposed MIMO antenna is compact size 32 mm $$\times$$ 56 mm and can be used with existing Wi-Fi routers to increase the capacity of the channel and throughput of the system.

The use of wireless networks has become a trend at this time. However, this can cause several problems in the use of this technology. One of the problems arising from this technology is the limited signal coverage in a certain place. To solve these problems, WDS technology is an alternative solution that can be done. WDS technology will be applied to each room. In this study, QOS analysis will be used to evaluate throughput performance and response time. The test scenario is performed with 1000 users (simulated) for seven days, four sampling times considering working hours and outside working hours. The analysis results show that with WDS technology, the resulting performance tends to be more stable with a throughput value of 500 KBps and a max response time of 5.5 ms.

The COVID-19 pandemic has exposed the dependence of diagnostic laboratories on a handful of large corporations with market monopolies on the worldwide supply of reagents, consumables, and hardware for molecular diagnostics. Global shortages of key consumables for RT-qPCR detection of SARS-CoV-2 RNA have impaired the ability to run essential, routine diagnostic services. Here, we describe a workflow for rapid detection of SARS-CoV-2 RNA in upper respiratory samples including nasal swabs and saliva, utilizing low-cost equipment and readily accessible reagents. Using repurposed Creality3D Ender-3 three-dimensional (3D) printers, we built a semiautomated paramagnetic bead RNA extraction platform. The hardware for the system was built for $300 USD, and the material cost per reaction was $1 USD. Named the Ender VX500, instrument performance when paired with RT-qPCR for SARS-CoV-2 detection in nasal and saliva specimens was two virus copies per microliter. There was a high-performance agreement (assessed using 458 COVID-19 nasal swab specimens) with the Aptima SARS-CoV-2 assay run on the Hologic Panther, a commercial automated RNA extraction and detection platform. Inter- and intrainstrument precision was excellent (coefficients of variation (CoV) of 1.10 and 0.66–1.32%, respectively) across four instruments. The platform is scalable with throughput ranging from 23 specimens on a single instrument run by one user in 50 min to 364 specimens on four instruments run by four users in 190 min. Step-by-step instructions and protocols for building and running the Ender VX500 have been made available without restriction.

The process of cluster head selection under the process of cluster formation in wireless sensor networks is determined to be essential for extending the lifetime of the network. In this paper, a Self Adapting Differential Search Strategies Improved Artificial Bee Colony Algorithm (SADSS-IABCA)-based Cluster Head Selection Scheme is proposed for prolonging the lifetime of the network with improved Quality of Service. The differential search strategies employed in the SADSS-IABCA-based Cluster Head Selection Scheme are reliable in updating the dependent variables in periodic intervals of time through the integration of mutation and crossover. This proposed SADSS-IABCA-based Cluster Head Selection Scheme incorporated diversified search strategies associated with differential evolution with the employee and onlooker bee phase in order to improve the local searching ability of ABC with the view to eliminate its limitations of delayed convergence. In addition, the appropriate selection of differential evolution strategies is computed through the probability-based self adapting process for effective selection of cluster heads. The simulation results of the proposed SADSS-IABCA-based Cluster Head Selection Scheme also confirmed a predominant improvement in percentage of alive nodes, throughput and mean residual energy compared to the benchmarked cluster head schemes used for investigation.

Purpose The small area network for data communication within routers is suffering from storage of packet, throughput, latency and power consumption. There are a lot of solutions to increase speed of commutation and optimization of power consumption; one among them is Network-on-chip (NoC). In the literature, there are several NoCs which can reconfigurable dynamically and can easily test and validate the results on FPGA. But still, NoCs have limitations which are regarding chip area, reconfigurable time and throughput. Design/methodology/approach To address these limitations, this research proposes the dynamically buffered and bufferless reconfigurable NoC (DB2R NoC) using X-Y algorithm for routing, Torus for switching and Flexible Direction Order (FDOR) for direction finding between source and destination nodes. Thus, the 3 × 3 and 4 × 4 DB2R NoCs are made free from deadlock, low power and latency and high throughput. To prove the applicability and performance analysis of DB2R NoC for 3 × 3 and 4 × 4 routers on FPGA, the 22 bits for buffered and 19 bit for bufferless designs have been successfully synthesized using Verilog HDL and implemented on Artix-7 FPGA development bond. The virtual input/output chips cope pro tool has been incorporated in the design to verify and debug the complete design on Artix-7 FPGA. Findings In the obtained result, it has been found that 35% improvement in throughput, 23% improvement in latency and 47% optimization in area has been made. The complete design has been tested for 28 packets of injection rate 0.01; the packets have been generated by using NLFSR. Originality/value In the obtained result, it has been found that 35% improvement in throughput, 23% improvement in latency and 47% optimization in area has been made. The complete design has been tested for 28 packets of injection rate 0.01; the packets have been generated by using NLFSR.

With the rapid increase in the functionality of IoT applications, the services provided by edge/IoT devices have surged in the recent past. Fog computing is gaining momentum as a contemporary computational framework for IoT-enabled smart applications that offers a latency sensitivity advantage over cloud computing. Effective task scheduling reduces application computation and latency durations while improves QoS. Many researchers have proposed a variety of heuristic and metaheuristic approaches for effective scheduling; however, there is still scope for improvement. The present work proposes a hybrid of heuristic and metaheuristic techniques for the scheduling of tasks. fireworks algorithm (FWA) is a metaheuristic algorithm, and Heterogeneous Earliest Finish Time (HEFT) is a heuristic algorithm. The bi-objective optimization approach is presented to minimize the makespan and cost factors. To compare the performance of the algorithm, the experiments have been conducted on distinct and independent scientific workflows. The results of exhaustive simulations have established the significance of the presented BH-FWA algorithm over other comparative approaches. To validate the relevance of the technique for fog computing networks, the metrics utilized for comparisons are makespan, cost, and throughput.

Radiomics is a high-throughput approach to image phenotyping. It uses computer algorithms to extract and analyze a large number of quantitative features from radiological images. These radiomic features collectively describe unique patterns that can serve as digital fingerprints of disease. They may also capture imaging characteristics that are difficult or impossible to characterize by the human eye. The rapid development of this field is motivated by systems biology, facilitated by data analytics, and powered by artificial intelligence. Here, as part of Abdominal Radiology's special issue on Quantitative Imaging, we provide an introduction to the field of radiomics. The technique is formally introduced as an advanced application of data analytics, with illustrating examples in abdominal radiology. Artificial intelligence is then presented as the main driving force of radiomics, and common techniques are defined and briefly compared. The complete step-by-step process of radiomic phenotyping is then broken down into five key phases. Potential pitfalls of each phase are highlighted, and recommendations are provided to reduce sources of variation, non-reproducibility, and error associated with radiomics.

In situ microplates are small in size, crystal cultivation and operation are difficult, and the efficiency of crystal screening is relatively low. To solve this problem, a novel combined crystallization plate was designed for high-throughput crystal cultivation and in situ data collection. A frame was used to hold 48 in situ microplates, and the in situ microplates were sealed on one side with an ultralow background-scattering Kapton film. An automatic liquid handler (Mosquito) was used to add a liquid drop to the in situ microplates in the frame, and CrystalClear HD tape was used to seal the frame. A sealed frame holding 48 microplates was developed as a novel combined crystallization plate and was used for crystal cultivation under different conditions and in situ data collection at the synchrotron beamline. Moreover, individual microplates can be separated from the combined crystal plate and then fixed on a magnetic base or loaded onto a UniPuck for in situ data collection. Automatic grid scanning was used to locate crystals. The efficiency of the combined crystallization plate for crystal screening was verified. This method avoids the manual manipulation of crystals during crystal screening and diffraction data collection; therefore, the combined crystallization plate is suitable for large-scale screening of microcrystals.

Massive MIMO is an important framework of 5G and the Internet of Things (IoT). Signals radiated from a 5G antenna share the complex architecture. A system is therefore needed that can minimize complexity and provides good spectral efficiency and throughput. This is achieved by reducing RF chains which is an important constraint in an energy-efficient 5G system. In this paper, a new semi-orthogonal-based scheme is proposed to formulate SINR which maximizes throughput when RF chains are combined without increasing computational complexity, and also the focus is laid on the improvement of the sum data rate of the system. More importantly, the optimization algorithm has been implemented and used to model the massive MIMO as a tool for an energy-efficient system.

Big data analytics in cloud environments introduces challenges such as real-time load balancing besides security, privacy, and energy efficiency. In this paper, we propose a novel load balancing algorithm in cloud environments that performs resource allocation and task scheduling efficiently. The proposed load balancer reduces the execution response time in big data applications performed on clouds. Scheduling, in general, is an NP-hard problem. In our proposed algorithm, we provide solutions to reduce the search area that leads to reduced complexity of the load balancing. We recommend two mathematical optimization models to perform dynamic resource allocation to virtual machines and task scheduling. The provided solution is based on the hill-climbing algorithm to minimize response time. We evaluate the performance of proposed algorithms in terms of response time, turnaround time, throughput metrics, and request distribution with some of the existing algorithms that show significant improvements, Comment: My supervisor has requested that i withdraw the article

Over the top (OTT) video streaming is widely deployed to deliver stored media. But, the OTT video quality assessment method is uncertain. In this paper, we introduce three application metrics of OTT video quality, including initial buffering time, mean re-buffering duration, and re-buffering frequency. We find the transmission control protocol (TCP) throughput can influence the application metrics; then, we improve the TCP model and forecast the TCP throughput more accuracy. Next, we give experiments to change the network environment and test the OTT video. Our ultimate goal is to characterize the correlation between the application metrics and user quality of experience using simplified fuzzy inference system. Firstly, the nonlinear support vector machine is used to divide the application metrics into different categories. Secondly, the different categories of fuzzy rules are designed to infer the objective scores. The proposed objective model can reduce the calculation steps effectively and improve the similarity. Lastly, the existing assessment methods are compared with it, the experimental results show that the method accords closely with human subjective judgment.

Compared with the current continuous development of the Internet, the unbalanced distribution of medical resources and low efficiency have made the traditional medical system stretched. In order to solve these problems and allow patients to get health advice from doctors at home, this article proposes an Internet-based telemedicine system. This article first elaborates the research background and significance of telemedicine, the development status at home and abroad, and the advantages and disadvantages of telemedicine. In the overall design, the system is divided into basic platform subsystem, application platform subsystem, and specific application subsystem, and the structure and implementation process of each module are designed. Aiming at the reliability of data transmission, this paper improves the traditional follow-up selection algorithm. The performance of the improved algorithm proposed is better than that of the traditional algorithm, which greatly improves the throughput of the system, avoids the repeated use of the same node, and improves the stability of the system. This paper designs a more comprehensive telemedicine system at the functional level, which can meet various functions required by telemedicine, enable patients and doctors to communicate in real time, and effectively improve doctors’ work efficiency.

Case‐control design based high‐throughput pharmacoinformatics study using large‐scale longitudinal health data is able to detect new adverse drug event (ADEs) signals. Existing control selection approaches for case‐control design included the dynamic/super control selection approach. The dynamic/super control selection approach requires all individuals to be evaluated at all ADE case index dates, as the individuals’ eligibilities as control depend on ADE/enrollment history. Thus, using large‐scale longitudinal health data, the dynamic/super control selection approach requires extraordinarily high computational time. We proposed a random control selection approach in which ADE case index dates were matched by randomly generated control index dates. The random control selection approach does not depend on ADE/enrollment history. It is able to significantly reduce computational time to prepare case‐control data sets, as it requires all individuals to be evaluated only once. We compared the performance metrics of all control selection approaches using two large‐scale longitudinal health data and a drug‐ADE gold standard including 399 drug‐ADE pairs. The F‐scores for the random control selection approach were between 0.586 and 0.600 compared to between 0.545 and 0.562 for dynamic/super control selection approaches. The random control selection approach was ~ 1000 times faster than dynamic/super control selection approach on preparing case‐control data sets. With large‐scale longitudinal health data, a case‐control design‐based pharmacoinformatics study using random control selection is able to generate comparable ADE signals than the existing control selection approaches. The random control selection approach also significantly reduces computational time to prepare the case‐control data sets.

This study proposes a train dispatching management algorithm for autonomous shuttle freight train (AST) service in a shared railway corridor with regular train traffic. The method operates the ASTs without disturbing the regular train paths and involves consecutive overtaking and waiting decisions at the intermediate stations. The developed methodology and algorithm were used for developing a real-time dispatching control module integrated with a multimethod railway microsimulation model capable of simulating the timetable and non-timetable-based regular trains and ASTs. The model was implemented for a railway corridor with regular train traffic and the efficiency of the control module was confirmed. The AST freight transport capacity was evaluated using several key performance indicators based on train throughput, punctuality, and delays. The main conclusions of the study are as follows: the control logic decreased the number of AST trips by 5.6% in favor of protecting the regular train paths for (1 + 1) locomotive couple (LC) and decreased down to 14.8% for (6 + 6) LC. The AST yard delay increased by 238.2% for (6 + 6) LC. The average number of station stops and total trip times increased up to 51.2% and 14.2% for the (6 + 6) LC. The slower AST speeds also triggered more dramatic capacity decreases, increased station stops and total trip times. The model was also evaluated for random regular train failures, and daily average of AST trips made per year decreased by 8.6% for (6 + 6) LC for the mean failure interval time of 5 days and a randomly selected regular train.

Ultra-high-definition (UHD) video standards demand processing speed from 60 to 120 fps. These standards require relatively huge resources for providing such high processing speed. In this paper, an area-efficient and high-speed two-dimensional (2D) $$2\times 2$$ and $$3\times 3$$ block parallel scalable recursive convolution (BPSRC) architectures are proposed. The $$2\times 2$$ and $$3 \times 3$$ BPSRC architectures are used to implement block parallel filters with small to large kernel sizes but limited to multiples of 2 and 3, respectively. For a block parallel convolution, the spatial window is partitioned into fixed size blocks for parallel processing of the block outputs. The algorithm proved effective with respect to area and computational time. The increase in kernel size does not affect the processing time but increases the hardware cost. However, the increase in hardware cost is considerably less when compared with conventional block parallel convolution (BPC). Overall, multiplier complexity is reduced by a factor of 4/9 and 9/16 for $$3\times 3$$ and $$2 \times 2$$ BPSRC implementation of 2D finite impulse response (FIR) filters, respectively, over conventional BPC. A throughput of 1.55 Giga operations per second is achieved with $$2\times 2$$ BPSRC, and that of 1.86 Giga operations per second is achieved with $$3\times 3$$ BPSRC on Virtex 7 XC7VX485T FPGA.

Modern web services use in-memory caching extensively to increase throughput and reduce latency. There have been several workload analyses of production systems that have fueled research in improving the effectiveness of in-memory caching systems. However, the coverage is still sparse considering the wide spectrum of industrial cache use cases. In this work, we significantly further the understanding of real-world cache workloads by collecting production traces from 153 in-memory cache clusters at Twitter, sifting through over 80 TB of data, and sometimes interpreting the workloads in the context of the business logic behind them. We perform a comprehensive analysis to characterize cache workloads based on traffic pattern, time-to-live (TTL), popularity distribution, and size distribution. A fine-grained view of different workloads uncover the diversity of use cases: many are far more write-heavy or more skewed than previously shown and some display unique temporal patterns. We also observe that TTL is an important and sometimes defining parameter of cache working sets. Our simulations show that ideal replacement strategy in production caches can be surprising, for example, FIFO works the best for a large number of workloads.

The traditional scheme for load balancing in a homogeneous Long Term Evolution (LTE) Network where User Equipment (UEs) associate to a node with the strongest received signal strength is not practical for LTE Heterogeneous Network (LTE HetNet) due to power disparity between the nodes. Therefore, dynamic Cell Range Expansion (CRE) based load-balancing schemes were employed by several scholars to address the challenges in the LTE HetNet. However, the fairness index in achieving the desired average user throughput and UE offloading effect is relatively low. In this work, an efficient load-balancing algorithm for LTE HetNet based on dynamic Cell Range Expansion (CRE) was developed to improve the fairness of the network for the desired throughput and UE offloading effect. The simulation results achieve a throughput gain improvement of up to 11%, while the fairness index improves by 6% compared to the existing algorithm. Further, the UEs offloading effect shows a significant improvement of 3% relative to the existing algorithm. Keywords: Fairness Index; Cell Range Expansion; Load Balancing; LTE Heterogeneous Network; Throughput

The article describes 30 small container terminals in the Baltic Sea Region and examines their competitive advantages. Terminals whose annual maximum throughput capacity does not exceed 150,000 TEU...

Container throughput forecasting plays an important role in port capacity planning and management. Regarding the issue of container throughput of Tianjin-Hebei Port Group, considering the container throughput is an incomplete grey information system affected by various factors, the effect is often unsatisfactory by adopting a single forecasting model. Therefore, this paper studies the issue by combining fractional GM (1, 1) and BP neural network. The comparison results show that the combination model performs better than other single models separately and has a higher level of forecasting accuracy. Furthermore, the combination model is adopted to forecast the container throughput of Tianjin-Hebei Port Group from 2021 to 2025, which would be a data reference for the future development optimization for the container operation of Tianjin-Hebei Port Group.

Purpose Digital signal processing (DSP) applications such as finite impulse response (FIR) filter, infinite impulse response and wavelet transformation functions are mainly constructed using multipliers and adders. The performance of any digital applications is dependent on larger size multipliers, area and power dissipation. To optimize power and area, an efficient zero product and feeder register-based multiplier (ZP and FRBM) is proposed. Another challenging task in multipliers is summation of partial products (PP), results in more delay. To address this issue, the modified parallel prefix adder (PPA) is incorporated in multiplier design. In this work, different methods are studied and analyzed for designing FIR filter, optimized with respect to area, power dissipation, speed, throughput, latency and hardware utilization. Design/methodology/approach The distributed arithmetic (DA)-based reconfigurable FIR design is found to be suitable filter for software-defined radio (SDR) applications. The performance of adder and multipliers in DA-FIR filter restricts the area and power dissipation due to their complexity in terms of generation of sum and carry bits. The hardware implementation time of an adder can be reduced by using PPA which is based on Ling equation. The MDA-RFIR filter is designed for higher filter length (N), i.e. N = 64 with 64 taps and this design is developed using Verilog hardware description language (HDL) and implemented on field-programmable gate array. The design is validated for SDR channel equalizer; both RFIR and SDR are integrated as single system and implemented on Artix-7 development board of part name XC7A100tCSG324. Findings The MDA-RFIR for N = 64 is optimized about 33% in terms of area-delay, power-speed product and energy efficiency. The theoretical and practical comparisons have been done, and the practically obtained results are compared with existing DA-RFIR designs in terms of throughput, latency, area-delay, power-speed product and energy efficiency are better about 3.5 times, 31, 45 and 29%, respectively. Originality/value The MDA-RFIR for N = 64 is optimized about 33% in terms of area-delay, power-speed product and energy efficiency.

We investigate the scheduling practices of multistage outpatient health programs that offer care plans customized to the needs of their patients. We formulate the scheduling problem as a Markov decision process (MDP) where patients can reschedule their appointment, may fail to show up, and may become ineligible. The MDP has an exponentially large state space and thus, we introduce a linear approximation to the value function. We then formulate an approximate dynamic program (ADP) and implement a dual variable aggregation procedure. This reduces the size of the ADP while still producing dual cost estimates that can be used to identify favorable scheduling actions. We use our scheduling model to study the effectiveness of customized-care plans for a heterogeneous patient population and find that system performance is better than clinics that do not offer such plans. We also demonstrate that our scheduling approach improves clinic profitability, increases throughput, and decreases practitioner idleness as compared to a policy that mimics human schedulers and a policy derived from a deep neural network. Finally, we show that our approach is fairly robust to errors introduced when practitioners inadvertently assign patients to the wrong care plan.

The throughput of a belt conveyor system is the primary design parameter when considering a new installation, determined by the cross-section of the material on the belt, coupled with the belt spee...

Organic electrochemistry has emerged as an enabling and sustainable technology in modern organic synthesis. Despite the recent renaissance of electrosynthesis, the broad adoption of electrochemistry in the synthetic community, and especially in industrial settings, has been hindered by the lack of general, standardized platforms for high-throughput experimentation (HTE). Herein, we disclose the design of the HTe–Chem, a high-throughput microscale electrochemical reactor that is compatible with existing HTE infrastructure and enables the rapid evaluation of a broad array of electrochemical reaction parameters. Utilizing the HTe–Chem to accelerate reaction optimization, reaction discovery, and chemical library synthesis is illustrated using a suite of oxidative and reductive transformations under constant current, constant voltage, and electrophotochemical conditions., Electrochemistry is an enabling technology in modern organic synthesis. The HTe−Chem is an HTE reactor that accelerates the discovery, optimization, and application of electrochemical reactions.

Low-latency and energy-efficient multi-Gbps LDPC decoding requires fast-converging iterative schedules. Hardware decoder architectures based on such schedules can achieve high throughput at low clock speeds, resulting in reduced power consumption and relaxed timing closure requirements for physical VLSI design. In this work, a fast column message-passing (FCMP) schedule for decoding LDPC codes is presented and investigated. FCMP converges in half the number of iterations compared to existing serial decoding schedules, has a significantly lower computational complexity than residual-belief-propagation (RBP)-based schedules, and consumes less power compared to state-of-the-art schedules. An FCMP decoder architecture supporting IEEE 802.11ad (WiGig) LDPC codes is presented. The decoder is fully pipelined to decode two frames with no idle cycles. The architecture is synthesized using the TSMC 40 nm and 65 nm CMOS technology nodes, and operates at a clock-frequency of 200 MHz. The decoder achieves a throughput of 8.4 Gbps, and it consumes 72 mW of power when synthesized using the 40 nm technology node. This results in an energy efficiency of 8.6 pJ/bit, which is the best-reported energy-efficiency in the literature for a WiGig LDPC decoder.

Based on nationwide aircraft performance and scheduling gathered from Variflight.com, the work evaluates how the identified factors, such as aircraft taxi-in and taxi-out time, airborne time, Origin–destination Pair Herfindahl–Hirschman Index (OD-Pair HHI), hub airport, and flight distance, coexisted in order to organize aviation scheduling more effectively and appropriately in the air transportation system. The result indicates that the variables of a hub airport, OD-Pair HHI and throughput have a more significant effect on SBT than the duration of each flight phase does (i.e., aircraft taxi-in and taxi-out time, airborne time, etc.). Fierce competition has prompted airlines to improve their on-time performance (OTP), instead of scheduling shorter block times. Moving forward, if the departure airport is a major one, the SBT will increase by 3.43 min, indicating that airport throughput is a primary determinant in adding or shortening SBT in China's aviation system.

In this mymargin letter, we investigate a backbone-assisted wireless powered communication network (WPCN) under the protocol of “harvest-then-transmit” protocol, where user nodes (UNs) harvest energy from the sinks in the downlink (DL), then they use the energy to transmit signals to the sinks by a non-orthogonal multiple access in the uplink (UL). We consider joint successive interference cancellation and backbone-assisted sink cooperation to reduce the impact of co-channel interference on the throughput performance. We maximize the system throughput by jointly optimizing user scheduling and the time allocation for DL energy harvesting and UL transmission. For solving the maximum throughput problem efficiently, we propose a column generation method based on a greedy algorithm. Numerical results show that our proposed energy and information transfer design is superior to other reference methods in WPCNs.

This letter studies the application of backscatter communications (BackCom) assisted non-orthogonal multiple access (BAC-NOMA) to the envisioned sixth-generation (6G) ultra-massive machine type communications (umMTC). In particular, the proposed BAC-NOMA transmission scheme can realize simultaneous energy and spectrum cooperation between uplink and downlink users, which is important to support massive connectivity and stringent energy constraints in umMTC. Furthermore, a resource allocation problem for maximizing the uplink throughput and suppressing the interference between downlink and uplink transmission is formulated as an optimization problem and the corresponding optimal resource allocation policy is obtained. Computer simulations are provided to demonstrate the superior performance of BAC-NOMA.

Library preparation for high-throughput sequencing applications is a critical step in producing representative, unbiased sequencing data. The iGenomX Riptide High Throughput Rapid Library Prep Kit purports to provide high-quality sequencing data with lower costs compared to other Illumina library kits. To test these claims, we compared sequence data quality of Riptide libraries to libraries constructed with KAPA Hyper and NEBNext Ultra. Across several single-source genome samples, mapping performance and de novo assembly of Riptide libraries were similar to conventional libraries prepared with the same DNA. Poor performance of some libraries resulted in low sequencing depth. In particular, degraded DNA samples may be challenging to sequence with Riptide. There was little cross-well plate contamination with the overwhelming majority of reads belong to the proper source genomes. The sequencing of metagenome samples using different Riptide primer sets resulted in variable taxonomic assignment of reads. Increased adoption of the Riptide kit will decrease library preparation costs. However, this method might not be suitable for degraded DNA.

Purpose: The paper aims to elaborate on long-term throughput forecasts in Small and Medium-Sized ports (SMPs). The research problem relates to determining a method that is relevant for the long-term transhipment forecasting in SMPs. Design/Methodology/Approach: Research was applied to the Polish ports system consisting of three major ports (Gdańsk, Gdynia, and Świnoujście) and the minor port of Szczecin. Forecasts for the cargo groups in the major ports were produced using regression models where the model parameters were estimated with the Ordinary Least Squares Method. The obtained forecasts of the throughput of cargo groups in major Polish seaports were converted into dynamics indices. The resulting matrices of the cargo throughput dynamics indices established for the leading Polish seaports were used to prepare cargo throughput forecasts for the port of Szczecin. Findings: Elaborated throughput forecasts indicate that, in the future, the port in Szczecin will retain its universal character, and a moderate increase in the cargo volumes confirms that it will serve as a complementary port to the major ports in the range. The forecasted modal split of hinterland transport indicates the increasing importance of road transport, and a decreased rail and inland waterway transport to/from the hinterland. That is likely due to a lower cargo throughput and difficulties in organizing the rail and inland waterway corridors based on smaller freight volumes. Practical implications: The study provides a practical tool for long-term throughput and hinterland traffic forecasting in SMPs. It is addressed to port authorities as the method may be used in decision-making on capacity expansion and academics exploring the phenomenon of predictions in maritime transport. Originality/Values: The novel method of demand forecasting in SMPs includes relations between major and minor ports in the range and multi-staged validation of results. Research advances studies on the dynamics of multi-port systems., peer-reviewed

The issues of increasing the operation efficiency of mobile communication systems of next generations using adaptive noiseless coding are considered. The optimization method of the turbo code encoder and decoder operation is proposed using an adaptive choice of the state diagrams size with a new parameter of decoding uncertainty value. The method implementation allows us to provide the specified values of information reliability without reducing the paths throughput of systems and networks of 5G generation.

As AI applications become pervasive on edge device, incrementally learning new tasks is demanded for deep neural network (DNN) models. In this article, we proposed AILC, a compute-in-memory (CIM)-based accelerator for on-chip incremental learning using STT-MRAM technology. On the software side, a network-expansion-based low-precision training algorithm is proposed for incremental learning, where the loss function is modified to handle the unbalanced training dataset. On the hardware side, the detailed CIM accelerator design for incremental learning is illustrated. A workload-aware hardware resources assignment protocol is proposed to improve the throughput when the workload of weight gradient calculation is low. The software simulation results on CIFAR-100 dataset show that the proposed algorithm can effectively support incremental learning despite the device conductance variation exists. System-level benchmark shows that AILC could achieve 147×, 3.7×∼28.7×, 2.05×∼2.9× higher energy efficiency than Nvidia Titan-V GPU, RRAM-based CIM accelerators and edge TPU/GPU, respectively. Compared to the baselines, the throughput of AILC is improved by 2.0×∼2.2× on average with the hardware resources assignment protocol, which results in 4.1×∼21.4× higher throughput than edge TPU/GPU.

To accelerate the discovery of materials through computations and experiments, a well-established protocol closely bridging these methods is required. We introduce a high-throughput screening protocol for the discovery of bimetallic catalysts that replace palladium (Pd), where the similarities in the electronic density of states patterns were employed as a screening descriptor. Using first-principles calculations, we screened 4350 bimetallic alloy structures and proposed eight candidates expected to have catalytic performance comparable to that of Pd. Our experiments demonstrate that four bimetallic catalysts indeed exhibit catalytic properties comparable to those of Pd. Moreover, we discover a bimetallic (Ni-Pt) catalyst that has not yet been reported for H2O2 direct synthesis. In particular, Ni61Pt39 outperforms the prototypical Pd catalyst for the chemical reaction and exhibits a 9.5-fold enhancement in cost-normalized productivity. This protocol provides an opportunity for the catalyst discovery for the replacement or reduction in the use of the platinum-group metals.

In this work, a ReRAM-based energy-efficient CIM accelerator is presented with two techniques for edge AI applications. Firstly, a circuit-algorithm co-design scheme is proposed to realize fully analog processing, which improves the energy efficiency and the throughput of neural network. To deal with the I-V nonlinearity of ReRAM, a nonlinear-aware training algorithm is proposed to improve the network accuracy. Secondly, a 1T2R cell is proposed to replace previous 2T2R for weight storage with 35% area saving. For evaluation, a neural network with two fully connected layers and one ReLU layer is built for the MNIST dataset. The error rate can be reduced by >46% and the energy efficiency is 99 TOPS/W@200 MHz, 2.6X improvement over the digital method.