With the advancement of the information society, the importance of acquiring and utilizing various types of information from specific users with designated attributes has been increasing. This paper proposes an information acquisition method that leverages the data forwarding capability of delay tolerant network (DTN) to collect such information with mobile devices in a shorter time, without causing data transfer loads and costs on the fixed communication infrastructure. Its effectiveness is shown through computer simulation.

Practical optical buffers have not yet been implemented in optical packet switching (OPS) networks, where packet-by-packet switching is performed using optical signals at optical switches. As a result, packet collisions occur when multiple optical packets simultaneously reach the same output port of optical switches. This leads to significant communication performance degradation, particularly when Transmission Control Protocol (TCP) is used. In this paper, we propose a method that the receiver host adaptively delays the transmission of TCP acknowledgment (ACK) to achieve a predefined target round-trip time (RTT) for the sender host. This target RTT is determined based on real-time pure RTT measurements obtained during data transmission and reception. By dynamically adjusting the RTT of data packets traversing optical paths of various lengths in an OPS network to one of a limited set of fixed values, we can reduce the likelihood of such collisions.

In an Optical Packet Switching (OPS) network environment, the absence of practical optical buffers results in frequent packet collisions at intermediate nodes, as multiple packets may arrive simultaneously at the same output port of a certain intermediate node. This leads to substantial packet loss even under non-congested conditions. Consequently, the conventional Transmission Control Protocol (TCP) does not fully utilize the bandwidth of the bottleneck links. Moreover, highly bursty data transmission can impose the above impact excessively on intermediate nodes in OPS networks, making it undesirable. To address this issue, we have presented a control mechanism consisting of two disciplines: microburst transmission and congestion window updating. However, the performance of microburst transmission control itself, which corresponds to TCP pacing, has not been quantitatively clarified. In this paper, we quantitatively evaluate the performance of microburst transmission control. Specifically, we investigate the impact of different microburst transmission intervals on throughput and clarify the potential fairness of throughput among optical paths with different route lengths.

Nowadays, the number of wireless stations (STAs) connected to a wireless local area network (WLAN) is increasing, and the quality of service (QoS) requirements of transferred data are diverse. Even in such congested situations, it is necessary to assign sufficient frame transmission opportunities to each STA and base station (BS) according to the QoS requirements of transferred data. IEEE 802.11ax introduces simultaneous transmission of multiple frames by Orthogonal Frequency Division Multiple Access (OFDMA) for efficiently accommodating many STAs in a WLAN. For backward compatibility, OFDMA transmission in IEEE 802.11ax is used in cooperation with Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) that is conventionally used in IEEE 802.11. In this paper, we propose a frame transmission method that selectively uses OFDMA transmission and CSMA/CA-based transmission, taking into account the QoS requirements of transferred data and the characteristics of the two types of transmissions.

Edge computing over Named Data Networking (NDN) is expected to be a promising approach for the deployment of applications such as connected cars and virtual or augmented reality. In conventional edge computing frameworks over NDN, computing request packets are not forwarded based on the resource availability status of the edge execution nodes (e.g., CPU and memory utilization). This may result in requests reaching fully loaded nodes that cannot handle these requests. To address this issue, this paper proposes a resource discovery scheme for edge computing over NDN. In this scheme, each execution node distributes its own resource availability status within its vicinity via scoped-flooding. Using this information, requests are guided toward available nodes that can satisfy the requirements of the requests. Moreover, the nodes attach their status to the computed response packets and distribute the status to each node along the response paths. Our extensive simulations show that the proposed scheme helps request packets to effectively discover available edge execution nodes.

In lossy networks, including optical packet switching (OPS) networks and wireless networks, data recovery from frequent packet loss is essential. Erasure-coding, a type of forward error correction (FEC), is one of effective solutions, and furthermore, it could be combined with a technique that feeds back decoding progress information from the receiver to the sender for efficient data transmission. However, conventional feedback methods cannot inform the sender of the decoding progress information at the receiver when acknowledgment (ACK) packets are lost. Therefore, we propose a new data transmission method that is robust to high packet loss by feeding back accumulated information of decoded symbols at the receiver, and furthermore that smartly reduces the calculation overhead of encoding/decoding process. The proposed method can reduce the redundancy by up to 5.84% compared to conventional methods in environments where random packet loss occurs.

Technologies such as volunteer computing, which use computing resources of nodes in the neighborhood of people, have been attracting attention. When using a large number of end nodes with computing power in the neighborhood of people as an “End Cloud,” main problems in selecting an appropriate group of computing nodes that communicate with each other from among the End Cloud include the long physical distances between the computing nodes and the large amount of messages required to precisely grasp the amount of computing resources. In this paper, we propose an algorithm for selecting an appropriate set of computing nodes from the End Cloud taking these problems into account.

Wireless sensor networks and their applications are becoming more diverse, and as an important future challenge, urgent high-priority data must be transferred with minimum delay and loss in a prioritized manner. This paper proposes strictly prioritized multihop transmission of high-priority data, composed of not only single frame but also multiple frames, by reserving the channel and exploiting medium access control coordination among sensor nodes on the route. Its effectiveness is shown through computer simulation.

With the evolution of optical networks, it is expected that each server and user terminal will be equipped with an optical interface and directly connected to all-optical networks. In particular, Optical Packet Switching (OPS) network is one of the candidate future networks, where it can realize ultra-high-speed transmission with a highly statistical multiplexing effect. Thus, in addition to simple one-to-one high-speed connections, Service Function Chaining (SFC), including the selection of servers, optical route, and signal resources, needs to be provided on an optical network operation framework. In this paper, we demonstrate the possibility of selecting appropriate servers and optical signal-level communication resources with an optical network testbed. As a realistic solution to realize SFC on an OPS network operation, we implement a centrally controlled OPS network with software-defined network technology. We verify the dynamic selection of servers and optical communication resources based on aggregated load information of servers and confirm the feasibility of our SFC management.

We present WhichTF, a computational method to identify functionally important transcription factors (TFs) from chromatin accessibility measurements. To rank TFs, WhichTF applies an ontology-guided functional approach to compute novel enrichment by integrating accessibility measurements, high-confidence pre-computed conservation-aware TF binding sites, and putative gene-regulatory models. Comparison with prior sheer abundance-based methods reveals the unique ability of WhichTF to identify context-specific TFs with functional relevance, including NF-κB family members in lymphocytes and GATA factors in cardiac cells. To distinguish the transcriptional regulatory landscape in closely related samples, we apply differential analysis and demonstrate its utility in lymphocyte, mesoderm developmental, and disease cells. We find suggestive, under-characterized TFs, such as RUNX3 in mesoderm development and GLI1 in systemic lupus erythematosus. We also find TFs known for stress response, suggesting routine experimental caveats that warrant careful consideration. WhichTF yields biological insight into known and novel molecular mechanisms of TF-mediated transcriptional regulation in diverse contexts, including human and mouse cell types, cell fate trajectories, and disease-associated cells.

Edge computing over Named Data Networking (NDN) is expected to be a promising approach for the deployment of applications such as connected cars and virtual/augmented reality. In conventional edge computing frameworks, computing request packets are not forwarded based on the resource availability statuses of edge execution nodes (e.g., CPU and memory utilization). This makes it possible for the requests to reach fully loaded nodes that cannot deal with these requests. To address this issue, this paper proposes a resource discovery scheme for edge computing over NDN. In this scheme, each execution node distributes its own resource availability status in its vicinity by scoped-flooding. This information guides the requests toward available nodes that meet the requirements of the requests. Our extensive simulations show that the proposed scheme helps request packets to discover available execution nodes effectively.

In recent years, wireless local area networks (WLANs) have become widespread, and the number of mobile stations (MSs) connected to WLANs and the number of WLANs installed have been increasing. In addition, owing to the diversification of applications using WLANs, transmission opportunities have been proportionally greater in the uplink direction, from MSs to base stations (BSs), than ever. Furthermore, most of this data is transmitted using Transmission Control Protocol (TCP). Therefore, WLANs need to provide opportunities for frame transmission in each direction according to the ratio of TCP traffic intensity. In this study, we propose a TCP-aware frame transmission method that makes effective use of Orthogonal Frequency Division Multiple Access (OFDMA) transmission as introduced in the IEEE 802.11ax standard to improve the TCP throughput of the entire WLAN while realizing the allocation of such frame transmission opportunities. In an environment where TCP flows are transmitted in both uplink and downlink directions, the proposed method transmits each frame by using OFDMA transmission and Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) adaptively and selectively according to the buffering status of TCP data segments and ACKs in each direction.

MOTIVATION: Large-scale and high-dimensional genome sequencing data poses computational challenges. General-purpose optimization tools are usually not optimal in terms of computational and memory performance for genetic data. RESULTS: We develop two efficient solvers for optimization problems arising from large-scale regularized regressions on millions of genetic variants sequenced from hundreds of thousands of individuals. These genetic variants are encoded by the values in the set {0,1,2,NA}. We take advantage of this fact and use two bits to represent each entry in a genetic matrix, which reduces memory requirement by a factor of 32 compared to a double precision floating point representation. Using this representation, we implemented an iteratively reweighted least square algorithm to solve Lasso regressions on genetic matrices, which we name snpnet-2.0. When the dataset contains many rare variants, the predictors can be encoded in a sparse matrix. We utilize the sparsity in the predictor matrix to further reduce memory requirement and computational speed. Our sparse genetic matrix implementation uses both the compact two-bit representation and a simplified version of compressed sparse block format so that matrix-vector multiplications can be effectively parallelized on multiple CPU cores. To demonstrate the effectiveness of this representation, we implement an accelerated proximal gradient method to solve group Lasso on these sparse genetic matrices. This solver is named sparse-snpnet, and will also be included as part of snpnet R package. Our implementation is able to solve Lasso and group Lasso, linear, logistic and Cox regression problems on sparse genetic matrices that contain 1 000 000 variants and almost 100 000 individuals within 10 min and using less than 32GB of memory. AVAILABILITY AND IMPLEMENTATION: https://github.com/rivas-lab/snpnet/tree/compact.

MOTIVATION: The prediction performance of Cox proportional hazard model suffers when there are only few uncensored events in the training data. RESULTS: We propose a Sparse-Group regularized Cox regression method to improve the prediction performance of large-scale and high-dimensional survival data with few observed events. Our approach is applicable when there is one or more other survival responses that 1. has a large number of observed events; 2. share a common set of associated predictors with the rare event response. This scenario is common in the UK Biobank dataset where records for a large number of common and less prevalent diseases of the same set of individuals are available. By analyzing these responses together, we hope to achieve higher prediction performance than when they are analyzed individually. To make this approach practical for large-scale data, we developed an accelerated proximal gradient optimization algorithm as well as a screening procedure inspired by Qian et al. AVAILABILITYANDIMPLEMENTATION: https://github.com/rivas-lab/multisnpnet-Cox.

Currently, many kinds of wireless sensor networks need to communicate aperiodically generated packets like event detection as well as periodically generated ones for environmental, vital monitoring, and so on. Thus, collisions should be avoided among both aperiodic and periodic packets. In this paper, we propose a method to schedule transmission timings of aperiodic and periodic packets at the application layer from the source nodes. This has advantages of no modification to the MAC layer standardized by IEEE 802.11, 802.15.4, etc. and low cost of sensor nodes, compared with existing approaches based on the MAC layer modifications. To the best of our knowledge, this is the first proposal that controls transmission timings of both aperiodically and periodically generated packets at the application layer. Performance evaluation verifies that the proposed method improves packet loss rate compared with existing application layer approaches that control transmission timings of only periodic packets.

Currently, in addition to wireless LAN (WLAN), equipments for longer-distance wireless communication like IEEE 802.11af and 802.11ah are being developed. Thus, in the near future, access points (APs) and wireless stations (STAs) equipped with functions for both WLAN communication and longer-distance communication are expected. Although longer-distance communication enables STAs to connect to their APs even outside the WLAN coverages, the transmission rates are smaller. Multihop WLAN communication via relay STAs provides larger transmission rates. However, frame transmission from relay STAs increases collisions, and additional routing is required. This study proposes collision-protected multihop transmission without preliminary routing in the environment in which both WLAN communication and longer-distance communication are available. Collision protection and routing are processed by a two-way handshake between data frame transmitter and receiver stations with the longer-distance communication. After the handshake, the data frame is transmitted via relay STAs by exclusively using the WLAN channel.

Mobile crowdsensing is gaining wide popularity, and its main aim is to collect significant amounts of data by leveraging mobile terminals such as smartphones that provide a ubiquitous communication environment. On the other hand, named data networking (NDN) has become one of the most popular information-centric networks. In this paper, as a new paradigm, we propose an advanced crowdsensing system using NDN. Unlike previous studies, data, including uncertain information such as people's notions, opinions, and knowledge, are gathered from the outer locations of NDN, thanks to the delay-tolerant feature, which opens up new possibilities for data collection. The request for the data is forwarded in an area-focused manner. In addition, with this approach, we expect to significantly reduce the redundant data responses.

With the rapid increase in IoT (Internet of Things) applications, more sensor devices, generating periodic data flows whose packets are transmitted at regular intervals, are being incorporated into WSNs (Wireless Sensor Networks). However, packet collision caused by the hidden node problem becomes serious particularly in large-scale multi-hop WSNs. Moreover, focusing on periodic data flows, continuous packet collisions among periodic data flows are caused once periodic packet transmission phases are synchronized. In this paper, we tackle the compounded negative effect of the hidden node problem and the continuous collision problem among periodic data flows. To realize this objective, we propose a new MAC layer mechanism. The proposed method predicts risky durations during which collision can be caused by hidden nodes by taking into account periodic characteristics of data packet generation. In the risky duration, each sensor node stops the transmission of packets in order to avoid collisions. To the best of our knowledge, this is the first paper that considers the compounded effect of hidden nodes and continuous collisions among periodic data flows. Other advantages of the proposed method include that any new control packets are not required and it can be implemented in widely diffused IEEE 802.11 and IEEE 802.15.4 devices.

We pursue herein the establishment of an advanced and feasible Content Delivery Network (CDN) system by exploiting a simple Information-Centric Network (ICN) capability on the IP network, introducing a Breadcrumbs-based in-network guidance framework in addition to user-side caches. Through caching property analyses, we first present an interesting and beneficial feature by cooperation as follows: each popular content is automatically (without explicit control) cached on a designated surrogate server without duplication. Moreover, we propose a new cooperation method between CDN and Breadcrumbs, which enhances the abovementioned feature and consequently reduces the workload of origin servers and cache miss ratio by effectively using the cache capacity with a smaller processing overhead.

Recently, wireless LANs are closely located in various places, and radio interference occurs among them. In such interfering environments, especially, it is necessary to maximally increase transmission opportunities by smartly and autonomously discovering multiple frames that can be transferred in parallel. One typical approach is to control CST (Carrier Sense Threshold). However, a frame, intended to transfer in parallel with the precedence frame transfer, may interrupt the precedence transfer, which results in the loss of the precedence frame. In addition, the interference with ACKs replied from receiver stations is not considered. Therefore, in this paper, we propose a parallel frame transfer method based on strict judgement for the parallel transfer of multiple frames. In this judgement, all the frames transferred in parallel are intended to result in success, and the possibility of the parallel transfer with the replied ACKs is also checked.

Mobile terminal users prefer WLAN (Wireless LAN) to cellular network because of no connection fee. WLANs are available in some places where many people gather in urban areas, whereas public access network like WLAN should be deployed at a low cost for rural leisure venues and momentary events. In this paper, we propose a wireless access network with larger coverage than WLAN by the cooperation between WLAN and another wireless network with larger coverage. This achieves the wider deployment of such wireless access networks with a smaller number (cost) of base stations. Connection control that decides which wireless network should be used for each wireless station is also proposed to increase the effective network capacity of the wireless access network. Performance evaluations show that the proposal achieves larger effective network capacity compared with greedy connection that all the wireless stations inside the WLAN coverage use the WLAN and the other ones do the other network with larger coverage, respectively.

SUMMARY: Large biobanks linking phenotype to genotype have led to an explosion of genetic association studies across a wide range of phenotypes. Sharing the knowledge generated by these resources with the scientific community remains a challenge due to patient privacy and the vast amount of data. Here, we present Global Biobank Engine (GBE), a web-based tool that enables exploration of the relationship between genotype and phenotype in biobank cohorts, such as the UK Biobank. GBE supports browsing for results from genome-wide association studies, phenome-wide association studies, gene-based tests and genetic correlation between phenotypes. We envision GBE as a platform that facilitates the dissemination of summary statistics from biobanks to the scientific and clinical communities. AVAILABILITY AND IMPLEMENTATION: GBE currently hosts data from the UK Biobank and can be found freely available at biobankengine.stanford.edu.

Recently, the use of big data has attracted attention as a profitable business strategy, and is expected to keep increasing in the future. In contrast to existing ways of big data analysis based on centralized computing environment such on a few huge data centers, we advocate a distributed and parallel computation environment aiming at fine-grained cloud computing, which includes so-called edge computing. In the advocated environment, it is assumed that many users, which own big data to be analyzed, dynamically participate in the network to request computer resources and leave after finishing their analyses. In such a dynamic and realistic environment, this paper improves the proximity of computer resources to big data by applying virtualized network in which nodes with each big data and the corresponding computer resources are mutually connected by proper optical paths. Optical path arrangement is periodically updated for new users, without affecting other users currently using computer resources. Moreover, a resource assignment algorithm suitable for such dynamic changes is also proposed to achieve fairness in terms of the network distance between big data and computer resources, and effective load balancing among resource suppliers. We evaluate its effectiveness by computer simulation.

Recently, wireless LAN has been diffused. Although unicast transmission is mainly used currently, multicast transmission, such as distribution of multimedia information, is expected to increase in near future. Thus, channel utilization of wireless LAN needs to be improved in general environments where multicast flows coexist with unicast. Although many packet transmission methods that achieve high channel utilization have been proposed, they focus on situations that only either multicast or unicast transmission exists. In this paper, focusing on general environments where multicast and unicast flows are communicated in parallel, we propose new packet transfer scheduling so that high network bandwidth and channel utilization can be realized by transmitting packets at as high rate as possible, by reducing packet loss, and by omitting the overhead to communicate control frames. These improvements are realized by effective cooperation between multicast and unicast flows. Evaluation results demonstrate that network throughputs of both unicast and multicast are increased while decreasing packet transfer delay.

In recent years, network function visualization (NFV), which attempts to provide services flexibly by virtualizing network functions on a general-purpose server, has attracted attention. Here, Service Chaining, which dynamically connects each network function (virtual network function: VNF), is critical. However, as it is difficult to select the optimum VNF, in a decentralized manner, especially the distances between the VNFs tend to increase, leading to larger communication and processing delays. Therefore, in this paper, we propose a distributed search method for ordered VNFs to suppress delays while considering the control server load, with an in-network guidance technology for query messages.

Recently, by the spread of Internet of Things (IoT), wireless sensor networks and their applications are becoming more diverse. Future tasks in IoT include healthcare and factory networks, in which urgent data, in case of sudden condition change of elderly people or disaster due to machine failure, must be transferred with minimum delay and loss. However, existing methods for such prioritized transmission support only relative prioritization among different data traffic. In this paper, we propose a strictly prioritized transmission method of urgent data for transmission with minimum delay and loss. Urgent data are transmitted with strictly prioritized manner in MAC coordination between neighbor nodes along the route.

Lately, ZigBee has been attracting attention as a short-range wireless communication standard. In a ZigBee network, it is necessary to suppress the packet transfer load because ZigBee needs to operate with low power consumption and within narrow communication bandwidth. On the other hand, in typical environments in which ZigBee is utilized, such as a smart home network, it generally coexists with WLAN (Wireless LAN), which has relatively long idle periods. Thus, to reduce the communication load in ZigBee, this paper proposes a hop counts reduction method, where some packets in the ZigBee network are transferred via WLAN when the WLAN is not overloaded. Existing studies that assume an environment in which ZigBee and WLAN coexist are only concerned with interference avoidance between the two networks. On the other hand, this paper focuses on adaptive cooperation, which enables more efficient communication with low power consumption and higher channel utilization.

In recent years, the use of big data as a profitable business strategy has attracted considerable attention, and is expected to drastically increase in the future. We advocate a distributed and parallel computation environment for big data analysis aiming at fine-grained cloud computing, which includes so-called edge computing. This approach differs from the way that data analysis was performed in the past, namely based on a centralized computing model involving a few huge data centers. In contrast, the realization of a distributed computational environment in which computer resources with relatively low capacity are utilized requires optimal matching between the many owners of big data and the suppliers of computer resources. In this paper, we apply network virtualization with optical paths to improve the proximity of the computer resources to the big data, and present a resource allocation method to achieve fairness in terms of the network distance between the big data and computer resources, and effective load balancing among suppliers. The effectiveness of our approach is evaluated by extensive computer simulation.

Coexistence between ZigBee and WiFi technologies, which operate within the same frequency band, is increasing with the widespread use of the IoT (Internet of Things). ZigBee devices suffer significant decrease in the sink arrival rate of packets in the presence of WiFi interference. To overcome this problem, many channel control methods have been proposed. The existing methods switch only ZigBee channels to avoid interference with WiFi. Here, we propose a method for improving ZigBee packet arrival rate by controlling both the WiFi and ZigBee channels. In this paper, we propose a cooperative channel control method designed to improve ZigBee's packet arrival rate. We propose a method that controls not only ZigBee devices and channels but also requests a temporary pause in the use of specific WiFi channels. Finally, we show the effectiveness of the proposed method using computer simulations.

Recently, usage of wireless LANs (WLANs) by mobile devices has spread rapidly. However, these devices consume large amounts of electricity when they communicate through WLAN. Although IEEE 802.11 specifies power saving mode (PSM), it cannot let wireless stations (STAs) sleep in an optimal manner and increases packet transfer delay. In this paper, we propose a power saving method that uses a different framework. Specifically, STAs reduce power consumption by sleeping for the network allocation vector duration that is extended by bidirectional burst transmission within the acceptable delay of each downlink packet. In addition, we put forth the combination of the proposed method with PSM to improve performance in the situations for which the amount of traffic is small.

An increasing number and variety of devices are connected to the Internet. Thus, it is expected that peer-to-peer (P2P) networks, which have no central servers, will solve the server overload issue in the client-server communication model. Among P2P information search schemes, we focus on Skip Graphs, because they exhibit strong churn resilience and have a range search function, which is desirable for advanced networks such as smart meter networks, smart agriculture, smart cities, vehicular ad hoc networks, and online social networks. However, the overlay of a Skip Graph is constructed regardless of the nodes' locations. As a result, the end-to-end delay between communication nodes becomes much longer than the potential minimum. In a conventional method, this problem is solved by using special landmark nodes. However, this approach sacrifices some advantages of Skip Graphs as a P2P system. Thus, we propose a Skip Graph construction method that does not use any special nodes. The simulation results show that the proposed method provides about 35% improvement in transmission distance without additional query hops.

In recent years, caring for the elderly has emerged as a significant problem, and the number of elderly people living alone has also been on the rise. Against this backdrop, the authors propose a network for watching over the elderly in local communities. In this network, a group to watch over an elderly person is designed in a local community, and the watchers communicate with the elderly through messages over Online Social Networking platforms. Since several people in a group watch the elderly in this network, careful monitoring of the elderly and a reduction in the mental and physical burdens on watchers are achieved. The groups are organized per elderly person by the proposed method inspired by some Meta-Heuristic Algorithms so that the multi-objective requirements to each group, to quickly respond emergent unhealthy states of the elderly, and to notice any kinds of signs and symptoms proactively, are satisfied. Moreover, constraints to alleviate the watchers' burden are also applied as our original contribution.

Recently, the demand for wireless networks to treat periodic traffic is increasing, which includes healthcare wireless sensor networks. This tendency causes the problem of continual packet collisions among different periodic flows. In this paper, we propose a new formula for predicting whether two heterogeneous periodical flows from different source nodes have overlapping packet transfer durations. From this formula, we propose a transfer scheduling method that shifts the packet generation phase (timing) to avoid future collisions. This method adopts naive random access control, like DCF, for the MAC layer process. In addition, source nodes do not require significant computational power, because only the sink intensively schedules the timing and informs them to the corresponding source. Therefore, compared to existing methods in which each source node completely schedules the timing of creating packets based on time division multiple access (TDMA), our method requires less functional expansion, complexity, and computational power. Finally, we demonstrate the effectiveness of our method through computer simulations.

Optical circuit and packet integrated network is a promising network architecture that can deal with various sizes/types of data efficiently. In this integrated network, paths (routes) for circuit and packet switching are assumed to be calculated and selected independently. However, this approach is insufficient in terms of efficiency and rapidity of total path calculation. Calculating appropriate paths with shorter computation time is necessary. In this paper, we propose and evaluate a high-speed multipath computation method that quickly calculates end-to-end paths for optical circuit switching by making use of the calculation results of next hop information for packet switching.

Lately, network usage has been dominated by content distribution, and hence, content oriented network (CON) has attracted much attention. In CON, however, we encounter a scalability problem regarding content acquisition, particularly in large-scale networks composed of multiple domains, because the total amount of content is much larger than the number of network nodes. CON requires a more scalable routing scheme than the traditional IP routing. We have proposed a breadcrumbs-based architecture to establish highly scalable content routing for the content-oriented paradigm. In this paper, we consider another novel approach for controlling guidance (routing) information. We propose a breadcrumbs-scoping (BC-scoping) framework that manages the distribution scope of guidance information according to some criteria. First, we propose BC-scoping on domain, which limits the distribution scope of guidance information within domains according to a hierarchical network structure. It promotes intradomain content acquisition and prevents the occurrence of cross-domain BC trails, because intradomain content acquisition is preferred, both technically and economically. Second, we consider the use of multiple criteria simultaneously with this framework when the criteria are independent of each other. Specifically, we propose BC-scoping on domain and popularity as a practical example of BC-scoping on multiple-criteria. For both of the proposed methods, we show the effectiveness through extensive computer simulation.

In recent years, wireless networks have been utilized in many situations. However, further increase in network capacity is essential in order to transfer large volume of traffic demanded by many and various clients and applications. By introducing multi-channel environment, network capacity is increased. However, in cases that packet generation rates from wireless stations to a base station are different among wireless stations, all the communication bandwidth over multiple channels is not properly allocated to wireless stations with larger generation rates in particular, even though there is enough available network bandwidth in total. This issue becomes significantly serious in actual wireless access networks where the distance between the base station and each wireless station is different, that means different maximum transmission rate of each station, under the fading condition. Therefore, in this paper, we adaptively allocate proper communication channel, and as a result, adaptive bandwidth to all wireless stations so that they transmit all generated packets by equalizing sojourn time in their sending buffers if there is enough total network bandwidth over all channels. Finally, we show the effectiveness of the proposed method by computer simulation.

Recently, wireless LANs are utilized in various situations. However, inefficient channel utilization is still a serious problem. One of effective solutions is to prevent packet loss and to transmit packets at a higher rate. While the backoff algorithm in IEEE 802.11 avoids only the packet loss caused by collision, other existing works tackle the packet loss caused by channel fading by transmitting packets at a lower rate. The authors have proposed a transmission scheduling algorithm which avoids decreasing transmission rate and improves channel utilization. However, these previous methods do not consider influence to TCP performance. In this paper, we propose a novel packet transmission scheduling which enhances TCP performance by effectively utilizing higher channel transmission rate and by keeping TCP congestion window size large. The proposed scheduling at MAC level avoids retransmission time out (RTO) of TCP ACK and duplicate ACK, without modifying TCP itself and adopting complicated cross-layer approaches. Evaluation results demonstrate that total throughput which aggregates all transferred TCP flows and the minimum throughput among the TCP flows are enhanced by explicitly decreasing the frequencies of duplicate ACK and RTO.

These days, in addition to host-to-host communication, Information-Centric Network (ICN) has emerged to reflect current content-centric network usage that many users are now interested not in where the content is but in acquired contents themselves. However, the authors believe that current IP network still remains, at least from deployment perspective, as one of near future network architectures. This is because ICN has various scalability and feasibility challenges, and host-to-host communication is also diffused like remote login, VoIP, and so on. Therefore, the authors' research aims to establish the feature of information-centric network on conventional IP network to achieve feasible architecture. In this paper, we propose to operate Breadcrumbs (BC) and Content Delivery Network (CDN) frameworks coordinately on IP network to improve the performance on content retrieval and acquisition. Both BC and CDN are important as a content-centric technique. Finally, we compare the proposed method with CDN that we carefully modeled through simulation. Simulation results show that our approach can reduce server load.

Currently, most of cloud computing services based on IaaS are provided by one specific provider for many users. However, this has some problems in terms of initial cost or ISP transit cost. In addition, individual small use is difficult because the services tend to be provided in a large unit to use. To solve these problems, we advocate a “participating cloud computing platform” that personal or small company level resource suppliers participate in, configure and provide cloud computing. Then, to efficiently search nearby resources among many and various ones in the platform, we propose Resource Breadcrumbs (RBC) as a key technology to provide in-network guidance capability for users' requests. This enables to supply users a smaller unit of resources: computing resource, memory resource, content, application etc. in comparison to traditional IaaS service. In the proposal, supplier-nodes distribute, around them, guidance information directed to themselves with dedicated control message. The distributed information enables user to autonomously locate nearby resources. In addition, distributed management reduces computational load on central database and enables to build cloud platform with lower cost.

What is an ultimately evolved form of cloud computing environment? The solution should have two important features; “Fine-Granular” and “Participating”. To realize such an attractive and feasible solution with the two features, we advocate a “participating fine-granular cloud computing platform” that a large number of personal or small company level resource suppliers participates in, configures and provides cloud computing. In particular, to efficiently search nearby resources among many and various ones in the platform, we propose Resource Breadcrumbs(RBC) as a key technology to provide in-network guidance capability autonomously for users' requests. This enables to supply users a smaller unit of resources: computing resource, memory resource, content, application etc. in comparison to traditional IaaS service. In the proposal, supplier-nodes distribute, around them, guidance information directed to themselves with dedicated control message, and additionally, the information may be logged along the trail of ACK message from central database server or supplier to user. The distributed information enables user to autonomously locate nearby resources. In addition, distributed management reduces computational load on central database and enables to build participating fine-granular cloud platform with lower cost.

In a wireless sensor network (WSN), one of the most important challenges is extending network lifetime because it is difficult to replace sensor nodes and recharge their batteries. In a WSN, nodes around a sink tend to die earlier than other nodes because these nodes have to transmit and relay more packets. To overcome this problem, many load-balancing methods were proposed. By the way, in recent years, two or more overlapping WSNs have been constructed in the same geographic area. In this situation, the overlapping WSNs are expected to extend their lifetime by cooperative packet forwarding. In this paper, we introduce shared nodes with multichannel communication capability as interchange points among respective WSN planes. With these shared nodes, we propose a novel cooperative routing method for WSNs to balance each other's load. Finally, we show the effectiveness of the proposed method by simulation experiments.

Recently, the demand of wireless sensor network treating periodical traffic is increasing. This tendency causes the problem of packet collisions among different periodical traffic. Specifically, packet collisions occur continuously in the case that packet generation and forwarding timing of a sensor node overlaps with other nodes. IEEE 802.11 DCF can detect packet collision and retransmit the lost packets to some extent. However, this approach does not solve the collision caused by the overlapping timing, because packets belonging to different traffic collide as long as the transmission timing is similar. In this paper, we propose a transmission scheduling method that shifts packet generation phase of periodical traffic to avoid the overlap of packet generation and forwarding. To realize appropriate phase shift, packet collision event is recorded at each relay node and the information is gathered at the sink. Then, the sink detects the traffic which experiences the most collisions and instructs the corresponding source node to shift the timing of creating packets.

These days, underwater exploration demand is increasing in the field of industry, science, military, and so on. Hence, Underwater Sensor Network to collect data by scattering sensors in water has been attracted attention. However, there are various problems specific to underwater environments including the communication medium. In this paper, we focus on some problems on broken or degrading link between two neighboring nodes moving away from each other along with water current. To recover such links, we propose to introduce and control mobile nodes to relay communication between such two nodes. Existing studies dealing with the controlled mobility focused on the movement from the sea bed towards the water surface in order to avoid broken links caused by a shadow zone. On the other hand, the proposed method recovers broken and degrading links caused by movement of sensors by the water flow. Different from exiting works, limited number of smart mobile nodes move in any directions. In addition, because the cost of node movement in UWSN is higher than that of terrestrial network, this method also aims at reducing the movement distance of mobile nodes. Furthermore, the proposed method can measure and deal with the link disconnection proactively because the movement of mobile nodes is controlled while grasping the network condition. Finally, we show the effectiveness of the proposed method by computer simulation.

Recently, multicast transmission in wireless LAN is increasing. On the other hand, due to wide diffusion of smartphones and tablet computers as well as laptop ones, electric power saving of STAs (wireless stations) becomes more important. However, when a few STAs which can use only a low transmission rate exist in multicast group, the AP (Access Point) needs to transmit packets to all STAs at the low rate. As a result, the power consumption of all the STAs extremely increases by extending operating period of transceiver. In this paper, to reduce total power consumption of STAs in multicast group, we propose multi-rate parallel transmission where packets are transmitted multiple times at high and low rates. This reduces power consumption of STAs on which a high transmission rate is available because packet reception at the high rate decreases operating period of transceiver and sleep period can be inserted. In other words, power consumption required for packet communication processing itself is reduced. This differs from existing power saving method in which STAs sleep only when there are no packets to communicate. Finally, we show the effectiveness of the proposed method by computer simulation.

In this paper, we focus on an information gathering service like human sensing systems and supervisory system for nursing care, where a server collects information from many clients at the same time, under the current IP network following so-called Best-Effort Discipline. In such applications, especially, communication quality should be enhanced in order to attain application-specific target throughput at maximum. To tackle this technical issue, we propose a method which adaptively tunes TCP socket buffer size and the number of TCP connections according to the policy which preferentially ensures target QoS for high priority clients. This is an unprecedented research about many-to-one transport protocol which has functions to coordinate between multiple sessions. In addition, the proposed method is suitable for the system which transports information to a server considering various network access and priority levels. Through extensive simulation experiments, we demonstrate that proposed methods can ensure the required communication quality within physical limitations according to priorities, regardless of TCP versions.

Inefficient channel utilization is still a serious problem in wireless LANs. One of effective solutions is to decrease the frequency of packet loss. While the backoff algorithm in IEEE 802.11 avoids only the packet loss caused by collision, other previous works that prevent the packet loss caused by channel fading improve channel utilization with transmission at a lower rate. However, utilizing lower transmission rate degrades transmission efficiency because the channel is occupied for a longer time. In this paper, we avoid decreasing transmission rate with a novel transmission scheduling. Specifically, transmission to stations under fading channel condition is interrupted until the condition improves, and other transmission is processed on ahead. Moreover, we propose an extended RTS/CTS handshake which quickly detects the improvement of channel condition and minimizes the wasted time even if fading loss occurs. Evaluation results demonstrate that network throughput is improved without degrading the throughput fairness between stations.

Recently, multicast delivery of multimedia data via wireless LAN is increasing. Though IEEE 802.11 lacks the mechanism to recover packet loss in multicast, network coding (NC) has demonstrated its potential to retransmit lost packets efficiently. However, in existing NC methods, transmission delay becomes worse as the number of receiver stations increases because lost packets are continuously retransmitted until all stations receive them. In transmission of multimedia data, especially, transmission delay should be minimized. Therefore, we decrease transmission delay and improve its scalability for the number of receiver stations, at the expense of small packet loss acceptable for multimedia data.

Lately, access loads on servers and network traffic are increasing due to larger content-size and higher request-frequency in content distribution networks. Breadcrumbs (an architecture with guidance information for locating a content cache) is designed to reduce server-loads and to form content oriented network autonomously in cooperation with cached contents. We also proposed Breadcrumbs+ which solves BC's problem that queries visit the same routers repeatedly with forming a loop of routing. However, Breadcrumbs itself still has a technical issue that some queries potentially traverse longer hops compared to conventional approach. In addition, Breadcrumbs takes only a passive approach; a query is routed by a trail of BCs only when the query happens to encounter a corresponding BC. This indicates that Breadcrumbs has a space to enhance its capability. We expect that active and adaptive control of BC improves user experience. In this paper, we propose Active Breadcrumbs which provides higher cache utilization through active spreading guidance information around cache-nodes. Finally, we show the effectiveness of Active Breadcrumbs in terms of traversing content/query hops and server utilization through the extensive computer simulation.

Recently, video streaming services using P2P(Peer-to-Peer) have attracted attention to solve the problem of load concentration on servers and reduce a lot of latency. Many P2P streaming systems, like Coolstreaming, however, have a complicated approach to control playback timings severely. And this leads to less churn resiliency and less adaptability to fluctuation of network traffic. Therefore, we focus on a simple and robust approach to realize “pseudo” streaming with high quality, which is based on BitTorrent. In the existing methods with the simple approach, peers download pieces just closer to playback timings to decrease the playback discontinuity. In contrast, in this paper, we propose a new P2P video streaming system considering the cooperation of three metrics; video structure, playback timings, and pieces dispersion on network. In this system, users vary three piece selections to suit the delivery status. Specifically, users preferentially download pieces which affect the video quality, are closer to playback timings, and improve the delivery efficiency. Moreover, we show the effectiveness of the proposed method by computer simulation.

Recently, optical network technologies are actively researched and developed. However, adaptive cooperation of optical network with network applications is not considered deeply. On the other hand, as a content distribution system, P2P(Peer-to-Peer) has attracted attention. Currently, content delivery with P2P is based on IP network, and its application to optical network has not been considered. Therefore, we consider adaptive cooperation between P2P and optical network, especially OCS(Optical Circuit Switching) network for high feasibility to distribute contents more efficiently.

Recently, network coding (NC) has been popularly applied to wireless networks in order to improve scarce wireless capacity. In wireless LANs, NC can be applied to packet retransmission in which multiple packets can be equivalently transmitted by a single retransmission trial at base station (BS). In this paper, retransmission based on NC cooperates with IEEE 802.11e EDCA. In EDCA, when network load is high, QoS is significantly degraded even in high priority class. To solve this, existing methods improve backoff control, and decrease packet loss caused by collision. However, backoff control cannot prevent packet loss caused by multipath channel fading. In the proposed cooperation between NC and EDCA, QoS of high priority class is improved from the aspect of efficient loss recovery. Unlike NC method with no QoS control, we encounter transmission scheduling problem among NC packet, single lost packet, and new packet. Moreover, in the constitution of packets encoded into NC packet, priority and traffic load of each priority class should be considered. Therefore, we propose how to determine the packet to be transmitted at next transmission opportunity in BS and how to constitute NC packets encoded by adequate set of lost packets. Finally, we show the effectiveness of the proposed method by extensive computer simulation.

Optical network technologies based on WDM (Wavelength Division Multiplexing) are actively researched and developed. However, adaptive cooperation of their property with network applications is not considered deeply. On the other hand, as one of attractive content distribution systems, many researchers focus on P2P(Peer-to-Peer) technologies. Currently, content delivery system with P2P is based on IP network, and its application to optical network has not been directly considered yet, though it is attractive to transfer huge content between data centers or between Origin servers, etc, in addition to forthcoming various large contents. Therefore we consider adaptive cooperation between P2P and optical network to distribute huge contents more efficiently but to supress the Origin server load, and propose the novel and future content distribution platform called BitTorrent over Optical Network (BT-ON). In BT-ON, the OCS (Optical Circuit Switching) is adopted as optical switching scheme due to its feasibility and suitability to large content transfer, while BitTorrent is done as P2P system due to its efficiency and generality. To overcome the technical issues caused by OCS network architecture where the physical resources are actually reserved by specific signaling processes, the BT-ON introduces some novel methods; application-triggered path set-up procedure, wavelength path allocation method, advanced improvement suitable for OCS network like wavelength-path based parallel request and enlargement of data transmission unit, and the peer grouping according to wavelength path reservation status.

Multi-Channel MAC protocols increase network throughput because multiple data transfers can take place simultaneously. However, existing Multi-Channel MAC protocols do not take full advantage of the multi-channel environment, because they lack a mechanism allowing wireless stations to acquire vacant channel and time resources. In this paper, to achieve further throughput increase, we newly propose the Multi-Channel MAC protocols that effectively utilize idle channels and potentially available time of stations by employing burst and interrupted frame transfers, under the condition that each station can use only two transceivers. We assume a transceiver can behave as either a transmitter or a receiver at a time. We show the effectiveness of the proposed methods by computer simulation. Furthermore, through the evaluation as to more than two transceivers' case, we confirm two transceivers' case is best solution in terms of both attained throughput and hardware complexity.

Recently, network coding (NC) has been popularly applied to wireless networks in order to improve channel utilization. In wireless LANs, when NC is applied to packet retransmission, a base station can simultaneously retransmit multiple packets destined to different wireless stations for a single retransmission trial. On the other hand, NC creates additional packet delay at both base station and wireless stations, and hence, packet transfer delay may increase seriously. However, existing NC-based retransmission methods do not consider this additional delay explicitly. In addition, when the number of flows is small, NC exhibits less benefits because the chances of NC-based retransmission are highly reduced. Therefore, in this paper, we propose a novel NC-based retransmission method in order to improve packet transfer delay and jitter of received packets. Moreover, to achieve further improvement of delay, jitter and retransmission efficiency, we propose a retransmission method in which NC-based retransmission cooperates with the typical ARQ method. We overcome the disadvantage of NC-based retransmission by combining with ARQ cooperatively. Finally, we show the effectiveness of the proposed methods by extensive computer simulation.

Recently, wireless LAN is achieving remarkable growth, and maturity. On the other hand, by the advance of the Internet, the demand for the multimedia communication services which include video and voice will be expected to grow. Therefore, in the future, the mechanism of QoS guarantee must be realized even in wireless LAN environment. So far, IEEE 802.11e EDCF has been proposed, which is a contention based channel access method to achieve the QoS guarantee in wireless LAN. However, this cannot realize the desired throughput ratio or deterministic target throughput in principle. In this paper, we expand the EDCF to solve such QoS issues and enable more flexible QoS control. Moreover, we show the effectiveness of our proposal by computer simulation.