Lossless Networks Research Articles

Compression-based inference of network motif sets.

Physical and functional constraints on biological networks lead to complex topological patterns across multiple scales in their organization. A particular type of higher-order network feature that has received considerable interest is network motifs, defined as statistically regular subgraphs. These may implement fundamental logical and computational circuits and are referred to as "building blocks of complex networks". Their well-defined structures and small sizes also enable the testing of their functions in synthetic and natural biological experiments. Here, we develop a framework for motif mining based on lossless network compression using subgraph contractions. This provides an alternative definition of motif significance which allows us to compare different motifs and select the collectively most significant set of motifs as well as other prominent network features in terms of their combined compression of the network. Our approach inherently accounts for multiple testing and correlations between subgraphs and does not rely on a priori specification of an appropriate null model. It thus overcomes common problems in hypothesis testing-based motif analysis and guarantees robust statistical inference. We validate our methodology on numerical data and then apply it on synaptic-resolution biological neural networks, as a medium for comparative connectomics, by evaluating their respective compressibility and characterize their inferred circuit motifs.

LC-TTFS: Toward Lossless Network Conversion for Spiking Neural Networks With TTFS Coding

LC-TTFS: Toward Lossless Network Conversion for Spiking Neural Networks With TTFS Coding

Hierarchical equalization scheme for retired lithium-ion battery packs based on inductor-flyback transformer

Some lithium-ion batteries can be reused after sorting and recombination. Equalization of these batteries during the process of step utilization is an effective way to improve the inconsistency of battery packs and extend their lifespan. This paper proposes a hierarchical equalization topology based on a combination of inductor and transformer, which is divided into intra-group and inter-group battery equalization. The intra-group equalization circuit achieves energy transfer between non-adjacent individual batteries through an improved Buck-Boost circuit. The inter-group equilibrium circuit exchanges energy between any battery subpackages and the entire battery pack using a flyback transformer. An LCD (Inductance-Capacitor-Diode) lossless absorption network is added to the transformer to absorb voltage spikes caused by leakage inductance while achieving soft switch to reduce switch losses. Considering that retired lithium-ion batteries can affect battery capacity and State of Charge (SOC) estimation accuracy due to aging issues, a modified Extended Kalman Filter (EKF) algorithm is proposed to improve battery SOC estimation accuracy. A fuzzy logic controller is introduced into the SOC-based control strategy to improve the equalization efficiency of the battery pack. Finally, to validate the effectiveness of the proposed equalization scheme, a model is built and simulated using MATLAB/Simulink software. The experimental results show that compared with the traditional double-layer inductor equalization scheme using the most value equalization algorithm and fuzzy logic control algorithm respectively, this equalization scheme has a faster equalization speed and higher equalization efficiency.

Predictive Queue-Based Rate Control for Low Latency in Lossless Data Center Networks

Predictive Queue-Based Rate Control for Low Latency in Lossless Data Center Networks

Characterizing Lossy Dielectric Materials in Shock Physics by Millimeter-Wave Interferometry Using One-Dimensional Convolutional Neural Networks and Nonlinear Optimization

When a dielectric material undergoes mechanical impact, it generates a shock wave, causing changes in its refractive index. Recent demonstrations have proven that the modified refractive index can be determined remotely using a millimeter-wave interferometer. However, these demonstrations are based on the resolution of an inverse electromagnetic problem, which assumes that the losses in the material are negligible. This restrictive assumption is overcome in this article, in which a new approach is proposed to solve the inverse electromagnetic problem in lossy and shocked dielectric materials. Our methodology combines a one-dimensional convolutional neural network architecture, namely Inverse problem of Lossless or Lossy Shocked Wavefront Network (ILSW-Net), with a nonlinear optimization technique based on the Nelder–Mead algorithm to estimate losses within dielectric materials under a mechanical impact. Experimental results for both simulated and real signals show that our method can successfully predict the velocities and the refractive index while accurately estimating the shock wavefront.

CLHHN: Category-aware Lossless Heterogeneous Hypergraph Neural Network for Session-based Recommendation

In recent years, session-based recommendation (SBR), which seeks to predict the target user’s next click based on anonymous interaction sequences, has drawn increasing interest for its practicality. The key to completing the SBR task is modeling user intent accurately. Due to the popularity of graph neural networks (GNNs), most state-of-the-art (SOTA) SBR approaches attempt to model user intent from the transitions among items in a session with GNNs. Despite their accomplishments, there are still two limitations. First, most existing SBR approaches utilize limited information from short user–item interaction sequences and suffer from the data sparsity problem of session data. Second, most GNN-based SBR approaches describe pairwise relations between items while neglecting complex and high-order data relations. Although some recent studies based on hypergraph neural networks have been proposed to model complex and high-order relations, they usually output unsatisfactory results due to insufficient relation modeling and information loss. To this end, we propose a category-aware lossless heterogeneous hypergraph neural network (CLHHN) in this article to recommend possible items to the target users by leveraging the category of items. More specifically, we convert each category-aware session sequence with repeated user clicks into a lossless heterogeneous hypergraph consisting of item and category nodes as well as three types of hyperedges, each of which can capture specific relations to reflect various user intents. Then, we design an attention-based lossless hypergraph convolutional network to generate sessionwise and multi-granularity intent-aware item representations. Experiments on three real-world datasets indicate that CLHHN can outperform the SOTA models in making a better tradeoff between prediction performance and training efficiency. An ablation study also demonstrates the necessity of CLHHN’s key components.

Model Self-Adaptive Display for 2D–3D Registration

In the 2D–3D registration process, due to the differences in CAD model sizes, models may be too large to be displayed in full or too small to have obvious features. To address these problems, previous studies have attempted to adjust parameters manually; however, this is imprecise and frequently requires multiple adjustments. Thus, in this paper, we propose the model self-adaptive display of fixed-distance and maximization (MSDFM) algorithm. The uncertainty of the model display affects the storage costs of pose images, and pose images themselves occupy a large amount of storage space; thus, we also propose the storage optimization based on the region of interest (SOBROI) method to reduce storage costs. The proposed MSDFM algorithm retrieves the farthest point of the model and then searches for the maximum pose image of the model display through the farthest point. The algorithm then changes the projection angle until the maximum pose image is maximized within the window. The pose images are then cropped by the proposed SOBROI method to reduce storage costs. By labeling the connected domains in the binary pose image, an external rectangle of the largest connected domain is applied to crop the pose image, which is then saved in the lossless compression portable network image (PNG) format. Experimental results demonstrate that the proposed MSDFM algorithm can automatically adjust models of different sizes. In addition, the results show that the proposed SOBROI method reduces the storage space of pose libraries by at least 89.66% and at most 99.86%.

Revisiting Congestion Detection in Lossless Networks

Congestion detection is the cornerstone of end-to-end congestion control. Through in-depth observations and understandings, we reveal that existing congestion detection mechanisms in mainstream lossless networks (i.e., Converged Enhanced Ethernet and InfiniBand) are improper, due to failing to cognize the interaction between hop-by-hop flow controls and congestion detection behaviors in switches. We define the ternary states of switch ports and present Ternary Congestion Detection (TCD) for mainstream lossless networks. TCD utilizes the ON-OFF sending pattern and the feature of queue length evolutions to detect the transitions among ternary states. We also enable TCD under the practical multiple queues scenario by TCD-MQ. Testbed and extensive simulations demonstrate that TCD can detect congestion ports accurately and identify flows contributing to congestion as well as flows only affected by hop-by-hop flow controls. Meanwhile, we shed light on how to incorporate TCD with rate control. Case studies show that existing congestion control algorithms can achieve 3.3 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\times$</tex-math> </inline-formula> and 2.0 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\times$</tex-math> </inline-formula> better median and 99th-percentile FCT slowdown by combining with TCD.

Research on an Improved Carbon Emission Flow Model Considering Electric Vehicle Charging Fluctuation and Hybrid Power Transaction

With the update of the power transaction mode and the access of increasing electric vehicles with high randomness to the power grid, the existing carbon emission flow calculation method cannot consider the influence of carbon emission due to the fluctuation in electric vehicle charging and various transaction modes. Given the above shortcomings, this paper proposes an improved carbon emission flow model considering electric vehicle charging fluctuations and hybrid power transactions. The model first considers different transaction modes, allocates the network loss to both sides of power generation, and forms a lossless network, realizing the decoupling of the bilateral transaction mode, pool transaction mode, and network loss and then calculating the day-ahead network’s carbon emission. Then, the changes in different transaction modes caused by the fluctuation in electric vehicle charging under different transaction modes are analyzed and the nodes of ‘spontaneous change’ are found to form a day-ahead intra-day deviation network and calculate its carbon emission flow. Finally, the calculation results are combined to obtain an improved power system carbon emission flow considering electric vehicle charging fluctuations and transaction modes. In this paper, the 33-node system is used for verification.

Approximately Lossless Model Compression-Based Multilayer Virtual Network Embedding for Edge–Cloud Collaborative Services

Edge-cloud collaboration integrated with network virtualization is indispensable for diversified edge services. Meanwhile, the multilayer elastic optical network (ML-EON) is a promising underlying network for virtual network requests (VNRs) customized for edge-cloud collaborative services. However, the joint allocation of computing resources and high-dimensional ML-EON resources in virtual network embedding (VNE) will pose great computational complexity for online service deployment. In this paper, we propose an approximately lossless model compression mechanism to ease the computing burden of the VNE over ML-EON for edge-cloud collaborative services. An integer quadratic constraint programming (IQCP) model is established for the problem. Model compression based on Virtual Link Mapping Cost Estimation (VLMCE) is investigated to shield the variables and constraints related to ML-EON. In particular, the resource metric and topology metric are introduced into VLMCE to cope with resource contentions among virtual links in the same VNR and improve estimation accuracy. The model solving relies on Hopfield neural network (HNN) is further studied, where optimizing the compressed model is losslessly converted to minimizing the energy function of HNN. The experimental results reveal that the proposed mechanism guarantees an approximately lossless algorithm performance and a high time efficiency compared with the original IQCP model. The performances of VNR cost and blocking ratio are also promoted compared with the benchmarks.

Effective switchless inter-FPGA memory networks

FPGA network bandwidth increases to 400 Gbps or higher by high-density optical integration, e.g., Co-Packaged Optics (CPO) and onboard Si-photonics transceivers. This study presents its lightweight switchless network architecture by exploiting an indirect path, consisting of up to k one-hop paths for a diameter-k network topology. It integrates interconnection network and memory channels into a lightweight memory-to-memory inter-FPGA network. It has a uni-directional network topology for connecting the largest number of FPGAs. A newly introduced indirect flow control enables lossless uni-directional networks. We present multi-path point-to-point communications, multi-port message-combine collective communications, and efficient WDM (Wavelength Division Multiplexing) cabling method for the uni-directional networks. We implement and evaluate the uni-directional switchless network on an FPGA cluster consisting of six custom Stratix10 MX2100 FPGA cards connected with a 16-wavelength Arrayed Waveguide Grating (AWG) router. Simulation results show that the multi-port message-combine collective communication achieves 7× faster than conventional communication with 272 FPGAs. Effective WDM cabling using 16 wavelengths decreases the number of cables by 88% compared with parallel optical cabling for 272 FPGAs.

Conditional invertible image re-scaling

One of the key issue for image re-scaling is modeling the down-sampling loss. The loss can be approached by a prior distribution through an invertible network in recent works. However, this assumption of independence between images and down-sampling losses is not always satisfied in practice. To address the above problem, we design a module to learn a latent variable representing the down-scaling loss conditional on low-resolution(LR) images according to the information entropy theory. Unlike previous methods, the down-scaling loss is recovered through the proposed module whose inputs are both LR images and predefined distributions. The difference between the high-resolution(HR) image and the low-resolution(LR) image is represented with a latent variable through a lossless invertible neural network. In addition, a conditional entropy loss is proposed to train the invertible neural network to suppress the conditional entropy between LR images and latent variables. It helps to decrease the effect of the latent variable to generate the HR image from LR images during the up-scaling procedure. We evaluate the proposed method on widely used data sets, and the experimental results demonstrate that our proposed method performs favorably against SOTA methods in terms of PSNR and SSIM metrics.

Congestion Control Using In-Network Telemetry for Lossless Datacenters

In the Ethernet lossless Data Center Networks (DCNs) deployed with Priority-based Flow Control (PFC), the head-of-line blocking problem is still difficult to prevent due to PFC triggering under burst traffic scenarios even with the existing congestion control solutions. To address the head-of-line blocking problem of PFC, we propose a new congestion control mechanism. The key point of Congestion Control Using In-Network Telemetry for Lossless Datacenters (ICC) is to use In-Network Telemetry (INT) technology to obtain comprehensive congestion information, which is then fed back to the sender to adjust the sending rate timely and accurately. It is possible to control congestion in time, converge to the target rate quickly, and maintain a near-zero queue length at the switch when using ICC. We conducted Network Simulator-3 (NS-3) simulation experiments to test the ICC’s performance. When compared to Congestion Control for Large-Scale RDMA Deployments (DCQCN), TIMELY: RTT-based Congestion Control for the Datacenter (TIMELY), and Re-architecting Congestion Management in Lossless Ethernet (PCN), ICC effectively reduces PFC pause messages and Flow Completion Time (FCT) by 47%, 56%, 34%, and 15.3×, 14.8×, and 11.2×, respectively.

Efficient Layer Compression without Pruning.

Network pruning is one of the chief means for improving the computational efficiency of Deep Neural Networks (DNNs). Pruning-based methods generally discard network kernels, channels, or layers, which however inevitably will disrupt original well-learned network correlation and thus lead to performance degeneration. In this work, we propose an Efficient Layer Compression (ELC) approach to efficiently compress serial layers by decoupling and merging rather than pruning. Specifically, we first propose a novel decoupling module to decouple the layers, enabling us readily merge serial layers that include both nonlinear and convolutional layers. Then, the decoupled network is losslessly merged based on the equivalent conversion of the parameters. In this way, our ELC can effectively reduce the depth of the network without destroying the correlation of the convolutional layers. To our best knowledge, we are the first to exploit the mergeability of serial convolutional layers for lossless network layer compression. Experimental results conducted on two datasets demonstrate that our method retains superior performance with a FLOPs reduction of 74.1% for VGG-16 and 54.6% for ResNet-56, respectively. In addition, our ELC improves the inference speed by 2× on Jetson AGX Xavier edge device.

Quantum fluctuations in electrical multiport linear systems

We present an extension of the classical Nyquist-Thevenin theorem for multiport classical electrical networks by Twiss to the quantum case. Conversely, we extend the quantum fluctuation-dissipation result for one port electrical systems to the multiport case, both reciprocal and nonreciprocal. Our results are extended to lossy systems by depicting resistive components as continuous limits of purely lossless lumped-element networks. Simple circuit examples are analyzed, including a linear system lacking a direct impedance representation.

On Packet Reordering in Time-Sensitive Networks

Time-sensitive networks (IEEE TSN or IETF DetNet) may tolerate some packet reordering. Re-sequencing buffers are then used to provide in-order delivery, the parameters of which (timeout, buffer size) may affect worst-case delay and delay jitter. There is so far no precise understanding of per-flow reordering metrics nor of the dimensioning of re-sequencing buffers in order to provide worst-case guarantees, as required in such networks. First, we show that a previously proposed per-flow metric, reordering late time offset (RTO), determines the timeout value. If the network is lossless, another previously defined metric, the reordering byte offset (RBO), determines the required buffer. If packet losses cannot be ignored, the required buffer may be larger than RBO, and depends on jitter, an arrival curve of the flow at its source, and the timeout. Then we develop a calculus to compute the RTO for a flow path; the method uses a novel relation with jitter and arrival curve, together with a decomposition of the path into non order-preserving and order-preserving elements. We also analyse the effect of re-sequencing buffers on worst-case delay, jitter and propagation of arrival curves. We show in particular that, in a lossless (but non order-preserving) network, re-sequencing is “for free”, namely, it does not increase worst-case delay nor jitter, whereas in a lossy network, re-sequencing increases the worst-case delay and jitter. We apply the analysis to evaluate the performance impact of placing re-sequencing buffers at intermediate points and illustrate the results on two industrial test cases.

DR-LL Gan: Diabetic Retinopathy Lesions Synthesis using Generative Adversarial Network

— Diabetic Retinopathy (DR) is a serious consequence of diabetes that seriously impact on the eyes and is a leading cause of blindness. If the lesions in DR arise in the central portion of the fundus, they may result in significant vision loss, which we refer to as Diabetic Macular Edema (DME). Deep learning (DL) techniques are commonly used utilized in ophthalmology for discriminative tasks such as diabetic retinopathy or age-related macular degeneration (AMD) diagnosis. Deep learning techniques typically need huge picture data sets for deep convolutional neural networks (DCNNs) training, it should be graded by human specialists. According to international protocol, it is classified into five severity categories. However, improving a grading model for high generality needs a significant quantity of balanced training data, which is challenging to obtain, especially at high levels of severity. Typical techniques for data augmentation, in many applications of deep learning in the retinal image processing domain, the difficulty of access to huge annotated datasets and legal concerns about patient privacy are limiting issues. As a result, the concept of creating synthetic retinal pictures that are indistinguishable from actual data has garnered more attention. GANs have been certain to be an effective framework for creating synthetic databases of anatomically accurate retinal fundus pictures. GANs, in particular, have garnered increasing attention in ophthalmology. in this article, we present a loss-less generative adversarial network (DR-LL GAN) to generate good resolution fundus pictures that May be adjusted to include random grading and information about the lesion. As a result, large-scale generated data may be used to train a DR grading and lesion segmentation model with more appropriate augmentation. Our model experiments evaluated on IDRID and MESSIDOR datasets, it's obtained a discrimination loss of 0.69374 and a generation loss of 1.10438, as well as a segmentation accuracy of 0.9840 in our tests. This might support in the optimization techniques of the neural network design and in computer-aided screening of medical picture, thus increasing diagnostic reliability for clinical assessment in the future of sophisticated technological healthcare.

RIFL: a reliable link layer network protocol for data center communication

More and more latency-sensitive services and applications are being deployed into the data center. Performance can be limited by the high latency of the network interconnect. Because the conventional network stack is designed not only for a local area network, but also for a wide area network, it carries a great amount of redundancy that is not required in a data center network. This paper introduces the concept of a three-layer protocol stack that can fulfill the exact demands of data center network communications. The detailed design and implementation of the first layer of the stack, which we call RIFL, is presented. A novel low latency in-band hop-by-hop re-transmission protocol is proposed and adopted in RIFL, which guarantees lossless transmission in a data center environment. Experimental results show that RIFL achieves 110 ns point-to-point latency on 10 m active optical cables, at a line rate of 112 Gbps. RIFL is a multi-lane protocol with scalable throughput up to multi-hundred gigabits per second. It can be the enabler of low latency, high throughput, flexible, scalable, and lossless data center networks.

A fully matched dual stage CMOS power amplifier with integrated passive linearizer attaining 23 db gain, 40% PAE and 28 DBM OIP3

PurposeThe purpose of this paper is to introduce a new linearization technique known as the passive linearizer technique which does not affect the power added efficiency (PAE) while maintaining a power gain of more than 20 dB for complementary metal oxide semiconductor (CMOS) power amplifier (PA).Design/methodology/approachThe linearization mechanism is executed with an aid of a passive linearizer implemented at the gate of the main amplifier to minimize the effect of Cgs capacitance through the generation of opposite phase response at the main amplifier. The inductor-less output matching network presents an almost lossless output matching network which contributes to high gain, PAE and output power. The linearity performance is improved without the penalty of power consumption, power gain and stability.FindingsWith this topology, the PA delivers more than 20 dB gain for the Bluetooth Low Energy (BLE) Band from 2.4 GHz to 2.5 GHz with a supply headroom of 1.8 V. At the center frequency of 2.45 GHz, the PA exhibits a gain of 23.3 dB with corresponding peak PAE of 40.11% at a maximum output power of 14.3 dBm. At a maximum linear output power of 12.7 dBm, a PAE of 37.3% has been achieved with a peak third order intermodulation product of 28.04 dBm with a power consumption of 50.58 mW. This corresponds to ACLR of – 20 dBc, thus qualifying the PA to operate for BLE operation.Practical implicationsThe proposed technique is able to boost up the efficiency and output power, as well as linearize the PA closer to 1 dB compression point. This reduces the trade-off between linear output power and PAE in CMOS PA design.Originality/valueThe proposed CMOS PA can be integrated comfortably to a BLE transmitter, allowing it to reduce the transceiver’s overall power consumption.

Gamma -Robust electricity market equilibrium models with transmission and generation investments

We consider uncertain robust electricity market equilibrium problems including transmission and generation investments. Electricity market equilibrium modeling has a long tradition but is, in most of the cases, applied in a deterministic setting in which all data of the model are known. Whereas there exist some literature on stochastic equilibrium problems, the field of robust equilibrium models is still in its infancy. We contribute to this new field of research by considering Gamma -robust electricity market equilibrium models on lossless DC networks with transmission and generation investments. We state the nominal market equilibrium problem as a mixed complementarity problem as well as its variational inequality and welfare optimization counterparts. For the latter, we then derive a Gamma -robust formulation and show that it is indeed the counterpart of a market equilibrium problem with robustified player problems. Finally, we present two case studies to gain insights into the general effects of robustification on electricity market models. In particular, our case studies reveal that the transmission system operator tends to act more risk-neutral in the robust setting, whereas generating firms clearly behave more risk-averse.