Optical circuit switching (OCS) is becoming used in some data center networks due to its low power consumption, low latency, and high bandwidth. Previous research introduced a design model for a twisted and folded Clos network (TF-Clos) as a data center network to maximize the switching network size, i.e., the number of connected terminals, while guaranteeing the admissible blocking probability. The previous model assumes that request active rates from all the terminals are identical. However, it is an overly conservative design when the active rates differ, resulting in a smaller switching network size than desired. This paper proposes a terminal-shuffling (TS) scheme for designing an OCS TF-Clos network with an admissible blocking probability guarantee, which supports different active rates. Each terminal can arbitrarily choose any leaf switch to connect, making the network design more adaptable to varying conditions. A patch panel or direct termination by operators can wire optical fibers between the terminals and the leaf switches. We formulate a TS-based TF-Clos design problem to maximize the switching network size. We develop an approximation approach to find a feasible solution to the optimization problem. Numerical results demonstrate that the switching network size of the proposed TS scheme is larger than that of baseline schemes.

The introduction of optical-circuit-switches (OCSes) has enabled the implementation of capacity- and energy-efficient networks in production datacenters. To correctly operate optical-circuit-switched datacenter networks (OCS DCNs), fibers between pairs of terminals (e.g., servers or top-of-rack switches) and OCSes should be verified before starting operations. However, this task is difficult because OCSes cannot use topology discovery or link monitoring functions, which are only available on electrical packet switches. Motivated by this challenge, we investigated a fiber topology and quality verification (FTQV) problem for OCS DCNs in this paper. Though a previous study inspected fibers in hierarchical OCS DCNs using only one dedicated tester for fiber probing, making the process time-consuming, we consider using digital diagnostic monitoring (DDM) functions at multiple transceivers for fiber inspection. We thus developed solid theories for correctly and quickly inspecting fibers even when multiple probes are sent in parallel. We also developed an algorithm that correctly and quickly solves the FTQV problem on the basis of our theories. Numerical experiments showed that our algorithm completes FTQV at most 48.7 times faster than a baseline algorithm.

Evaluating network reliability is crucial for designing communication infrastructures, particularly for such next-generation networks as 6 G, which require extremely high reliability levels, often reaching seven 9's (99.99999 %). However, network reliability evaluation is computationally tough, making it computationally intractable to achieve accurate results for large-scale networks within a reasonable time. Consequently, existing methods either sacrifice computational efficiency or lack guaranteed error bounds. This paper presents an efficient approximation method for network reliability evaluation that guarantees that the approximation error remains within a specified bound. Our proposed method first determines the maximum degree of simultaneous link failures that can be safely ignored without exceeding the bound. With binary decision diagrams, our method efficiently computes both the lower and upper bounds of network reliability. We validate our method through numerical experiments on real-world networks. Our method, which outperformed existing methods by several orders of magnitude, evaluated a network with 434 links in just 7 seconds; an existing method required more than 28 hours.

Data center networks facilitate large-scale data processing by interconnecting multiple switching devices. Optical circuit switching (OCS) provides high transmission capacity and energy efficiency. It establishes dedicated paths for data transfer, ensuring reliable communication. A Clos network is widely used among multi-stage switching architectures due to its scalability and structured design. This paper investigates models for designing folded/unfolded Clos networks with an admissible blocking probability to maximize OCS network size. While previous studies have examined fundamental and stage-extended Clos networks, they have not addressed unfolded Clos network structures that maintain an admissible blocking probability across different configurations. To fill this gap, we introduce unfolded Clos network structures that ensure an admissible blocking probability for both fundamental and extended stages. We also discuss connection admission control mechanisms tailored to these network models. A key focus of this study is a comprehensive performance evaluation, including switching network size and computation time. Furthermore, we explore an alternative approach employing multiple network planes to enhance scalability and flexibility. The findings provide valuable insights into the design of large-scale OCS networks with controlled blocking probabilities.

Future advancements in data centers are anticipated to incorporate advanced circuit switching technologies, especially optical switching, which achieve high transmission capacity and energy efficiency. Previous studies addressed a Clos-network design problem to guarantee an admissible blocking probability to maximize the switching capacity, which is defined by the number of terminals connected to the network. However, as the number of available${N} \times {N}$switches increases, the switching capacity no longer increases due to the switch port limitation. This paper proposes a design of a multiple-plane twisted-folded (TF) Clos network, named MP-TF, to enhance the switching capacity, which is limited by the original TF-Clos, by guaranteeing an admissible blocking probability. MP-TF consists of identical M TF-Clos planes and pairs of a$1\times {M}$selector and an${M} \times 1$selector, each pair of which is associated with a transmitter and receiver pair. We formulate a design model of MP-TF as an optimization problem to maximize the switching capacity. We introduce connection admission control in MP-TF, named MP-CAC. We derive the theorem that the MP-TF design model using MP-CAC guarantees the admissible blocking probability. Numerical results observe that MP-TF increases the switching capacity as the number of TF-Clos planes when available${N} \times {N}$switches are sufficient; for example, with seven planes, the switching capacity is 1.97 times larger than that of one plane, given a request active probability of 0.6 and an admissible blocking probability of 0.01. We find that the computation time for MP-TF diminishes with an increase in the number of TF-Clos planes. Designing MP-TF is similar to designing a single TF-Clos plane, differing mainly in the handling of connection admission control. With a larger number of${N} \times {N}$switches, MP-TF enables the design of a smaller TF-Clos plane. We provide the analyses of optical power management and network cost of MP-TF.

Modern society is underpinned by several network infrastructures, such as telecommunications and transportation. In these network infrastructures, every node pair should remain connected even if some network components fail. Thus, network operators must accurately evaluate the reliability between each node pair and effectively enhance unreliable pairs. Unfortunately, network reliability evaluation is known to be computationally difficult even for a single pair, and no accurate and efficient algorithm for all pairs has been developed. This paper introduces the problem of 2-NR++, evaluating network reliability for all node pairs, and proposes an efficient and exact algorithm. Unlike the previous algorithms, the proposed algorithm constructs only one data structure to compute the reliability of all node pairs. Numerical evaluation using real network benchmarks shows that our algorithm is faster than the state-of-the-art by an order of magnitude; e.g., our algorithm solves 2-NR++ in less than one hour even for a large network with 240 nodes, whereas the state-of-the-art requires more than one day. As an application of our algorithm, we also introduce a link augmentation problem to improve the least unreliable pair. We demonstrate that our algorithm reduces the computation time by around 30–90 times compared to the state-of-the-art.

As the reliability requirements increase for communication networks, e.g., the seven 9’s reliability for 6G, network operators must evaluate the reliability of their networks more accurately. Since public-communication networks are designed to be less prone to large-scale outages, their (un)reliability should be evaluated per outage scale (the number of disconnected nodes). Traditionally, scale-wise unreliability has been evaluated approximately or requires several hours due to its computational hardness. This paper proposes an efficient algorithm that exactly evaluates scale-wise unreliability by leveraging the given reliability requirements for computation acceleration. We first define a sub-problem that computes the probability of disconnecting x or more nodes and introduce a core algorithm that solves this problem with binary decision diagrams. The core algorithm is used to evaluate the scale-wise unreliability according to the given requirements, e.g., the outage probability or outage scale. The time complexity is theoretically analyzed, and its performance is experimentally verified. The proposed algorithm successfully evaluates in 82 minutes large benchmark networks with 400 links and the seven 9’s requirements; the state-of-the-art algorithm fails to evaluate it within 12 hours.

Low earth orbit (LEO) satellite constellations are a promising system to enable high-throughput space networks, but they also pose us many challenges. From a networking viewpoint, how to select appropriate optical inter-satellite links (OISLs) and maintain high network performance is an important problem. Regarding this “dynamic topology control”, how to do this particularly for non-uniform traffic conditions remains an open issue. In non-uniform traffic conditions, the link layer requires information from not only the physical layer as considered in current topology design but also the routing layer, because the non-uniform traffic distribution poses non-uniform congestion as a result of routing. Therefore, this paper describes a multilayer topology control model comprising a link and routing layers and proposes, as the main contribution, a link control algorithm for non-uniform traffic. In the proposed algorithm, each link has a link weight that reflects the link utilization after future route assignment, and links with high weights are preferentially selected so that links exist in the location where traffic will be concentrated in the future. We also introduce, as the secondary contribution, a hierarchical topology model: the multilayer model was formulated on the basis of this topology model to describe feedback from the routing layer and to enable us to evaluate system independent of constellation models such as Walker Star and Delta. Numerical experiments showed that the proposed method accommodated 7–17% more traffic than the baseline method over enough operational time and handled almost double the traffic during peak periods.

Future data centers are anticipated to embrace cutting-edge circuit switching technologies, particularly optical switching, renowned for their heightened transmission capacity and energy efficiency. Optical circuit switching guarantees consistent communication quality by establishing dedicated connections for data transmission and maintaining their integrity. Data centers favor Clos-network structures due to their scalability. This paper examines the advancements in designing Clos networks to boost switching capacity while maintaining internal blocking quality. We analyze the characteristics of Clos network design models, providing essentials. We extensively discuss their performances regarding their switching capacities and computation times. Drawing from the review of research progress, we address the challenges in enhancing the switching capacity of Clos networks for future studies.

Acquiring purpose-specific data is crucial for applying deep learning to applications. However, creating such data can be labor-intensive, necessitating the development of datasets at minimal cost. Consequently, there has been a recent trend towards utilizing computer-generated imagery (CGI) to generate purpose-specific datasets. Nonetheless, models trained on CGI data often exhibit suboptimal performance on real images due to domain gap issues. Thus, bridging this gap between synthetic and real data domains is crucial. We are developing a real parts classification application using 3D CAD models as training data. This study discusses our strategies for addressing the domain gap challenge in this application. To mitigate this gap, we explore three primary approaches: selecting architectures with a strong shape bias, leveraging large-scale pre-training to enhance generalization performance, and enhancing the shape bias of classifiers by manipulating color characteristics in the training data. We present evidence demonstrating the effectiveness of these meth-ods in bridging the domain gap between CG I training data and real-world images in our part classification application: the top-l accuracy became 98.34 % at the best case.

Machine Learning (ML) and Artificial Intelligence (AI) can harness the immense amount of operational data from clouds to services, to social and communication networks. In the era of data science and connected devices of all varieties, Intelligence have found ways to improve operations and management of next generation networks, systems, and services. Further research is therefore needed to understand and improve the potential and suitability of ML/AI in the context of network, system, and service operations and management. This will provide deeper understanding and better decision making based on largely collected and available operational and management data. It will also present opportunities for improving ML/AI algorithms on aspects such as reliability, dependability, and scalability, as well as demonstrate the benefits of these methods in control and management systems. Moreover, there is an opportunity to define novel platforms that can harness the vast operational data and advance ML/AI algorithms to drive management decisions in open and highly programmable networks, clouds, and data centers.

Data center networks (DCNs) consisting of optical-circuit switches (OCSes) have been intensively studied due to optical transmission's high capacity and energy efficiency. Since current DCNs consist of packet switches, the condition and cabling of optical fibers can be inspected easily by probing neighboring switches. However, OCS networks cannot be inspected in the same manner because OCSes only pass through optical signals. We have had to attach and detach a tester device to every switch for probing all the fibers, which is very time-consuming. This paper proposes a method for automatically inspecting fibers in an entire DCN without repeating tester reattachment. Our method is based on (1) theories on quickly estimating the fiber condition on the basis of the intensity of received probe signals and involves (2) an algorithm that reduces the number of probes needed. Numerical evaluation showed that our method can be used to inspect a huge DCN with 32,000 fibers in at most 2 days, whereas a baseline method involving repeated tester reattachment would take 2 weeks. An experiment using actual OCSes was also conducted to confirm the feasibility of our method.

Optical communication networks are operated with great care to avoid outages. In particular, significant outages that have a devastating impact on society must be avoided, and several countries specify “significance standards” for network outages. Although the impact of possible outages strongly depends on routing methods for requested connections, there has been no existing work that directly deals with these standards. This paper for the first time introduces a routing problem to minimize the outage impact on a given significance standard. The problem is shown to be NP-hard, and an efficient heuristic method is proposed. This method searches for a good solution by considering the problem as a shortest-path search, where the significance of a link failure is regarded as the link weight. Numerical experiments showed that the proposed method reduces significant outages by 30–40% compared with the conventional shortest-path method and 10-20% compared with the most-reliable-path method.

In communication networks, the significance of an outage is measured mainly by its scale (number of disconnected nodes). To avoid serious outages, operators design their networks so that the reliability meets the specification for each outage scale, where the more significant the outage, the less likely it is to occur. Although scale-wise unreliability has been evaluated with rough approximation, sixth-generation (6G) mobile communication requires more accurate reliability evaluation with seven 9's accuracy. Unfortunately, accurate scale-wise reliability evaluation is a computationally very tough problem, so no previous literature has studied evaluation methods rigorous enough for 6G. This paper proposes an efficient algorithm to exactly compute the probability for each number of disconnected nodes. Our algorithm performs the scale-wise unreliability evaluation in a dynamic programming manner without redundant repetition for each outage scale. Numerical experiments using real network topologies show its great efficiency, e.g., our algorithm computes exact probabilities for every outage scale in just two hours for a network with nearly 200 links. We also provide several interesting insights on the reliability of real topologies from the scale-wise perspective, since our work is the first to present the scale-wise unreliability of real large topologies.

Contemporary society survives on several network infrastructures, such as communication and transportation. These network infrastructures are required to keep all nodes connected, although these nodes are occasionally disconnected due to failures. Thus, the expected number of connected node pairs (ECP) during an operation period is a reasonable reliability measure in network design. However, no work has studied ECP due to its computational hardness; we have to solve the reliability evaluation problem, which is a computationally tough problem, for O(n2) times where n is the number of nodes in a network. This paper proposes an efficient method that exactly computes ECP. Our method performs dynamic programming just once without explicit repetition for each node pair and obtains an exact ECP value weighted by the number of users at each node. A thorough complexity analysis reveals that our method is faster than an existing reliability evaluation method, which can be transferred to ECP computation, by O(n). Numerical experiments using real topologies show great efficiency; e.g., our method computes the ECP of an 821-link network in ten seconds; the existing method cannot complete it in an hour. This paper also presents two applications: critical link identification and optimal resource (e.g., a server) placement.

Multispectral sensors are used to ensure visibility in various applications. However, when multiple sensors are used for capturing images, a misalignment may occur between the images taken by each sensor unless special care is taken. To correct such misalignments, image registration based on feature matching is conducted. However, the features captured by each sensor differ, thereby complicating the registration process. In this study, we develop an approach to overcome these challenges and to improve the registration accuracy between short-wave infrared and mid-wave infrared (SWIR and MWIR, respectively) images. First, we compare and validate SiLK, a detector-based feature matching method, and LoFTR, a detector-free feature matching method. The results clearly demonstrate the superior accuracy of LoFTR. Moreover, SWIR and MWIR images exhibit a characteristic color inversion according to Kirchhoff’ s law. Therefore, by inverting the color of a SWIR image and aligning the color tone between image pairs, we can improve the matching accuracy. Furthermore, by diversifying the color tones of the training data through augmentation, we can handle the domain gap between SWIR and MWIR images, thereby further enhancing the matching accuracy.

The Clos network has been studied for decades as a class of nonblocking switching networks. However, the structure is based on assumptions made at the time of its design, and the assumptions may not remain entirely valid. For instance, an optimal Clos network usually consists of non-square switches (numbers of input and output ports differ), though the switches available today are square ones. In addition, all paths have to be of the same length in Clos networks (presumably to simplify signal setting), but this assumption is no longer valid since signals can be dynamically established nowadays. This paper carefully identifies the implicit assumptions of the Clos network and redesigns it to increase its capacity. Although the conventional Clos network using square switches has to leave several ports unused to realize the nonblocking property, our network has almost no unused ports, which greatly increases network capacity. In addition, our network does not fix the path length and well utilizes shortcut connections if available. Comprehensive theoretical analyses show that our network has a larger capacity than the Clos network under most conditions. Numerical evaluations demonstrate that our network increases the capacity 15% on average and up to 50% at most.

Machine learning and artificial intelligence can harness the immense stream of operational data from clouds, to services, to social and communication networks. In the era of big data and connected devices of all varieties, machine learning and artificial intelligence have found ways to improve operations and management of information technology and communications.

Network reliability, that is, the probability of con-necting specified nodes under link failures, is a key metric of network infrastructure. Because network reliability evaluation is a computationally heavy task, past research has relied on unre-alistically simple failure models such as the independent failure model, wherein each link fails stochastically and independently, ignoring large-scale failures such as disasters, or the deterministic-correlated failure model, wherein all links within a disaster area always fail. However, actual networks follow the probabilistic-correlated (PC) failure model, wherein links in a disaster area fail stochastically with respect to each disaster. This paper proposes an efficient method to accurately compute network reliability under the PC model. Following a conventional method for the independent model, the proposed method uses binary decision diagrams (BDDs) to efficiently handle an exponential number of failure states. Additionally, it employs a probabilistic inference technique to support probabilistic correlation, which is represented as another BDD for integration with the conventional method. The computational complexity was theoretically analyzed, and its performance was experimentally verified; it can compute the network reliability within 1 h for a large network with nearly 200 links and 100 potential disasters.

Since modern society largely depends on several network infrastructures such as telecommunications and various types of power, the reliability of networked systems needs to be accurately evaluated. Traditionally, network reliability has been evaluated on the assumption that the availability of each link is precisely given. However, in reality, it may be given with a degree of uncertainty, e.g., with variance, which must be propagated into the network reliability. To the best of our knowledge, no literature has investigated this issue because, since even computing the reliability itself belongs to a computationally tough class, computing the variance seems much harder.This paper proposes an efficient algorithm to compute the variance of the network reliability given the variance in link availability. We experimentally verify the performance of our algorithm and show that it can compute the variance of network reliability within 0.1 seconds for real topologies with nearly 200 links. We also perform extensive analyses on the variance of network reliability and reveal that network reliability tends to be accurate and that the variance in reliability does not significantly exceed the variance in link availability. Even when some links have a substantial variance in availability, the impact on network reliability is marginal.

Machine learning and artificial intelligence can harness the immense stream of operational data from clouds, to services, to social and communication networks. In the era of big data and connected devices of all varieties, machine learning and artificial intelligence have found ways to improve operations and management of information technology and communications.

Network reliability, i.e., the probability of connecting a set of specified nodes under stochastic link failure, is a key indicator of network infrastructure, such as communication and power. Since network reliability evaluation is a computationally heavy task, several methods have been proposed to efficiently perform it. However, modern network infrastructures follow the client-server model, where many clients are served independently, so we have to evaluate the network reliability for every set consisting of the servers and each client. This evaluation process involves repetitive evaluations while changing the set, which imposes a heavy burden on network operators. This paper proposes a method that efficiently performs network reliability evaluation for the client-server model. Since our method is designed to evaluate reliability for multiple clients without explicit repetition, the computational complexity does not increase compared to the case where existing methods evaluate reliability for a single client. Numerical experiments using datasets of various topologies, including real communication networks, reveal great efficiency. Our method is more than 100 times faster than an existing method that requires repeated evaluation, e.g., it takes only 27 seconds to compute the reliability for 670 clients on a large network with 821 links.

Multi-slot-ahead forecasting on network traffic provides an extra degree of freedom to proactively manipulate the network resources when immediate reconfiguration of networks is expensive or infeasible. In return, it challenges the existing data-driven learning-based approaches on accuracy, especially when considering the evolving property of the traffic process. To this end, we establish an adaptive learning framework for multi-slot-ahead network traffic prediction based on Gaussian Process (GP). GP facilitates learning and comprehending the traffic process from a Bayesian perspective, where the main characteristics can be encoded into the kernel function for performance enhancement. The contributions of this paper are two-fold: 1). To track the evolving traffic characteristics, we approximate the optimal kernel adapting to the current traffic. 2). To predict in a large time horizon without significantly hurt the performance, Linear Model of Co-regionalization (LMC) is utilized to better make use of the correlation among subsequent multiple time-slots. Finally, we demonstrate the high tracking capability as well as the superiority of the proposed framework in terms of prediction accuracy through simulation.

With accurate knowledge of future Channel State Information (CSI), it becomes possible to better comprehend the radio propagating environment and manipulate the wireless resources in a proactive manner, so as to provide solid support to smart and high quality wireless transmission. However, in mobile environment, the evolving correlation patterns in CSI series challenge the existing data-driven algorithms to adaptively learn and predict its behavior. In this article, an adaptive learning algorithm is proposed based on Gaussian Process (GP), to discover and utilize the spatial correlation within a channel and across channels, and produce accurate CSI prediction. Specifically, 1). To track the evolving correlation of a channel, we tailor Spectrum Mixture (SM) kernel to not only approximate the optimal kernel adapting to the current CSI, but also capture the combined effect of path loss and User Equipment (UE) motion. 2). The correlation across channels is encoded into the GP-based learning framework through Linear Model of Co-regionalization (LMC). Finally, we verify the performance improvements through simulation.

Artificial Intelligence (AI) has the potential to leverage the immense amount of operational data of clouds, services, and social and communication networks. As a concrete example, AI techniques have been adopted by telcom operators to develop virtual assistants based on advances in natural language processing (NLP) for interaction with customers and machine learning (ML) to enhance the customer experience by improving customer flow. Machine learning has also been applied to finding fraud patterns which enables operators to focus on dealing with the activity as opposed to the previous focus on detecting fraud.

Network and Service analytics can harness the immense stream of operational data from clouds, to services, to social and communication networks. In the era of big data and connected devices of all varieties, analytics and machine learning have found ways to improve reliability, configuration, performance, fault and security management. In particular, we see a growing trend towards using machine learning, artificial intelligence and data analytics to improve operations and management of information technology services, systems and networks.

Clos networks are widely used as an efficient physical structure due to their nonblocking property. However, in strictly nonblocking networks composed of ordinary optical switches (switches with equal numbers of input/output ports), we find that a substantial fraction of ports can remain unused, which decreases the efficiency of Clos networks. This inefficiency comes from the implicit restriction that the two ''sides'' of switches have distinct roles, i.e., ports on one side are connected to endpoints (terminals) while those on the other side are linked to switches. Removing this restriction brings greater freedom in structuring the network and can increase the capacity without losing the strictly nonblocking property. This paper proposes a new physical structure and provides several theorems that address network capacity. Numerical experiments show that our structure increases the capacity by up to about 30%.

Cloud and network analytics can harness the immense stream of operational data from clouds and networks, and can perform analytics processing to improve reliability, configuration, performance, fault and security management. In particular, we see a growing trend towards using statistical analysis, Artificial Intelligence (AI) and machine learning to improve operations and management of IT systems and networks.

Our contemporary society survives on the services provided by several network infrastructures, such as communication, power, and transportation, so their reliability should be accurately evaluated from various aspects. While past studies defined network reliability in terms of connectivity, some of the network connections that have been established may suffer from insufficient resources, i.e., vertices may be connected on a network but flows might fail due to resource contention. As a first step to address resource protection, this paper introduces path disjointness to the field of network reliability analysis, i.e., we evaluate the probability that given terminals are connected via edge- or vertex-disjoint paths. In addition, we also deal with identifying critical links under path disjointness. Since network reliability analysis is a computationally tough problem, we propose efficient algorithms utilizing the data structure of binary decision diagrams. Numerical experiments show that our method scales up to a network with 189 links. We show that network reliability and the criticality of links are greatly dependent on path disjointness; this validates the importance of the proposed method.

Communication networks are an essential infrastructure and must be designed carefully to ensure high reliability. Identifying a fully reliable design is, however, computationally very tough since it requires that a reliability evaluation, which is known to be #P-complete, be repeated an exponential number of times. Existing studies, therefore, attempt to avoid exact optimization to reduce the computational burden by applying heuristics. Due to the importance of communication networks and to better assess the accuracy of heuristic approaches, exact optimization remains a key goal. This paper proposes an exact method for two network design problems: reliability maximization under budget constraints and cost minimization with assurance of reliability. Our method employs a common idea to solve these problems, i.e., a best-first search algorithm that runs on decision diagrams. Our method employs just a single binary decision diagram (BDD) to compute the reliability for any solution and is also used as the basis of a novel heuristic function, called the cost-aware BDD heuristic function, as a search guide. Numerical experiments show that our method scales well; it successfully optimizes a network with 189 links. In addition, our method reveals the poor performance of existing heuristic approaches; a well-known existing heuristic method is shown to yield a solution that offers less than half the optimal reliability.

Recently, network operators are confronting the challenge of exploding traffic and more complex network environments due to the increasing number of access terminals having various requirements for delay and package loss rate. However, traditional routing methods based on the maximum or minimum single metric value aim at improving the network quality of only one aspect, which makes them become incapable to deal with the increasingly complicated network traffic. Considering the improvement of deep learning techniques in recent years, in this paper, we propose a smart packet routing strategy with Tensor-based Deep Belief Architectures (TDBAs) that considers multiple parameters of network traffic. For better modeling the data in TDBAs, we use the tensors to represent the units in every layer as well as the weights and biases. The proposed TDBAs can be trained to predict the whole paths for every edge router. Simulation results demonstrate that our proposal outperforms the conventional Open Shortest Path First (OSPF) protocol in terms of overall packet loss rate and average delay per hop.

Traffic matrix (TM) estimation has been extensively studied for decades. Although conventional estimation techniques assume that traffic volumes are unchanged between origins and destinations, packets are often discarded on a path due to traffic burstiness, silent failures, etc. This paper proposes a novel TM estimation method that works correctly even under packet drops. The method is established on a Boolean fault localization technique; the technique requires fewer counters though it only determines whether each link is healthy. This paper extends the Boolean technique so as to deal with traffic volumes with error bounds just by a small number of counters. Along with submodular optimization for the minimum counter placement, we evaluate our method with real network datasets.

Recent years, Software Defined Routers (SDRs) (programmable routers) have emerged as a viable solution to provide a cost-effective packet processing platform with easy extensibility and programmability. Multi-core platforms significantly promote SDRs' parallel computing capacities, enabling them to adopt artificial intelligent techniques, i.e., deep learning, to manage routing paths. In this paper, we explore new opportunities in packet processing with deep learning to inexpensively shift the computing needs from rule-based route computation to deep learning based route estimation for high-throughput packet processing. Even though deep learning techniques have been extensively exploited in various computing areas, researchers have, to date, not been able to effectively utilize deep learning based route computation for high-speed core networks. We envision a supervised deep learning system to construct the routing tables and show how the proposed method can be integrated with programmable routers using both Central Processing Units (CPUs) and Graphics Processing Units (GPUs). We demonstrate how our uniquely characterized input and output traffic patterns can enhance the route computation of the deep learning based SDRs through both analysis and extensive computer simulations. In particular, the simulation results demonstrate that our proposal outperforms the benchmark method in terms of delay, throughput, and signaling overhead.

Modern networks have complex configurations to provide advanced functions. Network softwarization, a promising new movement in the networking community, could make networks more complexly configured due to the nature of software. Since these complexities make the networks error-prone, network verification is attracting attention as a key technology to detect inconsistencies between a configuration and an operational policy. Existing verifiers are, unfortunately, either inefficient or incomplete (operational policies are not rigorously checked). This paper presents a novel framework of data-plane verification. So as to efficiently manage the large search space defined by packet headers, our framework formalizes the consistency check by applying simple set operations defined in a small quotient space of packet header. This paper also reveals that the two spaces can be connected via the windowing query in computational geometry. Two windowing algorithms are proposed and backed by solid theoretical analyses. Experiments on real network datasets show that our framework with the windowing algorithms is surprisingly fast; when verifying policy compliance in a real network with thousands of switches, our framework reduces the verification time of all-pairs reachability from ten hours to ten minutes.

Wireless communication devices have spread widely in our society. However, they usually depend heavily on communication infrastructure, leaving them vulnerable to disasters or congestion of base stations. In these situations, a method to send out data without the support of infrastructure is required. Data transmission by D2D communication is a reliable method that does not rely on infrastructure. In this paper, we aim to improve the data dissemination using D2D transmission by applying the concept of assigning “modes” to devices according to their own mobility. In our study, we assume a multi-channel environment, where devices will be allocated different amounts of frequency channels according to their modes. We propose a mode selection function that uses velocity information of the devices to assign modes. By using this function, it is possible to allocate more frequency channels to devices of high mobility, so that they can transmit their data to more devices as they move through a wide area. By mathematical analysis, we evaluate the fairness of disseminated data density among devices of various velocities and the obtained results indicate the effectiveness of the proposed method for improving the efficiency of data dissemination.

Network verification is attracting attention as a key technology to detect configuration errors before deploying the network. In verification, a set of packets to be inspected is usually specified by a window -- a multi-dimensional rectangle defined by packet header fields (e.g., address prefixes and port ranges). Network operators have to know the forwarding behaviors of packets inside the window, this can be regarded as the windowing query problem in computation geometry. This paper proposes a novel windowing algorithm for network verification. Unlike existing windowing algorithms, our algorithm runs on a compressed data structure, because the search space has to be represented in a compressed form due to the space complexity.

Modern networks have complex configurations to provide advanced functions, but the complexity also makes them error-prone. Network verification is attracting attention as a key technology to detect inconsistencies between a configuration and a policy before deployment. Existing verifiers, however, either generally verify various properties over the policy at the cost of efficiency, or efficiently perform configuration analysis without paying much attention to the policy. This paper presents a novel framework of data-plane verification, which flexibly checks the inconsistency with great efficiency. For the purpose of generality, our framework formalizes a verification process with three abstract steps: each step is related to 1) packet behaviors defined by a configuration, 2) operator intentions described in a policy, and 3) the inspection of their relation. These steps work efficiently with each other on the simple quotient set of packet headers. This paper also reveals how the second step can be regarded as the windowing query problem in computational geometry. Two novel windowing algorithms are proposed with solid theoretical analyses. Experiments on real network datasets show that our framework with the windowing algorithms is surprisingly fast even when verifying the policy compliance; e.g., in a medium-scale network with thousands of switches, our framework reduces the verification time of all-pairs reachability from ten hours to ten minutes.

Virtual network embedding has been intensively studied for a decade. The time complexity of most conventional methods has been reduced to the cube of the number of links. Since customers are likely to request a dense virtual network that connects every node pair directly (|E| = O(|V|2)) based on a traffic matrix, the time complexity is actually O(|E|3 = |V|6). If we were allowed to reduce this dense network into a sparse one before embedding, the time complexity could be decreased to O(|V|3); the time gap can be a million times for |V| = 100. The network reduction, however, combines several virtual links into a broader link, which makes the embedding cost (solution quality) much worse. This paper analytically and empirically investigates the trade-off between the embedding time and cost for the virtual network reduction. We define two simple reduction algorithms and analyze them with several interesting theorems. The analysis indicates that the embedding cost increases only linearly with exponential decay of embedding time. Thorough numerical evaluation justifies the desirability of the trade-off.

There are many applications which cannot be executed by mobile devices due to their limitations in memory, processing, battery, among others. One solution to this would be offloading heavy tasks to cloud servers in the edge of the network, in a service model called Edge Cloud Computing. The main Quality of Service requirement of this model is a low Service Delay, which can be achieved by lowering Transmission Delay and Processing Delay. Works in literature focus on either one of those two types of delay. This paper, however, argues that an approach which combines transmission and processing technologies to lower Service Delay would be more efficient. This idea is defended by an analysis of the service model and existing stochastic modeling of the Edge Cloud Computing system. We conclude that a dual focus approach would be the only way of truly minimizing the Service Delay, therefore being the desired method to improve Quality of Service. We conclude by laying the foundation for a future model that follows such concept.

Several graph libraries have been developed in the past few decades, and they were basically designed to work with a few graphs. However, there are many problems in which we have to consider all subgraphs satisfying certain constraints on a given graph. Since the number of subgraphs can increase exponentially with the graph size, explicitly representing these sets is infeasible. Hence, libraries concerned with efficiently representing a single graph instance are not suitable for such problems. In this paper, we develop Graphillion, a software library for very large sets of (vertex-)labeled graphs, based on zero-suppressed binary decision diagrams. Graphillion is not based on a traditional representation of graphs. Instead, a graph set is simply regarded as a “set of edge sets” ignoring vertices, which allows us to employ powerful tools of a “family of sets” (a set of sets) and permits large graph sets to be handled efficiently. We also utilize advanced graph enumeration algorithms, which enable the simple family tools to understand the graph structure. Graphillion is implemented as a Python library to encourage easy development of its applications, without introducing significant performance overheads. In experiments, we consider two case studies, a puzzle solver and a power network optimizer, in which several operations and heavy optimization have to be performed over very large sets of constrained graphs (i.e., cycles or forests with complicated conditions). The results show that Graphillion allows us to manage a huge number of graphs with very low development effort.

Building optimal virtual networks across multiple domains is an essential technology for offering flexible network services. However, existing research is founded on an unrealistic assumption: providers will share their private information including resource costs. Providers, as well known, never actually do that so as to remain competitive. Secure multi-party computation, a computational technique based on cryptography, can be used to secure optimization, but it is too time consuming. This paper presents a novel method that can optimize virtual networks built over multiple domains efficiently without revealing any private information. Our method employs secure multi-party computation only for masking sensitive values; it can optimize virtual networks under limited information without applying any time-consuming techniques. It is solidly based on the theory of optimality and is assured of finding reasonably optimal solutions. Experiments show that our method is fast and optimal in practice, even though it conceals private information; it finds near optimal solutions in just a few minutes for large virtual networks with tens of nodes. This is the first work that can be implemented in practice for building optimal virtual networks across multiple domains.

The routing efficiency of structured overlay networks depends on the consistency of pointers between nodes. This consistency can, however, break temporarily when some overlay nodes fail, since it takes time to repair the broken pointers in a distributed manner. Conventional solutions utilize "backpointers" to quickly know the failure among the pointing nodes, which allows them to fix the pointers in a short time. Overlay nodes are, however, required to maintain backpointers for every pointing node which incurs significant consistency check and memory overheads. This paper proposes a novel light-weight protocol; an overlay node gives a "testament" containing its acquaintances (backpointers) only to its successor (i.e., clockwise closest node), and other nodes are freed from maintaining it. Our carefully-designed protocol guarantees that all acquaintances are registered with the testament even in the presence of churn, and the successor notifies the acquaintances for the deceased. Even if the successor passes away and the testament is lost, the successor of the successor can identify the acquaintances at a high success ratio. Simulations show that our protocol greatly reduces the mean time to repair (MTTR) and memory overheads while messaging cost increases.

Popularity ranking of Internet services is an important metric for network operators, because it enables mid- to-long term planning of their network facilities and root cause analysis for unexpected traffic. The service-oriented traffic monitoring is much helpful to infer the popularity, hence it has been gathering much attention from both researchers and practitioners. Lately, service identification of a given flow has become very difficult due to the rapid growth of CDNs and/or encrypted traffic, while some research works employed preceding DNS traffic as a hint. However, because of its cache mechanism, the DNS message count deviates from the actual number of flows, which can greatly degrade the ranking reliability. We propose a theoretical model for inferring the user's number of accesses per domain name by exploiting the characteristics of the DNS message count. To the best of our knowledge, this paper is the first attempt to formulate the effect of user's stub resolvers; previous studies were focused on analyzing the effect of cache servers. We evaluated the precision of our model with a real dataset of traffic of thousands of users. By analyzing the top-50 domain names by the number of users, we can infer the number of flows within a 24% error rate on average in 42 out of 50 FQDNs.

Building optimal virtual networks across multiple domains is an essential technology to offer flexible network services. However, existing research is founded on an unrealistic assumption; providers will share their private information including resource costs. Providers, as is well known, never actually do that to remain competitive. Technically, secure multiparty computation, which is a computational technique based on the cryptography, can be used to secure optimization, but it is too time-consuming. This paper presents a novel method to optimize virtual networks built over multiple domains, with great efficiency but without revealing any private information. Our method employs secure multi-party computation but only for masking sensitive values; it can optimize virtual networks under limited information without any time-consuming technique. It is solidly based on the theory of optimality, and is assured of finding reasonably optimal solutions. Experiments show that our method is fast and optimal in practice even concealing private information; it finds nearly optimal solutions in just a few minutes for large virtual networks with tens of nodes. This is the first work that can be implemented in practice for building optimal virtual networks across multiple domains.

In software-defined networking, applications are allowed to access a global view of the network so as to provide sophisticated functionalities, such as quality-oriented service delivery, automatic fault localization, and network verification. All of these functionalities commonly rely on a well-studied technology, packet classification. Unlike the conventional classification problem to search for the action taken at a single switch, the global network view requires to identify the network-wide behavior of the packet, which is defined as a combination of switch actions. Conventional classification methods, however, fail to well support network-wide behaviors, since the search space is complicatedly partitioned due to the combinations. This paper proposes a novel packet classification method that efficiently supports network-wide packet behaviors. Our method utilizes a compressed data structure named the multi-valued decision diagram, allowing it to manipulate the complex search space with several algorithms. Through detailed analysis, we optimize the classification performance as well as the construction of decision diagrams. Experiments with real network datasets show that our method identifies the packet behavior at 20.1 Mpps on a single CPU core with only 8.4 MB memory, by contrast, conventional methods failed to work even with 16 GB memory. We believe that our method is essential for realizing advanced applications that can fully leverage the potential of software defined networking.

The modern Web architecture basically follows the Representational State Transfer (REST) style. This style offers the architectural properties necessary to implement the Internet-scale Web. However, most authentication and delegation technologies that rely on session state actually deviate from the REST style. It must be noted, however, that the diversity of these technologies is imperative for the success of the Web. In this paper, we make a detailed analysis of current authentication and delegation technologies including OpenID and OAuth as well as HTML forms and cookies, and find that session state has an important role in terms of the diversity of the technologies. We also clarify that the negative impact of session state on the REST style is rather limited. On the basis of our analysis, this paper introduces the REST Using Session (RESTUS) architectural style, which is an extended REST style for sessions; this style places a session constraint on component interactions and so induces some properties required for the diversity.

In this paper, we propose a novel method to map locations of interest by utilizing the "scent" model. Location-aware services are being actively developed as they are a key component in realizing the Ambient Intelligence (AmI) world. Our proposal provides the basics needed to create location-aware services including navigation and notification. Users' interests are loosely shared by mirroring the diffusion of a real scent, and the users can "sniff out" the locations that match their interests. Since our model is based on a simple physical phenomenon, data processing is tractable and its user interface is intuitive. Moreover, user privacy is protected by the random "wind" effect without reducing map accuracy. We conduct thorough outdoor experiments in Kanagawa Prefecture, Japan, to demonstrate our model; twelve subjects were given mobile phones with GPS devices. The results show that the "scent" model provides an intriguing map with which to locate points of interest, and that the developed applications are promising.