Pair Of Nodes Research Articles

Orchid: enhancing HPC interconnection networks through infrequent topology reconfiguration

Interconnection networks are key components of high-performance computing (HPC) systems. As HPC evolves towards the exascale era, providing sufficient bisection bandwidth between computing node pairs through oversubscription in traditional networks becomes prohibitively expensive and impractical. Over the past decade, several architectures leveraging optical circuit switches (OCSs) for dynamic link bandwidth allocation have gained traction. These architectures require frequent network topology reconfiguration to adapt to changing traffic demands. However, practical implementation remains hampered by the long reconfiguration delays inherent in OCS technology. We propose Orchid, an architecture that leverages OCSs to achieve infrequent topology reconfigurations, effectively addressing the problem of long reconfiguration delays. A key innovation of Orchid is its ability to extract stable traffic matrices from historical data. This functionality guides the reconfiguration of the topology without the need for adjustments with each traffic matrix, thereby enabling the sharing of OCS overhead over an extended timeframe. Furthermore, Orchid addresses potential congestion arising from unexpected traffic through the joint design of OCS configuration and routing, ensuring an even distribution of traffic across global links. Extensive experiments using real HPC application traces and synthetic traffic demonstrate that Orchid achieves significant performance improvements compared to existing HPC interconnection networks. Specifically, Orchid reduces packet delay by at least 3× and enhances throughput by up to 60%.

Self-attention empowered graph convolutional network for structure learning and node embedding

In representation learning on graph-structured data, many popular graph neural networks (GNNs) fail to capture long-range dependencies, leading to performance degradation. Furthermore, this weakness is magnified when the concerned graph is characterized by heterophily (low homophily). To solve this issue, this paper proposes a novel graph representation learning framework called the graph convolutional network with self-attention (GCN-SA). The proposed scheme exhibits an exceptional generalization capability in node-level representation learning. The proposed GCN-SA contains two enhancements corresponding to edges and node features. For edges, we utilize a self-attention mechanism to design a stable and effective graph-structure-learning module that can capture the internal correlation between any pair of nodes. This graph-structure-learning module can identify reliable neighbors for each node from the entire graph. Regarding the node features, we modify the transformer block to enable it to assist the GCN in integrating valuable information from the entire graph. These two enhancements work in distinct ways to help our GCN-SA capture long-range dependencies, enabling it to perform representation learning on graphs with varying levels of homophily. The experimental results on real-world benchmark datasets demonstrate the effectiveness of the proposed GCN-SA. Compared to other outstanding GNN counterparts, the proposed GCN-SA is competitive. The source code is available at https://github.com/mengyingjiang/GCN-SA.

Ensemble learning using multivariate variational mode decomposition based on the Transformer for multi-step-ahead streamflow forecasting

Accurate streamflow forecasting is critical in the domain of water resources management. However, the inherently non-stationary and stochastic nature of streamflow poses a significant challenge to achieving accuracy in streamflow forecasting. In this study, we introduce an MVMD-ensembled Transformer model (MVMD-Transformer), which incorporates the MVMD for concurrent time–frequency analysis of streamflow and related potential influencing variables. The model aligns common modes in the decomposition results, ensuring that the different variables corresponding to each mode have the same center frequency. This alignment overcomes frequency mismatches and helps uncover the intrinsic patterns and essential features between streamflow and associated variables. During the forecasting phase, the Transformer component of the MVMD-Transformer model establishes connections among streamflow and other influencing variables across pairs of nodes in each mode. We tested the performance of the MVMD-Transformer model in forecasting streamflow across 1-, 3-, 5-, and 7-day horizons within the Shiyang River, Heihe River, and Shule River basins situated in the Hexi Corridor of Northwest China. The MVMD-Transformer model harnesses MVMD to concurrently decompose both predictor variables (precipitation, air temperature, air pressure, soil moisture) and the response variable (streamflow). Subsequently, the modes drived from the MVMD were fed into the Transformer, serving as the predictive analytics engine, to forecast streamflow. Furthermore, we conducted a comprehensive performance evaluation by comparing the MVMD-Transformer model against four alternatives: the VMD-ensembled Transformer model (VMD-Transformer), CEEMDAN-ensembled Transformer model (CEEMDAN-Transformer), stand-alone Transformer model, and LSTM model. The results indicate that MVMD-Transformer outperformed all other models, achieving Nash-Sutcliffe coefficient (NSE) values exceeding 0.85 in the majority of the forecasting scenarios. This superior performance highlights the proficiency of the MVMD approach in more accurately unraveling the intricate interdependencies between streamflow and its various potential influencing variables, thus significantly improving the precision of streamflow forecasting.

Abstract PS03-01: Single cell analysis of paired lymph nodes and primary tumors in breast cancer patients

Abstract A sentinel lymph node (SN) is the primary node draining the tumor and is assumed to be affected early in the metastatic process. The sentinel node holds a key position in the immune response against tumor in breast, and represents a unique connection between the tumor and the host immune response. We hypothesize that the immune profile in the primary tumor and the paired lymph node (LN) is different during tumor progression. 3.7 million single cells from paired primary tumor and lymph nodes from 33 breast cancer patients (Oslo2 cohort) was analyzed by single cell mass cytometry (CyTOF) with a 47 antibody immune panel and characterized by a semi-automatic gating approach (FlowSOM). Tumor cells were found in 11 LN. When analyzing the leucocytes from LN we identified a significant difference in immune cell type skewing towards higher abundance of memory CD4 and CD8 T cells expressing an exhausted phenotype in LN with metastasis. In addition, a higher abundance of activated Tregs and significantly lower abundance of resting Tregs was found in the metastatic LNs compared to the sentinel lymph nodes. The change in immune composition and exhaustion was correlated to the metastatic tumor burden. The skewing towards an exhausted immune profile was also found in larger primary tumors compared to smaller primary tumors. We further analyzed tumor cells from 8 patients with paired primary tumor and LN. No differences were identified in the primary tumor when stratified by LN status, but LNs with smaller metastasis expressed lower levels of epithelial markers essential in the Epithelial-to-Mesenchymal Transition such as E-cadherin, Pan Cytokeratin and EpCAM – this in contrast to the LN with manifested metastasis in the axilla which expressed higher levels of epithelial markers and lower levels of mesenchymal markers such as Vimentin and CD44. We identified a skewing in the immune profile from a naive phenotype towards a memory and exhausted phenotype in CD8 and CD4 T cell population in LN with manifested metastasis, as well as in larger primary tumors. We also identified that tumor cells in smaller metastatic tumors resembled a “mesenchymal like” phenotype compared to in the larger manifested tumors. These results suggest that the immune suppression is correlated with the tumor burden. Citation Format: Marit Fjørtoft, Inga Rye, Kanutte Huse, June Myklebust, Hege Russnes, Margit Riis, Ole Christian Lingjærde, Øystein Garred, Karin Teien Lande, Inger Riise Bergheim, Ellen Schlichting. Single cell analysis of paired lymph nodes and primary tumors in breast cancer patients [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PS03-01.

Photoinduced Floquet higher-order Weyl semimetal in C6 symmetric Dirac semimetals

Abstract Topological Dirac semimetals are a parent state from which other exotic topological phases of matter, such as Weyl semimetals and topological insulators, can emerge. In this study, we investigate a Dirac semimetal possessing sixfold rotational symmetry and hosting higher-order topological hinge Fermi arc states, which is irradiated by circularly polarized light. Our findings reveal that circularly polarized light splits each Dirac node into a pair of Weyl nodes due to the breaking of time-reversal symmetry, resulting in the realization of the Weyl semimetal phase. This Weyl semimetal phase exhibits rich boundary states, including two-dimensional surface Fermi arc states and hinge Fermi arc states confined to six hinges. Furthermore, by adjusting the incident direction of the circularly polarized light, we can control the degree of tilt of the resulting Weyl cones, enabling the realization of different types of Weyl semimetals.

Abstract PO4-25-01: Selective elimination of CD169+ macrophages in lymph nodes invaded by breast cancers

Abstract Lymph node metastasis is a prognostically significant factor in breast cancer, though the colonization mechanism of cancer cells in this immune cell-rich organ remains elusive. This study aimed to elucidate the impact of breast cancer on the lymph node immune cell landscape. Multiscale transcriptomic analyses were performed on both metastatic and non-metastatic lymph node samples from breast cancer patients with lymph node metastasis. Twenty laser-micro-dissected sections were obtained from 17 lymph nodes across six breast cancer patients at stages II–III, with each patient contributing both types of lymph nodes for direct comparison. Comparing the transcriptomes of paired lymph nodes with and without metastasis from the same patients revealed selective downregulation of CD169+ macrophage-related genes in metastatic lymph nodes. The spatial transcriptome indicated a potential depletion of CD169+ macrophages, initiators of anticancer immunity, from their residence (sinuses) in metastatic lymph nodes, while other principal immune cell types were unaltered. Mass spectrometry imaging revealed that the numbers of CD169+ macrophages were smaller in the metastatic lymph nodes than in the non-metastatic lymph nodes, suggesting that cancer cells uprooted CD169+ macrophages from the lymph nodes. Conversely, the count of B, T, Treg, and CD11c+ cells remained comparable in both lymph node types, albeit an enrichment of Treg cells around metastasized cancer tissues was observed. Additional immunohistochemistry analysis of 315 non-metastatic lymph nodes and 159 metastatic lymph nodes from 58 patients with breast cancer showed that a reduced CD169+ macrophage population was prevalent in various breast cancer subtypes. A subset of metastatic lymph nodes (37 out of 159) displayed a complete absence of detectable CD169+ macrophages. The data also depicted a gradual decline in CD169+ macrophages correlating with the pN classification, while no correlation was identified with pathological tumor size classification (pT) or metastasized cancer volume. The data suggest precedence of CD169+ macrophage elimination over other reported immune cell abnormalities, such as cell number and metabolic irregularities. CD169+ macrophages are a unique type of resident macrophages in the lymphoid organs that present cancer-derived antigens to CD8+ cells. The antigen-presenting role of CD169+ macrophages to T cells, a pivotal step in adaptive immunity, signifies the catastrophic implications of their suppression. This study underscores CD169+ macrophage suppression as a pronounced pathological phenotype in lymph nodes with breast cancer metastasis, thereby establishing it as a critical future therapeutic target. Citation Format: Yurina Maeshima, Tatsuki R. Kataoka, Alexis Vandenbon, Masahiro Hirata, Yasuhide Takeuchi, Yutaka Suzuki, Yukiko Fukui, Yumiko Ibi, Hironori Haga, Satoshi Morita, Masakazu Toi, Shinpei Kawaoka, Kosuke Kawaguchi. Selective elimination of CD169+ macrophages in lymph nodes invaded by breast cancers [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-25-01.

Research on Brain Networks of Human Balance Based on Phase Estimation Synchronization.

Phase synchronization serves as an effective method for analyzing the synchronization of electroencephalogram (EEG) signals among brain regions and the dynamic changes of the brain. The purpose of this paper is to study the construction of the functional brain network (FBN) based on phase synchronization, with a special focus on neural processes related to human balance regulation. This paper designed four balance paradigms of different difficulty by blocking vision or proprioception and collected 19-channel EEG signals. Firstly, the EEG sequences are segmented by sliding windows. The phase-locking value (PLV) of core node pairs serves as the phase-screening index to extract the valid data segments, which are recombined into new EEG sequences. Subsequently, the multichannel weighted phase lag index (wPLI) is calculated based on the new EEG sequences to construct the FBN. The experimental results show that due to the randomness of the time points of body balance adjustment, the degree of phase synchronization of the datasets screened by PLV is more obvious, improving the effective information expression of the subsequent EEG data segments. The FBN topological structures of the wPLI show that the connectivity of various brain regions changes structurally as the difficulty of human balance tasks increases. The frontal lobe area is the core brain region for information integration. When vision or proprioception is obstructed, the EEG synchronization level of the corresponding occipital lobe area or central area decreases. The synchronization level of the frontal lobe area increases, which strengthens the synergistic effect among the brain regions and compensates for the imbalanced response caused by the lack of sensory information. These results show the brain regional characteristics of the process of human balance regulation under different balance paradigms, providing new insights into endogenous neural mechanisms of standing balance and methods of constructing brain networks.

Topological broadband invisibility

AbstractWhen an electromagnetic wave is incident onto an object, the scattering from the object leads to its exposure. An invisibility cloak can bend the electromagnetic wave around the object with phase velocities exceeding the light speed in free space, which seems only possible over a narrow bandwidth, as manifested in the existing approaches, such as transformation optics and scattering cancellation. Here, it is experimentally demonstrated that this bandwidth limitation can be overcome by strategically positioning objects at topological nodes, characterized by minimal electromagnetic field amplitudes and indeterminate phases. This is accomplished using all‐dielectric photonic crystal slabs, which are engineered to exhibit a pair of topological nodes at fixed planes across an extensive bandwidth. This bandwidth expansion is facilitated not by resonance, but by the enforcement of mirror and time‐reversal symmetry. These findings thus introduce a novel topological paradigm for the broadband invisibility devices.

Message-passing on hypergraphs: detectability, phase transitions and higher-order information

Hypergraphs are widely adopted tools to examine systems with higher-order interactions. Despite recent advancements in methods for community detection in these systems, we still lack a theoretical analysis of their detectability limits. Here, we derive closed-form bounds for community detection in hypergraphs. Using a message-passing formulation, we demonstrate that detectability depends on the hypergraphs’ structural properties, such as the distribution of hyperedge sizes or their assortativity. Our formulation enables a characterization of the entropy of a hypergraph in relation to that of its clique expansion, showing that community detection is enhanced when hyperedges highly overlap on pairs of nodes. We develop an efficient message-passing algorithm to learn communities and model parameters on large systems. Additionally, we devise an exact sampling routine to generate synthetic data from our probabilistic model. Using these methods, we numerically investigate the boundaries of community detection in synthetic datasets, and extract communities from real systems. Our results extend our understanding of the limits of community detection in hypergraphs and introduce flexible mathematical tools to study systems with higher-order interactions.

BRAF mutation, selected miRNAs and genes expression in primary papillary thyroid carcinomas and local lymph node metastases

Mutations in cancer-related genes are now known to be accompanied by epigenetic events in carcinogenesis by modification of the regulatory pathways and expression of genes involved in the pathobiology. Such cancer-related mutations, miRNAs and gene expression may be promising molecular markers of the most common papillary thyroid carcinoma (PTC). However, there are limited data on their relationships. The aim of this study was to analyse the interactions between BRAF mutations, selected microRNAs (miR-21, miR-34a, miR-146b, and miR-9) and the expression of selected genes (LGALS3, NKX2–1, TACSTD2, TPO) involved in the pathogenesis of PTC. The study cohort included 60 primary papillary thyroid carcinomas (PTC) that were classified as classical (PTC/C; n=50) and invasive follicular variant (PTC/F; n=10), and 40 paired lymph node metastases (LNM). BRAF mutation status in primary and recurrent/persistent papillary thyroid carcinomas was determined. The mutation results were compared both between primary and metastatic cancer tissue, and between BRAF mutation status and selected genes and miRNA expression in primary PTC. Furthermore, miRNAs and gene expression were compared between primary PTCs and non-neoplastic tissue, and local lymph node metastatic tumor, respectively. All studied markers showed several significant mutual interactions and contexts. In conclusion, to the best our knowledge, this is the first integrated study of BRAF mutational status, the expression levels of mRNAs of selected genes and miRNAs in primary PTC, and paired LNM.

Detecting Overlapping Communities Based on Influence-Spreading Matrix and Local Maxima of a Quality Function

Community detection is a widely studied topic in network structure analysis. We propose a community detection method based on the search for the local maxima of an objective function. This objective function reflects the quality of candidate communities in the network structure. The objective function can be constructed from a probability matrix that describes interactions in a network. Different models, such as network structure models and network flow models, can be used to build the probability matrix, and it acts as a link between network models and community detection models. In our influence-spreading model, the probability matrix is called an influence-spreading matrix, which describes the directed influence between all pairs of nodes in the network. By using the local maxima of an objective function, our method can standardise and help in comparing different definitions and approaches of community detection. Our proposed approach can detect overlapping and hierarchical communities and their building blocks within a network. To compare different structures in the network, we define a cohesion measure. The objective function can be expressed as a sum of these cohesion measures. We also discuss the probability of community formation to analyse a different aspect of group behaviour in a network. It is essential to recognise that this concept is separate from the notion of community cohesion, which emphasises the need for varying objective functions in different applications. Furthermore, we demonstrate that normalising objective functions by the size of detected communities can alter their rankings.

GAN-based self-supervised message passing graph representation learning

Graph representation learning aims to learn embedding representations of nodes in graph-structured data through supervised or unsupervised methods. However, for unlabeled or sparsely labeled graph datasets, conventional supervised approaches fail to perform adequately, and existing unsupervised methods have difficulty achieving high accuracy on multiple downstream tasks such as node classification and link prediction at lower embedding dimensions. To address this issue, we propose a novel self-supervised graph representation learning approach called GAN-based Message Passing Graph Representation Learning. This approach significantly improves the accuracy of node embeddings without relying on label information. Specifically, GMP-GL has developed a generator based on neighbor-weighted random walk to obtain positive and negative node sample pairs, as well as a discriminator that distinguishes whether a real connection exists between node pairs. Through adversarial learning, GMP-GL automatically learns node embeddings. Furthermore, GMP-GL optimizes the training outcomes of the generative adversarial model through self-supervised embedding aggregation learning, thus introducing rich structural information from the graph for the subsequent rounds of generative adversarial training. In addition, GMP-GL employs multiple autoencoders to generate and filter original embeddings, ensuring that the pre-embeddings preserve a significant amount of original feature information for formal learning. The study selects six real-world datasets and conducts comparative experiments with ten SOTA baseline methods. The experimental results demonstrate that the proposed model not only surpasses previous unsupervised methods but also matches or even exceeds the performance of supervised methods. This indicates the great potential of the proposed approach for representation learning tasks on unlabeled graph datasets and its effective applicability to real-world problems.

A 45nm AND 65nm CMOS IMPLEMENTATION OF 12T SRAM CELL WITH SOFT ERROR READ STABILITY OF MULTINODE UPSET RECOVERABILITY FOR AEROSPACE APPLICATION

With technology advancing rapidly, the size of transistors and the distance between them is decreasing quickly. This reduction in size is causing the critical charge of sensitive nodes in SRAM cells to decrease, making them more susceptible to soft errors, especially in aerospace applications. When a radiation particle strikes a open node in a standard 6T SRAM cell, the storage data in the cell can be changed, leading to a single-event upset (SEU). To address this issue, a new Soft Error Read and write Stability improves Low Power 12T (SARP12T) SRAM cell has been introduced in this study to reduce the impact of SEUs. The performance of SARP12T has been compared to other recently developed soft-error-aware SRAM cells like QUCCE12T, QUATRO12T, RHD12T, RHPD12T, and RSP14T. SARP12T is designed for the recoverability of all stored data in sensitive nodes even if they are affected by radiation strikes. Additionally, SARP12T can overcome single-event multi-node upsets (SEMNUs) that may happened at its sensitive node pair. This proposed cell also boasts the higher read stability, as the storage node holding '0' can regain from any upset during read operations. Moreover, SARP12T consumes the minimum amount of hold power, has greater write ability, and lower write delay compared to other similar cells. Despite a almost greater read delay and higher read and write energy consumption, the improvements in the proposed SARP12T cell make it a promising solution for mitigating soft errors in SRAM cells

Quantifying intragroup and intergroup connections in non-disjoint groups in social networks: A comprehensive analysis incorporating edges' and nodes' weights

In this paper, we propose generalizations of the EI index to quantify the distinction between intragroup and intergroup connections in social networks. These generalizations enable the analysis of interactions between non-disjoint groups, as large-scale disjoint groups are rarely found in many empirical networks. By expanding social network analysis, these measures facilitate the identification of actors with greater similarities or differences, generating previously untapped knowledge. Furthermore, we propose incorporating both edges' and nodes' weights in assessing the EI index. Intuitively, our approach classifies each network edge as internal or external, depending on the attribute groups that each node linked by the given edge belongs to. Since nodes can belong to more than one attribute group, three methods of classifying a network edge are proposed which depend on the number of shared groups by the pair of nodes. We evaluate the new measures in two distinct network contexts. The first context involves a co-authorship network, where researchers, acting as network actors, are categorized based on their respective areas of expertise in management engineering. The second network consists of trade relations among countries in the Americas, with countries grouped according to their participation in trade agreements. The analysis showed which area of management engineering is more independent by having more internal connections and made it possible to investigate the role of trade agreements in the actual trades made by countries. These proposals provide a more comprehensive and nuanced understanding of network dynamics, allowing for a more precise comprehension of interactions and patterns present in social systems.

A two-phase spatiotemporal chaos-based protocol for data integrity in IoT

One of the biggest problems with Internet of Things (IoT) applications in the real world is ensuring data integrity. This problem becomes increasingly significant as IoT expands quickly across a variety of industries. This study presents a brand-new data integrity methodology for Internet of Things applications. The “sequence sharing” and “data exchange” stages of the suggested protocol are divided into two parts. During the first phase, each pair of nodes uses a new chaotic model for securely exchanging their identity information to generate a common sequence. This phase’s objectives include user authentication and timing calculations for the second phase of the recommended method’s packet validation phase. The recommended approach was tested in numerous settings, and various analyses were taken into account to guarantee its effectiveness. Also, the results were compared with the conventional data integrity control protocol of IoT. According to the results, the proposed method is an efficient and cost-effective integrity-ensuring mechanism with eliminates the need for third-party auditors and leads to reducing energy consumption and packet overhead. The results also show that the suggested approach is safe against a variety of threats and may be used as a successful integrity control mechanism in practical applications.

Enhancing Wireless Sensor Network Security Through Mutual Information Analysis for Intrusion Detection and Resilience

Intrusion detection is a critical aspect of network security, involving the identification and response to unauthorized access or malicious activities within a system. It plays a crucial role in safeguarding networks and ensuring data integrity and confidentiality. The Mutual Information (MI) analysis is a vital component in the realm of intrusion detection and security within Wireless Sensor Networks (WSNs). By assessing the information shared between pairs of sensor nodes, MI analysis reveals underlying patterns and relationships in network data. This process involves calculating the mutual information for each node pair based on statistical properties of observed values. Strong correlations or dependencies between nodes are identified, aiding in the detection of critical nodes or clusters vulnerable to compromise. Additionally, MI analysis informs feature extraction by guiding the selection of informative features that capture network structure. It also serves as a proactive tool for identifying anomalies or deviations from expected patterns, which may signal intrusion attempts or malicious activities. Through monitoring changes in mutual information over time, MI analysis facilitates prompt responses to emerging threats, enhancing the resilience and security of WSNs. The work exhibits high accuracy, recall rates, and detection rates across various attack scenarios, underscoring its efficacy in identifying and mitigating security threats. The algorithm's efficiency, effectiveness, and reliability make it a promising solution for enhancing WSN security. Through sophisticated methodologies and adaptive defensive strategies, the proposed algorithm strengthens the robustness and dependability of WSNs, minimizing the risk of potential security vulnerabilities and ensuring comprehensive threat detection in real-world deployment scenarios.

Encounter Network and Its Application to Studying Epidemic Spread through Public Transportation Networks

Epidemic diseases have always been a challenging issue in public health. Densely populated urban public transportation systems could be the perfect candidate for transmitting epidemics by direct contact with the infected population and contaminated surfaces. In network analysis, betweenness centrality (BC), as the primary metric, has been used to measure an individual’s communication in the network. This research uses this concept to measure the risk of transmitting disease in public transportation links. For each network node, the set of shortest paths from that node to all other nodes compose a shortest paths sub-graph (tree). Source node total betweenness centrality (TBC) is defined as the sum of BC of all edges of this tree. Similarly, the shortest path sub-graph consisting of the set of the shortest paths from all nodes to a single target node is considered, and the target node TBC is defined as the sum of BC of all edges of this tree. The node encounter matrix is also defined to measure each node pair encounter in the network. The source node TBC of a node depicts the risk of transmitting the disease from that node to other nodes. Conversely, the target node TBC of a node is related to the risk of transmitting the disease from other nodes to that node. The susceptible-infected model is used to evaluate the proposed method by measuring the number of infected nodes.

Upgrading strategies for long-haul quantum key distribution networks

Quantum key distribution (QKD) has become a promising option for transmitting sensitive data due to the increased maturity of QKD devices and the threat scalable quantum computers imposes on asymmetric public-key cryptosystems. By utilizing existing infrastructure, e.g., amplifier huts and dark fibers, the cost and complexity of deploying QKD networks (QKDNs) can be reduced. In this study, we develop a topology optimization algorithm that minimizes the cost of the QKDN deployment and maximizes the key capacity between any pair of nodes. We present three deployment upgrade strategies for facilitating the encryption of increasing throughput of QKD-secured optical transport networks and enabling a pay-as-you-grow approach. Comparing different strategies in a capacity-planning study allows operators to assess the scalability of deployments and upgrades. Depending on the availability requirements, our results compare the cost of protection measures. We confirm that adding trusted nodes is the most cost-efficient upgrading strategy based on our analysis.

Local and global optimization methods for optical line control based on quality of transmission

The ever-increasing demand for data traffic in recent decades has pushed network operators to give importance to the aspect of infrastructure control to facilitate its scalability and maximize its capacity. A generic lightpath (LP) is deployed starting from a traffic request between a given pair of nodes in a network. LPs are operated in the network based on an estimate of the quality of transmission (QoT), which is derived from the physical layer characteristics of a selected route. Regardless of the model used to estimate QoT, it is necessary to calibrate the model to maximize its accuracy and define minimum design margins. The model calibration process depends significantly on the type of data that can be collected in the field (i.e., type of metric, resolution) and therefore on the available monitoring devices. In this work, a systematic evaluation of the QoT estimation is carried out on a multi-span erbium-doped-fiber-amplified optical line system (OLS) using in the first case only total power monitors and in the second experimentally emulating optical channel monitors (OCMs). Given the type of monitoring devices available, three different physical models are calibrated, and six optimization methods are used to define the optimal configuration of the target gain and tilt parameters of the optical amplifiers, jointly optimizing the working point of all amplifiers (global approach) or proceeding span by span (local approach). Subsequently, the OLS was set in each configuration obtained, and the generalized signal-to-noise ratio (GSNR) profile was measured at the end.

GUIDE: Training Deep Graph Neural Networks via Guided Dropout Over Edges.

Graph neural networks (GNNs) have made great progress in graph-based semi-supervised learning (GSSL). However, most existing GNNs are confronted with the oversmoothing issue that limits their expressive ability. A key factor that leads to this problem is the excessive aggregation of information from other classes when updating the node representation. To alleviate this limitation, we propose an effective method called GUIded Dropout over Edges (GUIDE) for training deep GNNs. The core of the method is to reduce the influence of nodes from other classes by removing a certain number of inter-class edges. In GUIDE, we drop edges according to the edge strength, which is defined as the time an edge acts as a bridge along the shortest path between node pairs. We find that the stronger the edge strength, the more likely it is to be an inter-class edge. In this way, GUIDE can drop more inter-class edges and keep more intra-class edges. Therefore, nodes in the same community or class are more similar, whereas different classes are more separated in the embedded space. In addition, we perform some theoretical analysis of the proposed method, which explains why it is effective in alleviating the oversmoothing problem. To validate its rationality and effectiveness, we conduct experiments on six public benchmarks with different GNNs backbones. Experimental results demonstrate that GUIDE consistently outperforms state-of-the-art methods in both shallow and deep GNNs.