Country：Japan

Country：Japan

Country：Japan

Country：Japan

Country：Japan

Country：Japan

<p>[背景] 扁平足や外反母趾などの足部変形は歩行機能を阻害する．足部のバイオメカニクスに着目すると，中足部を構成する舟状骨高や足底圧分布から足部特性が計測されている．中足部は上下運動のみではなく3次元的に動作する．そこで本研究では，スマートフォンを用いて3次元的に足部骨格を計測し，中足部に加わる回転モーメント評価システムを開発することを目的とした．[方法] 対象者は中高年の989名(39-89歳)とした．提案する中足部回転モーメントは本システムで抽出する舟状骨の足部中心線からの距離(モーメントアーム)と体重から導出した．そして中足部回転モーメントと足部骨格の変形の相関を求めた．中足部回転モーメントは体重の影響を受けるため，BMI別に解析を行った．[結果] 静止立位時の中足部回転モーメントは7～15 Nmと導出された．中足部回転モーメントの要素である舟状骨のモーメントアームはBMIが大きくことで増大した．さらに中足部回転モーメントは足部の骨格のずれと BMIが大きくなることで相関が高まった．[考察] 中足部回転モーメントと足部骨格の変形には関連があることが明らかになった．閉鎖運動連鎖(CKC)の観点から中足部の衝撃吸収機能は変形性膝関節症などの関節疾患と関係すると予想され，中足部回転モーメントとの関連が考えられる．以上より，足部骨格の3次元的な特徴量計測の意義が示唆された．</p>

DOI： 10.11239/jsmbe.annual60.109_2

<p>Deformities of the foot result in loss of walking function. The skeletal structure of the foot develops during childhood, and quantitative assessment of the foot is warranted. In healthy children, foot length and navicular height have been studied individually; however, the foot has not been sufficiently studied in its entirety, and the growth curve of the foot has not been clarified. In particular, the heel is an important site influencing foot biomechanics. In this study, we developed a three-dimensional (3D) foot scanner and digital footprint device to examine the foot characteristics of both children and adults. This study aimed to compare the foot characteristics of developing children with those of adults. Overall, 154 children (aged 8–10 years) and 403 adults (aged 40–89 years) were included in the study. The 3D foot scanner and digital footprint device were used for measurements. Foot length, forefoot width, foot height, navicular height, and heel width were evaluated. In particular, the ratio of each indicator to foot length was used to evaluate the characteristics of the child's foot. The measurement indices were significantly larger in adults than in children, indicating that the indices increase as children grow, irrespective of sex. The results of the measurements and the ratios to foot length showed large individual differences between adults and children in some of the measurement indices. However, there was no difference in the heel width between children and adults. In contrast, forefoot flattening was greater in adults. Increases in foot length and heel width were found to vary greatly among individuals and those with different characteristics. While foot length and navicular height increased in adults, heel characteristics did not differ significantly between children aged 8–10 years and adults. Children appear to show foot geometries and proportions different from adults, which would have important implications for the ergonomics of children's footwear.</p>

DOI： 10.14326/abe.11.172

<p>[背景] 外反母趾は，子どもから高齢者まで幅広く発生する一般的な足部変形である．しかし外反母趾の発生メカニズムは十分に解明されていない．外反母趾には靴などのフットウェアや体重などの外的要因が挙げられるが，足部骨格特性の影響を明らかにすることが重要である．本研究では，スマートフォンを用いて3次元に足部の表面構造を計測し，外反母趾に関連する足部骨格の特徴を明らかにすることを目的とした．[方法] 対象者は中高年女性268名とした．着目点は本システムで新たに設定した拇指角，舟状骨位置・高さ，足部骨格のずれ，横アーチの高さ・幅等とした．さらに舟状骨の位置と体重から中足部回転モーメントを導出することを試みた．[結果] 拇指角と相関があった特徴量は骨格のずれ，舟状骨の位置，横アーチ幅，中足部回転モーメントであった．骨格のずれと相関が高かったのは舟状骨の位置，足部高，横アーチ幅であった．拇指角は中足部回転モーメントが大きくなることで有意に増加した．[考察] 外反母趾に影響を与えるメカニズムとして，1.舟状骨の位置・高さに起因し，回転モーメントが増大するものと，2.前足部の幅に起因する2つが推察された．1により舟状骨と内側・中間楔状骨の可動性が大きくなり，第１中足骨の捻じれが発生する可能性が推察された．以上より，足部骨格の特徴量を定量的に計測することで外反母趾のリスク推定を行える可能性が示唆された．</p>

DOI： 10.11239/jsmbe.annual60.109_1

<p>We consider the determination of three-dimensional maps of received signal strength (3D-RSS maps) for disaster-recovery networks enabled by unmanned aerial vehicles (UAVs). In this paper, we extend the existing tensor completion based estimator to propose an efficient new 3D-RSS map estimator. To reduce the sensing route length for the UAV, the proposed method utilizes two approaches for estimating the RSS maps (the tensor completion-based and path-loss-based approaches), depending upon the number of high buildings. We show by simulation experiments that the proposed method can achieve a data-collection time comparable to those of existing methods with a shorter sensing route.</p>

DOI： 10.1587/comex.2021xbl0139

CiNii Article

DOI： 10.1016/j.medengphy.2021.08.001

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (scientific journal)  

DOI： 10.14923/transfunj.2020JAL2011

<p>The function of the medial longitudinal arch of the foot is essential for maintaining a stable gait.To assess this, many studies measure the outer dimensions of the foot, utilizing footprints. However, it has been highlighted that inconsistent results have been obtained during measurements.Therefore, this study compared the evaluation index of the medial longitudinal arch using the results obtained from a developed 3D foot-scanning system and footprint. The target group consisted of 403 people (mean age, 63.7±12.0 years).Resultantly, the 3D foot-scanning system index moderately contributed to CSI, which are the indices of the medial longitudinal arch associated with the digital footprint. It became clear that the use of quantitative indices for 3D foot measurements is important for evaluating the medial longitudinal arch.</p>

DOI： 10.11239/jsmbe.annual59.369

CiNii Article

Publishing type：Research paper (international conference proceedings)   Kind of work：Joint Work  

DOI： 10.1109/MSN50589.2020.00061

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (international conference proceedings)   Kind of work：Joint Work  

DOI： 10.1109/NOMS47738.2020.9110448

Publishing type：Research paper (international conference proceedings)   Kind of work：Joint Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

DOI： 10.1109/ISPACS48206.2019.8986318

DOI： 10.1109/ISPACS48206.2019.8986292

Publishing type：Research paper (scientific journal)   Kind of work：Single Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (international conference proceedings)   Kind of work：Joint Work  

Publishing type：Research paper (scientific journal)  

DOI： 10.1109/JCN.2018.000069

Publishing type：Research paper (international conference proceedings)   Kind of work：Joint Work  

Publishing type：Research paper (scientific journal)   Kind of work：Joint Work  

<p>Information-centric networking (ICN) has gained attention from network research communities due to its capability of efficient content dissemination. In-network caching function in ICN plays an important role to achieve the design motivation. However, many researchers on in-network caching due to its ability to efficiently disseminate content. The in-network caching function in ICN plays an important role in realizing the design goals. However, many in-network caching researchers have focused on where to cache rather than how to cache: the former is known as content deployment in the network and the latter is known as cache replacement in an ICN router. Although the cache replacement has been intensively researched in the context of web-caching and content delivery network previously, networks, the conventional approaches cannot be directly applied to ICN due to the fine granularity of chunks in ICN, which eventually changes the access patterns. In this paper, we argue that ICN requires a novel cache replacement algorithm to fulfill the requirements in the design of a high performance ICN router. Then, we propose a novel cache replacement algorithm to satisfy the requirements named Compact CLOCK with Adaptive Replacement (Compact CAR), which can reduce the consumption of cache memory to one-tenth compared to conventional approaches. In this paper, we argue that ICN requires a novel cache replacement algorithm to fulfill the requirements set for high performance ICN routers. Our solution, Compact CLOCK with Adaptive Replacement (Compact CAR), is a novel cache replacement algorithm that satisfies the requirements. The evaluation result shows that the consumption of cache memory required to achieve a desired performance can be reduced by 90% compared to conventional approaches such as FIFO and CLOCK.</p>

DOI： 10.1587/transcom.2017EBP3299

CiNii Article

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (international conference proceedings)   Kind of work：Joint Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (scientific journal)  

DOI： 10.1109/CCNC.2018.8319303

Publishing type：Research paper (scientific journal)  

<p>Information-centric networking (ICN) has received increasing attention from all over the world. The novel aspects of ICN (e.g., the combination of caching, multicasting, and aggregating requests) is based on names that act as addresses for content. The communication with name has the potential to cope with the growing and complicating Internet technology, for example, Internet of Things, cloud computing, and a smart society. To realize ICN, router hardware must implement an innovative cache replacement algorithm that offers performance far superior to a simple policy-based algorithm while still operating with feasible computational and memory overhead. However, most previous studies on cache replacement policies in ICN have proposed policies that are too blunt to achieve significant performance improvement, such as first-in first-out (popularly, FIFO) and random policies, or impractical policies in a resource-restricted environment, such as least recently used (LRU). Thus, we propose CLOCK-Pro Using Switching Hash-tables (CUSH) as the suitable policy for network caching. CUSH can identify and keep popular content worth caching in a network environment. CUSH also employs CLOCK and hash-tables, which are low-overhead data structure, to satisfy the cost requirement. We numerically evaluate our proposed approach, showing that our proposal can achieve cache hits against the traffic traces that simple conventional algorithms hardly cause any hits.</p>

DOI： 10.1587/transcom.2017ITP0006

CiNii Article

Publishing type：Research paper (international conference proceedings)   Kind of work：Joint Work  

Publishing type：Research paper (international conference proceedings)   Kind of work：Joint Work  

Publishing type：Research paper (scientific journal)   Kind of work：Joint Work  

Publishing type：Research paper (scientific journal)  

<p>As a result of large-scale and complicated of ICT (Information and Communication Technology), the technologies of operations and management of ICT infrastructure and systems are changing to adopt the variation and diversity of usage and communication characteristics. In retrospect, operations and management technologies cover widely from the support of telecommunications operations and remote monitoring for maintaining network equipments, to an integrated network management framework for configuration, monitoring, testing and controls. Recently, the spread of network virtualization technologies enable agility, integrity, and resilience to ICT services. Operations and management technologies will be more important in future, for the support of integrated management of ICT infrastructure including computing resource, and automation of service operations. In this paper, we review research activities of the Technical Committee on Information and Communications Management (ICM) with discussions of individual research category. We then describe the state-of-the-art topics and future directions in the area of ICM.</p>

DOI： 10.1587/transcom.2016PFI0017

CiNii Article

Publishing type：Research paper (international conference proceedings)   Kind of work：Joint Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

Publishing type：Research paper (conference, symposium, etc.)   Kind of work：Joint Work  

A signal detection and its classification are important to cognize radio environments, which include a radio wave monitoring or cognitive radio applications. In this paper, a new QAM classification method is presented. The method is based on two amplitude moments. The classification error probability is derived in the presence of timing offset, and the numerical results demonstrate the superior classification performance of the method.

DOI： 10.14923/transcomj.2016jbl4002

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

Modulation classification is a key technology in current wireless communication systems. The technology is used to find the modulation employed or estimate channel state information (CSI). In this paper, a modulation classification method is proposed for phase shift keying (PSK) and quadrature amplitude modulation (QAM) under an unknown signal-to-noise ratio (SNR) condition. The method works without carrier synchronization and is based on likelihood estimation using the probability density function (pdf) of the received signal envelope. The superior classification error rate of the proposed method is demonstrated by computer simulations.

DOI： 10.2299/jsp.20.183

CiNii Article

Publishing type：Research paper (scientific journal)   Kind of work：Single Work  

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

A method of application identification is important for network management because it is used for traffic controls and anomaly traffic detection. To apply traffic controls based on applications immediately, real-time application idenfitication technique is essential. We previously proposed a method of real-time application identification, and it achieved high accuracy. However, our proposed method needs the same number of packets to identify each application because it uses only one classifier. The problem in our proposed method is to need more packets than necessary in order to identify applications even if it can identify applications earlier. In this paper, we propose a method of multi-stage application identification, which use multi-classifier. First, we reveal that accuracy is degraded when increase number of applications. Second, we examine traffic features used by group classification. Finally, we describe how to classify applications into groups. We demonstrate that our proposed method can identify applications earlier than using one classifier.

CiNii Article

Publishing type：Research paper (scientific journal)  

OpenFlow, one of the famous implementation of Software Defined Network (SDN), has a goal of eliminating vendor lock in. However, since OpenFlow only provides the mechanism to decide how packet will be transferred as the protocol, it could not provide whole control via OpenFlow protocol. Therefore, the OpenFlow switch available for its network system strongly depends on how the switch supports the configuration interface and functions for its switch. In this paper we propose the SDN switch architecture, which conclude the OpenFlow controller inside the switch, to avoid vendor lock in for this reason. Furthermore, we also show that our proposal architecture can overcome the issue of using OpenFlow including the large initial costs and lack of number of flow table.

CiNii Article

Publishing type：Research paper (scientific journal)  

A signal detection and its classification are important to cognize radio environments, which include a radio wave monitering or cognitive radio applications. In this paper, a new QAM classification method is presented. The method is based on two amplitude moments. The classification error probability is derived in the presence of timing offset, and the humerical results demonstrate the superior classification performance of the method.

CiNii Article

Publishing type：Research paper (scientific journal)  

The naming scheme in a content-centric networking (CCN) is important because it directly affects the user-friendliness of the client. In particular, it is usual to introduce a hierarchical naming scheme to archive global uniqueness of names without conflicts, is overly demanding for the client to collect complete names of content because hierarchical complete names require knowing the entirety of the structure, which makes end users difficult to specify the content without knowledge of complete naming structure. In this paper we propose name completion, which is a process for completing partially specified names to fully specified names, and we describe its design and implementation. We next demonstrate an implementation of name completion in an experimental network to show the effectiveness of the name completion procedure.

CiNii Article

Publishing type：Research paper (scientific journal)  

According to radio channel conditions, the adaptive system changes modulation, coding, or power dynamically. The exact estimation of channel SNR is important in the adaptive system. In this paper, the new methods of maximum likelihood estimation are presented for SNR by using the probability density function (pdf) of envelope of received signals. The methods work in unknown signal power conditions, and don't require the carrier synchronization. The estimation accuracy of the proposed method is evaluated by normalized mean square error (NMSE).

CiNii Article

Publishing type：Research paper (scientific journal)   Kind of work：Single Work  

DOI： 10.24544/ocu.20171218-087

CiNii Article

Publishing type：Research paper (scientific journal)  

Since network virtualization such as Software Defined Network (SDN) and Network Function Virtualization (NFV) is pervading in network services, network infrastructure became available with low cost and short period. Therefore, it became easier for 2nd providers to provide their own service by using network infrastructures provided by 1st provider. However, since 1st provider only provides limited functions of management plane, it is difficult for 2nd providers to conduct a smooth management of service. In this paper, we propose the architecture, Service Dependent Management plane (SDM), to provide partial management plane for 2nd provider to overcome this issue. Furthermore we introduce the user command specified implementation, Simple Management of Access-Restriction Translator Gateway (SMART-GW), and demonstrate its feasibility and performance so that our proposed model is feasible enough for practical use in the carrier class networks.

CiNii Article

Publishing type：Research paper (scientific journal)  

IPv6 has huge wide address space. By utilizing this good feature, we can easily emphasize the security of IPv6 communications. We are proposing a Unified Multiplex Communication Architecture that protects privacy and emphasizes the security by changing IPv6 addresses for each session actively. Since this mechanism is based on innovating usage of IPv6 addresses, anyone can use this easily without modifying existing applications. However, this type of actively address changing approach is completely new. In order to provide a simple and appropriate management method to control many IPv6 addresses, there are many problems to solve. In this paper, we report design and implementation of the method to solve them by setting Tag to each address in the kernel address pool.

CiNii Article

Publishing type：Research paper (scientific journal)  

Spinal codes are the new approach for rateless codes. Spinal codes are decoded by a dynamic search as well as convolutional codes which have a long constraint length. Near capacity performance of spinal codes is demostrated. However, the error performance of spinal codes has not been studied enough. In this paper, the bit error rate of fixed rate spinal codes is examined by computer simulations. The error rate of fixed rate spinal codes is presented for rates from 1/15 to 8/9. A new method for symbol mapping is proposed to improve the bit error rate of high-rate codes.

CiNii Article

Publishing type：Research paper (scientific journal)  

Spinal codes are new class of rate-compatible codes. Decoder of the spinal codes has employed M-algorithm to find the least cost path on the code tree and sorting operation of M-algorithm increases the complexity of the decoder. To reduce decoder complexity, threshold algorithms have been proposed. These algorithms require the channel state information to determine the threshold, therefore, inaccurate estimation of channel state information causes performance degradation and sensitivity of the threshold should be considered. In this paper, we propose the symbol mapping method considering sensitivity of thresholds. In addition, we also evaluate the sensitivity of thresholds of symbol counting based threshold algorithms over interference channels.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

Spinal codes were proposed in 2011 by J.Perry, et al. as a new class of rateless codes. Spinal codes have a high randomness by using a hash function, witch realizes the sperior performance. The performance approaches the channel capacity in the additive white Gaussian noise channel. Spinal codes are decoded by a dynamic tree search which is employed for comvolutional codes if long constraint length. Howerver, it has not been studied enough about the characteristics of the Spinal codes. In this paper, the error probability is evaluated in the condition of a fixed rate. The error probability is found to have the error floor. To improve the error floor, the methods of systematic coding and tail symbol insertion are presented.

CiNii Article

Publishing type：Research paper (scientific journal)  

The modulation identifications is one of the key issues in the cognitive radios, and many works were devoted to the modulation classification of symbol-by-symbol modulations. However, few works were presented of block modulations. They are, for example, a signal detection of OFDM. In this paper, the decision method is improved in the block length estimation, which exploits the cyclostationarity. Simulation results of correct estimation rate show the superior performance of the improved method. The method does not require the accurate symbol rate and center frequency.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

TCAM (Ternary Content Addressable Memory) is widely used for high speed searching applications on networking equipment. Unlike general RAMs, a TCAM consumes vast amounts of power, because it is designed to search its entire memory in a single cycle. As size of matching table grows continuously, it recently becomes serious. In this paper, we propose a novel TCAM-based hardware architecture, which solves potential restrictions of TCAM by enabling selective partial activate according to "don't care bits." For the purpose of reducing power consumed by TCAMs, we subdivide a TCAM into multiple sub TCAMs in LSI stage and stop look-up on sub TCAMs that consist of don't care bits.

CiNii Article

Publishing type：Research paper (scientific journal)  

Cache on the content deployment in network system plays the important role in the aspect of efficient content deliver. Since the capacity of cache on each network node is limited, a number of algorithms have been proposed. However, because of previous methods do not consider that the demand of content fluctuates rapidly, ineffectual cache of content remains on the cache. In this paper, we propose the method of using the prediction result of content demand. We first demonstrate that by applying prediction can improve the efficiency of cache utilization. However, predicting the demand of whole contents is not realistic in the aspect of computational cost. We therefore propose the method of reducing computational cost greatly without lack of cache efficiency. We demonstrate that by applying it to the real monitored data, cache hit rate can be increased about 60%.

CiNii Article

Publishing type：Research paper (scientific journal)  

In cognitive radios, signal detection and modulation classification are becoming more and more important to recognize wireless environments. The modulation is well classified by a simple method using moments. However, the classification performance degrades significantly by timing offset offset. In this paper, we present both estimation method of symbol timing by using amplitude moment and a QAM classification method based on combined moments to improve the classification error probability in the presence of symbol timing error.

CiNii Article

Publishing type：Research paper (scientific journal)  

Spinal codes are recently proposed rateless codes which employ the hash functions to generate codewords. In spinal codes, codewords are obtained by the hash function with splitting the encoder states into subsequences. However, design methods of spinal codes and hash function have not been proposed. In this paper, we consider the design method of the rateless codes by replacing the hash function of spinal codes with other random structures. Specifically, we propose the rateless codes construction based on codewords splitting using generator matrices of block codes.

CiNii Article

Publishing type：Research paper (scientific journal)  

Content-centric networking (CCN) that is an innovative network architecture requires routers with performance far superior to that offered by today's Internet routers. There are, however, few router-level designs incorporating all the necessary components. The design and evaluation of a complete router is the primary contribution of this paper. We provide a concrete hardware design for a router model incorporating two entities that we propose. Our contributions are (1) presenting a proper algorithm for looking up and matching name addresses in CCN communication, (2) proposing a method to process CCN packets in a way that achieves high throughput and very low latency, and (3) demonstrating performance and cost on the basis of a concrete hardware design.

CiNii Article

Publishing type：Research paper (scientific journal)  

In wireless sensor networks, multi-hop communications are often employed for power limitation of sensor nodes. A path diversity is an effective technique to get highly reliable communication since there are many paths from source to destination. In fading channels, by removing small signal, we can get highly reliable communication. In this paper, we propose a new relay method with threshold. The error probability of majority decision is analyzed, and an effect of a threshold change on error probability is presented to show that this proposed method is effective to fading in tree networks.

CiNii Article

Publishing type：Research paper (scientific journal)  

In wireless sensor networks, multi-hop communications are often employed for the power limitation of sensor nodes. A path diversity is an effective technique to get highly reliable communications since there are many paths from source to destination. In fading channels, a small signal has low reliability. In this paper, we propose a new multi-hop method which does not relay the small signal in a tree network. Simulation results show that our method is effective to fading in binary tree fading channels.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)   Kind of work：Single Work  

DOI： 10.11511/jacc.57.0_1099

CiNii Article

Publishing type：Research paper (scientific journal)  

IPv6 Link-Local address is supporting Auto Configuration and Neighbor Discovery of IPv6, and isn't in the specification of IPv4 and is one of the remarkable specifications that are produced by the IPv6. They are set to all IPv6 nodes automatically. Since IPv6 Link-Local Address owns different characteristics from accustomed existing Global address, new restrictions that require accompanying Zone-ID information to address information anytime are caused. This is the big problem for ordinal end users, and it makes difficult for them to deal with IPv6 Link-Local addresses. After analyzing the problem for using the IPv6 Link-Local address and Zone-ID, we propose "Zone ID Free" functions that solve above problems. They have been implemented and verified. Design and implementation of the "Zone ID Free" functions are reported in this paper.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

CCN is considered as a future network architecture to exchange contents without specifying addresses of nodes. One of advantages to adopt CCN is to improve the availability of network resource by using in-network caching. However, since nodes having a lot of popular contents may suffer heavy traffic load which leads frequent updates of caches in neighbor nodes, the efficiency of cache usage may be degraded. To solve this problem, we propose a method to widely spread contents by using random encoded addresses. From our simulation results, we show that our method can reduce up to 72% of cache size to achieve the same cache hit ratio when the access frequencies are biased to a specific node.

CiNii Article

Publishing type：Research paper (scientific journal)  

Software-Defined Network (SDN) is to decouple the control plane and the forwarding plane, and to provide programmable APIs to operate a centralized management of the whole network by describing a software program. In such centralized managements, network administrators can manage their own network flexibly and uniformly according to the knowledge of whole network. As the network size grows, a controller has to process a huge number of control message, and also has to handle the network information certainly, which are requested to update frequently. Thus, scalability of the controller and consistency of information update become important problems in SDN. In this paper, we propose a database-oriented management system for asynchronous configuration in SDN. We use database to ensure consistency of information update and configure network according to network conditions. Moreover, we apply our management system of SDN to Open Flow network and validate that Open Flow Network dynamically are reconfigured according to the database update.

CiNii Article

Publishing type：Research paper (scientific journal)  

Application identification in the middle is one of important technique for network operators to manage application based traffic and policy controls. However, due to the increase of encrypted traffic, it is becoming harder to read application specific information from packets. We previously proposed a method of application identification of encrypted traffic, and it can achieve high accuracy as same as the case of normal traffic. However, it requires the offline identification which uses statistics of the whole of flows, and it can't identify flows in real-time. In this paper, we propose a method of application identification in high accuracy and real-time. First, we clarify the sufficient number of packets to identify the application of encrypted traffic with high accuracy. Second, we proposed the method of inferring flow statistics in order to achieve high accuracy even when we can use smaller packets than required packets in identification. According to evaluation of using the real traffic trace, proposed method can achieve high accuracy almost as the same as in offline method by observing smaller packets than required packets in identification.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)   Kind of work：Single Work  

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Spinal codes are recently proposed rateless codes that achieve Shannon capacity over binary symmetric channel and additive white Gaussian channel. In the decoding of spinal codes, we compute a minimum cost path on the tree of codes. Tail symbol technique is adopted for improving the convergence of surviving path, however, degradation in the information rate due to the tail symbol cannot be negligible in the case of small information block size. In this paper, we propose the sliding block decoding of spinal codes and we also demonstrate the superiority of the sliding block decoding in terms of the information rate for the small information block size systems.

CiNii Article

Publishing type：Research paper (scientific journal)  

It is necessary to update the prediction model's parameter with the progress of the time to reduce the deterioration of the accuracy of prediction, since the time series data of traffic changes every moment. However, the frequency of parameter re-estimation differs greatly due to the characteristic of time series data. In this paper, we propose a method to detect the phase for re-estimation automatically by using the residual of prediction as feedback information. Our method reduces the prediction error by handling the phase of re-estimation and also reduces unnecessary re-estimation. We realized a lightweight and general control structure by using moving average and moving standard deviation. We demonstrate that by applying it to the real monitored data in a major P2P file-sharing system the traceability against variations of time series increased about 20%.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Content-Centric Networking (CCN) is a novel architecture for dealing with various problems facing the Internet. However, there are deployment issues and the shortage of testbed environments. We focus on OpenFlow, which is a programmable environment without disrupting existing networks. Specific implementations of CCN using OpenFlow have not been examined in sufficient detail, although there have been investigations into the conceptual design. Our approach can solve the issues in existing designs by using a map from content names to hierarchically structured hash values and the longest prefix match.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

If there is a system that can show information of each IP communication node and its users will which types of communications are wished or not wished properly to outside, networks are properly managed and secure and stable communication environments are accomplished. In order to enable and realize such a system. the following problems must be solved. 1. There is no proper place where is connected to each node and services that run on the node and show the information of them. 2. Even if there is a proper place for it, it is not easy to guarantee that the shown information is properly reflected wills of the node and the users who run it. In order to solve above problems, we propose a system that can provide information of the IP communication nodes reliably. By using the DNS applicably, above problems are solved and it becomes possible to operate the system easily and surely. We propose such a system and report design and implementation of it.

CiNii Article

Publishing type：Research paper (scientific journal)  

In free-space optical communications, the transmitter position estimation is required to keep LOS (Line of Sight)link to communicate steadily in case of low-power signals or existence of background noise. The accuracy of tracking is expected to be improved by using path history of transmitter. In this paper a method for tracking and communication is presented. The method is based on the extended trellis diagram. Simulation results show the performance improvement by our method.

CiNii Article

Publishing type：Research paper (scientific journal)  

Spinal codes are a recently proposed rateless code that achieves Shannon capacity on additive white Gaussian noise channel. Some breadth-first tree search algorithms are introduced for decoding spinal codes, how- ever, sorting operations of conventional decoding algorithms remain as critical throughput bottleneck. To achieve higher throughput of the decoder, non-sorting algorithms and highly parallerism are attractive. In this paper, we investigate a property of threshold algorithms for decoding spinal codes, i.e. , T-algorithm and metric-based thresh- old algorithm. These algorithms are evaluated from the view points of computational complexity and throughput, and we demonstrate the superiority of threshold decoding algorithms.

CiNii Article

Publishing type：Research paper (scientific journal)  

Content-Centric Networking (CCN) is promising as a future communication paradigm to exchange con- tents without specifying address of nodes. However, deployment of CCN from clean-slate would take a long time since all routing facilities are needed to be replaced. For early deployment of CCN it is feasible to use the current Internet infrastructure as much as possible, and achieve minimum modifications of applications and/or end nodes. Based on such background we propose a novel approach which enables to deploy CCN rapidly with extremely small modifications compared to other approaches. Our solution is based on the use of IPv6 network as a lower layer routing scheme. To realize we design some important components such as Name Mapper and Address Mapper. We also discuss about a deployment scenario for early adaptation of CCN in IPv6 networks.

CiNii Article

Publishing type：Research paper (scientific journal)  

Recently, energy consumption of routers has become a serious problem, hence power reduction is an urgent challenge. In particular, the occupancy of packet buffer is very low in core routers, which leads to unnec- essary consumption of electric power. To solve this problem, we proposed a new buffer architecture called Two Dimensional Sliced Packet Buffer. Dividing a whole buffer into multiple sub buffers at the LSI level enables to control power management independently according to its occupancy and traffic volume. In this paper, we show the optimal parameters to minimize power consumption. Through performance evaluation on power consumption and overhead, we show that optimal parameters are parallels 64 and serials 8, and our proposed packet buffer can reduce an average 50% of power consumption when an average input rate is 30%.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

We have proposed Sliced Router Architecture, which reduces the power consumption of routers by adjusting routers' performance according to the volume of traffic. In this architecture, it is crucial to consider not only packet loss but traffic latency. In this paper, we extend this architecture to reduce traffic latency under conditions of causing no packet loss. We also discuss the trade off in power-saving and worst traffic latency.

CiNii Article

Publishing type：Research paper (scientific journal)  

Due to the wide use and spread of network communications, importance of provide appropriate qualities for network service, for incetance, to each user or each service has become greater. We are trying to provide these qualities in access networks because facilities could be introduced easily in access networks compared to core and metro networks. On the contrary, recentry Software Defined Network (SDN) came to be actively conducted research region. Especially OpenFlow has attracted considerable attention as it enables flexible network control through dealing with traffic in each flow. Hence it can be suitable for providing qualities on the goals of us. Therefore in the first pahase we propose a route control method in cases where OpenFlow is utilizable in multihomed access networks. In addition we offer its evaluation results in OpenFlow multihomed networks established across the Internet.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

In wireless sensor networks, multi-hop communications are often employed for the power limitation of sensor node. A path diversity is an effective technique to get highly reliable communications. In this paper, we analyze the error performance of a majority decision in binary tree networks composed of fading channels. We evaluate the variance of bit error rate and and outage probability by a correlation matrix of decision variable, and show that the majority decision is effective to fading in binary tree fading channels.

CiNii Article

Publishing type：Research paper (scientific journal)   Kind of work：Single Work  

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)   Kind of work：Single Work  

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

This document discusses notion and actual mechanisms of Corresponding Auto Names for IPv6 Addresses. With this mechanism, all IPv6 addresses (even if they are link-local scoped addresses) can obtain their own Names, and it will be able to use Names anywhere instead of IPv6 Addresses. IPv6 address is too long and complicated to remember, and it is very nuisance thing to type a literal IPv6 address manually as an argument of applications. Also, it is very difficult for human beings to tell which IPv6 address is set to which actual IPv6 node. In this sense, literal IPv6 address information can be called meaningless information for human beings. In order to solve above problems and to make the information meaningful, mechanisms called Corresponding Auto Names for IPv6 addresses have designed and implemented. They become human friendly information. By applying a simple naming rule to the Auto Names (e.g., use the same Auto Name Prefix for IPv6 addresses that are set to the same interface (node)), this will contribute to help people to understand which IPv6 address / Name indicates which actual IPv6 node.

CiNii Article

Publishing type：Research paper (scientific journal)   Kind of work：Single Work  

DOI： 10.1587/transcom.E95.B.1901

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Our research group proposes Secure and Resilient Virtual Trust Routing (SeRViTR) to be able to provide trustworthy routing framework for both user-centric and network-centric routings. We currently develop a conceptual design and verify technical problems to realize SeRViTR. In this paper, we describe an implementation model of functional components in SeRViTR, and verify the trustworthy routing function and the trust management mechanism in inter-domain environment.

CiNii Article

Publishing type：Research paper (scientific journal)  

In this paper, we propose a new communication scheme, which utilizes IP address as a key of session, to realize a secure remote communication from outside network. In this scheme, IP address is an access key for a session by itself, and it varies for every session to reduce a security risk to be reused by anonymous parties. To realize this access scheme, we discuss on the listen model at the server, key updating, and key synchronizing issues. We also describe an implementation model of proposed scheme to show the feasibility.

CiNii Article

Publishing type：Research paper (scientific journal)  

Cognitive radio attracts much attention as a system to realize an efficient use of radio spectrum. In the cognitive radio, signal detection and modulation classification are key issues to recognize the radio channel condition. In this paper, new signal detection and modulation classification methods based on distribution tests are proposed. Both Anderson-Darling test and Cramer-Von-Mises test are examined for signal detection error probability and modulation classification error probability. Numerical results of error probability show that the proposed methods are effective to identify signals in the radio channel.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

With a trigger of the IPv4 address exhaustion, it is naturally supposed that the IPv6 introductions will be accelerate to various network environments. Also, it will become general that both IPv4 and IPv6 addresses are set to one communication node and IPv4 and IPv6 mixed communications become popular. From a viewpoint of the DNS, it becomes general that both A record (for IPv4) and AAAA record (for IPv6) are associated with one domain name. In regular DNS resolving that obtain IP address(es) with a domain name argument, it is needed to achieve to obtain heterogeneous IP addresses (IPv4 and IPv6) (heterogeneous records (A and AAAA)) with a domain name argument. With the current DNS resolving mechanism, it is possible to achieve above features. However, it is inefficient and problematic. Also, for general end users, it is not necessary to be strongly conscious of the difference of IPv4 and IPv6 addresses. Therefore, this DNS resolving becomes a significant operation that judges the IPv6 introduction to the network is done smoothly or not. This paper clarifies problems of the current DNS resolving mechanism under IPv4/IPv6 mixed environment that is significant operation for the IPv6 introduction and discusses efficient and improved DNS resolving mechanisms.

CiNii Article

Publishing type：Research paper (scientific journal)  

The complete network service of infrastructure is an urgent issue by means of continuous growth in network traffic. Apparently, the energy consumption of network routers may arise one of global environmental affairs, and therefore Research and Development works regarding power reduction is well desired. Our group proposes a unique feature embedded onto routers, which consists of plurality of slice, and is dynamically controlled by the prediction of network traffic flow. In this paper, we've examined slice control and LSI architecture, and have shown the validity of slice-architectural router.

CiNii Article

Publishing type：Research paper (scientific journal)  

The current Internet is designed to optimize the communications among the known hosts; the Internet protocol (IP) address of the destination node is known or otherwise resolved by inquiring the domain name system (DNS) server. However, highly intelligent network is expected in the future Internet that can route packets to the appropriate destination not only with an numerical identifier but also with information which is often referred as content-centric network. In this paper, we discuss that routing by topic names is feasible in network layer rather than in conventional application layer. In addition, we show that looking up topic names is feasible by evaluating the proposed router architecture.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Ternary content addressable memory (TCAM) is becoming very popular for designing high-throughput forwarding engines on routers. However, TCAM has potential problems in terms of cost of hardware and power. We have proposed the RAM-based new hardware architecture for a fast forwarding engine that looks up a route by a single access. In this report, we evaluate the hardware cost of our forwarding engine by chip area. From the evaluation results, we confirm the optimal parameter which minimizes the hardware cost.

CiNii Article

Publishing type：Research paper (scientific journal)  

Some studies have shown that using by a single scale-free network is not appropriate for modeling social networks that represent human relationships in real world. In the social networks, multiple community networks based on common interests or values exist together. However, there is no study on communication model where multiple community networks are taken into consideration. Moreover, the evaluation that considers the load-balancing of the hub node in the scale-free network is done as an application of the information search by crossing of multiple networks. In this paper, we investigate how the performance of information retrieval is affected by using multiple community networks. To evaluate the performance, we use a degree of density which denotes the processing cost of hub nodes as well as the distance.We show that the performance of information retrieval by crossing different community networks improves in terms of the number of hops and degree of density on any path between two nodes.

CiNii Article

Publishing type：Research paper (scientific journal)   Kind of work：Single Work  

CiNii Article

Publishing type：Research paper (scientific journal)  

Recent years, there are many studies on identification of application from the monitored traffic in the Internet where various types of application traffic coexist. Application Identification is one of effective technologies for network management because application-based traffic control would be realizable based on results of identification. However, accurate identification is becoming difficult by using methods proposed so far due to the increase of encrypted traffic. In application identification, encryption is one of serious factors to degrade the accuracy of identification because it is so hard to infer the original (unencrypted) traffic from the encrypted one. Our research goal is to improve the accuracy of application identification whatever the target traffic is encrypted. In this paper, we propose a method of application identification from encrypted traffic based on characteristic changes by encryption. By applying this method, we show that the identification accuracy of application for the encrypted traffic can be improved.

CiNii Article

Publishing type：Research paper (scientific journal)  

Cognitive radios are intelligent communications, and are expected to realize an efficient use of the radio channel. The modulation identifications is one of the problem in the cognitive radio. As for the modulation identifications for block orthogonal modulation, the method of clustering and labeling was proposed. However, the identification performance has been evaluated only by the computer simulation. In this paper, error probability analysis is presented for a block orthogonal modulation identification using General Orthogonal Modulation(GOM). The analysis includes an inner product of two signals and noise. This causes difficulties in the analysis. The multidimensional rotation is employed for the difficulties in order to derive error probability. Numerical results confirm the validity of analysis.

CiNii Article

Publishing type：Research paper (scientific journal)  

In network management, application identification of traffic is important. However, since encrypted traffic has been increasing in recent years, it is becoming difficult to identify the kind of application by methods proposed so far. In application identification, encryption is one of serious factors to degrade the accuracy of identification because it is so hard to infer the original (unencrypted) traffic from the encrypted one. Our research goal is to increase the accuracy of application identification even if there is encrypted traffic. In this paper, we numerically analyze how the characteristics of original traffic are changed by encryption. Specifically, we monitor packets of the same flow, at the points of before/after encryption. We then collect statistical information on monitored flows, and numerically analyze how this information has been changed by encryption. By applying regression analysis, we show that we can estimate the statistical characteristics of original traffic from measured encrypted traffic.

CiNii Article

Publishing type：Research paper (scientific journal)  

Computer simulations are widely used for an error probability estimation of convolutional codes. Few studies have been made for exact bit error probability of convolutional codes of more than 4-state. In this paper, an analytical approach presented for convolutional code with MAP decoding. Moments of decision variable are desired by the recurrence relations, and discussed with its probability density function, which leads to the exact error probability.

CiNii Article

Publishing type：Research paper (scientific journal)  

Ternary content addressable memory (TCAM) is becoming very popular for designing high-throughput forwarding engines on routers. However, TCAM has potential problems in terms of hardware cost and energy consumption, which limits it deploying large amounts of capacity. In this paper, we propose a new hardware architecture for a fast forwarding engine that fundamentally solves the potential problems of TCAM. We also develop a hardware design for our architecture. Our results show that the proposed hardware significantly reduces hardware cost and power consumption.

CiNii Article

Publishing type：Research paper (scientific journal)  

In recent years, in distributed cooperative systems such as distributed file systems and online games, equalization of performance between collaborative nodes is one of impotant issues as well as explicit performance such as bandwidth, delay and packet loss rate. Selection of the server which can offer equal performance to collaborative nodes is necessary to offer equalized performance. To achieve this problem, we have proposed so far a server selection method that select the server by calculating the circumcenter of collaborative nodes according to the distance between nodes obtained by performance measures. However, depending on the location of nodes, there is a problem that using circumcenter is not always sufficient to find the most appropriate server. In this paper, we propose a method to improve the accuracy of server selection by solving the problem of using circumcenter by downhill simplex method. We also show probability of selection of the most suitable server can be improved around 55% compared to the previous proposal.

CiNii Article

Publishing type：Research paper (scientific journal)  

In P2P file sharing systems, time dependent characteristics of queries (query trends) become much important to forecast the demand of file in future. Prediction of future demand would be effective for the efficient use of e.g., caching mechanism, however, accurate prediction of query trend is difficult because patterns of query trends may differ significantly according to a nature of keyword. Classification of query trend patterns is on of effective ways to forecast method for future demand. We previously proposed the method to classify trend patterns based on DFT and Ward clustering method. In this paper, we propose a method for automatic classification of query trend patterns accurately. According to the results, we show our proposed method can classify trend patterns by the high probability of 88.3%.

CiNii Article

Publishing type：Research paper (scientific journal)  

This paper describes IPv6 neighbor cache update/delete methods from external nodes. In IPv6, mapping between datalink layer (MAC address) and network layer (IPv6) addresses are maintained by a neighbor cache in each node. After the neighbor discovery protocol, an entry of neighbor cache is created, and used the cached information during the communication. RFC4861 defines neighbor discovery methods and describes how to manage neighbor cache entries by neighbor discovery messages. However, RFC4861 does not define quick and clear neighbor cache update (delete) functions. This document proposes two types of IPv6 neighbor cache update (delete) methods from external nodes. One is a heuristic type method that does NOT require neighbor discovery message extensions. The other is an explicit type method that requires small neighbor discovery message extensions. We also provide a combined method to enable fast and sure deletion of the entry.

CiNii Article

Publishing type：Research paper (scientific journal)  

We have proposed a new communication architecture called "Unified Multiplex Communication Architecture." Two new types of addresses are introduced. One is EA (Ephemeral Address) used at client nodes, the other is SSA (Specific Service Address) used at server nodes. They are exclusively used only for each communication session. Since they are introduced only by modifying the kernel, it is not necessary to modify the existing applications to use these addresses. End users can easily acquire the advantages of new types of communications. These addresses are enabled only when their sessions are valid. Therefore, it is easy to provide security enhancement and privacy protection functions. The new communication architecture can evolve the communication style from one-fixed to multiple-floated and from share to dedicate. The paper discusses the detailed functions that enable these features from the viewpoints of relationship between IP addresses and communication processes / sessions. The described functions have been implemented and verified.

CiNii Article

Publishing type：Research paper (scientific journal)  

We have proposed a new communication architecture that is called "Unified Multiplex Communication Architecture." By introducing new types of addresses that are assigned to and consumed at each communication session, the architecture can provide secure communication services without requiring special operations to end-users. We should reconsider the role of ICMP, since an address changes a concept from the identifier of a node to the identifier of service. In our architecture the target of ICMP, which is used to advertise control message for the node, change to the service. Therefore ICMP advertises control message in cooperation with a service by limiting response term or target node. In this paper, we clarify the function of ICMP suitable for Specific Service Address. First we show that the new role of ICMP for Unified Multiplex Communication Architecture. Furthermore, we propose a new technique to realize the flexible information mechanism of ICMP.

CiNii Article

Publishing type：Research paper (scientific journal)  

In future generation network, routing on contents or resource is anticipated as the post-Internet Protocol (IP) routing method. The final goal of our research is to show the feasibility of routing based on resource in Network Layer. As an initial step, we use 'name' for routing, particularly 'Fully Qualified Domain Name (FQDN)' to show the feasibility and to generalize it into resource-based routing. When writing the routing information of FQDN which has a hierarchical structured name into hierarchical logical topology, mismatch can occur if the physical topology's information is missing. In addition, routing tables should be reorganized to reflect the different access frequency among the FQDNs. In this paper, we propose a mapping algorithm of logical to physical topology and the reconstruction of routing tables. As a result, we show that using access frequency and physical topology's information increases the efficiency for searching FQDN.

CiNii Article

Publishing type：Research paper (scientific journal)  

We have proposed Unified Multiplex Communication Architecture that providing specific service addresses. The architecture can provide secure communication service by utilizing huge addressing space of IPv6. In this paper, we realize more flexible access control mechanism by embedding attribute information to IP addresses. For realizing this mechanism, all nodes on the network operate based on the rule that is made by network administrator. Finally, we design and implement address management mechanism that enables to provide diversity of specific service addresses.

CiNii Article

Publishing type：Research paper (scientific journal)  

In this paper, a new modulation classification method is presented for multiple signals composed of large desired signal and small interference. Using higher-order moments of the received signal, the transmitted signal modulation and power are estimated by the method. Interference is classified by the estimated signal modulation with the higher-order moments. Numerical results of the classification error probability show that this method is effective for the classification of multiple signals.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

When some wireless communication systems operate in the same frequency band, it is essential to employ the techniques for interference mitigation. The goal of this paper is to investigate the performance of some mitigation techniques and to show the superiority of metric based limiting, which yields the better bit error rate performance.

DOI： 10.11485/iteac.2010.0_9-6-1

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

無線センサネットワークでは,センサノードで取得した情報をネットワークを用いて収集する.個々のノードは容量や消費電力に制限があるため,大電力による通信は不向きである.このような無線センサネットワークにおいて,小電力を維持したまま信頼度の改善を行うためにはマルチホップ通信が有効である.本稿では,センサノードの情報を2次元平面にランダムに配置した中継ノードを介してフュージョンセンターに伝送する無線センサネットワークについて考える.多数のマルチホップ経路を2分木でモデル化し,フュージョンセンターで多数決判定する場合の誤り率を導出し,誤り率と中継段数の関係を明らかにする.

CiNii Article

Publishing type：Research paper (scientific journal)  

近年の電波需要の増加に対し利用可能な周波数帯は限られたものであり,802.11無線LANやBluetoothなどの複数の無線規格が同一の周波数帯に混在するような状況が生じている.このような状況下では機器間での干渉により深刻な性能劣化が引き起こされ,干渉の影響を軽減する方式が重要となる.本稿では,ビタビ復号を用いたシステムにおいて,干渉の影響を考慮した複数の方式について比較を行い,メトリックに対するリミッティングが干渉波の影響を軽減する方式として有効であることを明らかにする.

CiNii Article

Publishing type：Research paper (scientific journal)  

In next-generation network architectures, the number of nodes and equipment connected to networks increase more than ever. Due to the explosive increase of network equipment, an information retrieval technique to access surely and flexibility will be required. For this, we have proposed Generalized Routing Protocol with variable-length address information. Generalized Routing Protocol can achieve flexible information retrieval that has been implemented on overlay networks, such as DHT and DNS etc.. In this paper, we implement routing mechanism for variable-length address of Generalized Routing Protocol. Moreover, we verify whether routing information including variable-length address can be stored in routing table of GRP routers and exchanged between GRP routers.

CiNii Article

Publishing type：Research paper (scientific journal)  

In recent years, various distributed applications such as file sharing systems, online games, and multicast streamings are widely used. Some of distributed applications need to equalize their performance to every participant. For example, from the viewpoint of user experience, it is preferable for distributed file system to provide a steady throughput wherever the client are connecting to the server. Also, in online games, synchronization between players is one of important factors for smooth progress of the game, therefore equalization of delays among joining players is important. However, there is no generalized method to selecting the best node for equalization of network metrics whose importance are different according to applications. In this paper, we propose a server selection method for equalizing various performances in distributed systems. Our approach is based on surveying techniques especially trilateration. As an example of equalization, we evaluate the proposed method in the case of equalization of RTTs, and show our method can reduce around 50% of performance difference compared to random server selection cases.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

In wireless sensor networks, many sensor nodes with relay transmission aggregate information. In this paper, we demonstrate error rate performance improvements due to majority decision which produce path diversity effects by a lot of routes in wireless sensor networks. We generate a topology from 2-dimensional randomly deployed nodes, and evaluate the performance by computer simulations and clarify that the path diversity by majority decision is effective for a reliable sensor network.

CiNii Article

Publishing type：Research paper (scientific journal)  

To design guaranteed high-performance router, it is problem that packet buffer is non-deterministic. We propose Head Buffer MMU which can realize high-speed data transfer guaranteed packet buffer. Moreover, the MMU can provide large buffers with cost-effective and low power consumption. In this paper, we design the bank selection circuit which can guarantee DRAM high-speed data transfer in Head Buffer MMU. We then show power saving of DRAM by using the bank selection circuit.

CiNii Article

Publishing type：Research paper (scientific journal)  

DOI： 10.1007/s10922-009-9121-9

Publishing type：Research paper (scientific journal)  

In recent years, the growth of wireless services has fostered many communication protocols in 2.4GHz band, including 802.11 Wireless LAN, Bluetooth, and so on. When some wireless communication systems work in the same frequency band, it is essential to employ the techniques for interference cancellation. In this paper, we present an interference cancellation method which employs the erasure bit decoding and ring trellis. The method is applicable to multiple pulse interferences, although the conventional erasure bit decoding method involves the number of erasure bits limitations. In addition, the method is evaluated in an interference channel from the view points of decoding bit error rate by computer simulations.

CiNii Article

Publishing type：Research paper (scientific journal)  

In P2P file sharing systems, caching of contents is one of practical methods to improve the search efficiency and the load balancing. Since the resources (e.g., disk spaces) for caching are limited, the contents to be cached may vary according to the cache replacement algorithm. However, typical cache replacement algorithms do not consider the time-dependent trends of query keywords. Therefore, the problem that many of caches that might not be effectively used remains comes to happen. Therefore, the cache replacement algorithms is necessary. In this paper, we analyze the time-dependent trends of query keywords in P2P file sharing system towards a new cache replacement method which takes query trends into consideration. First, we collect keyword (filename) queries by crawling the P2P file sharing system. We then analyze similarities of time-dependent trends of queries by using Discrete Fourier Transform. We then categorize trend pattern by applying the hierarchical clustering. As a result, we show that time-dependent trend patterns of query can be categorized into four types.

CiNii Article

Publishing type：Research paper (scientific journal)  

In recent years, due to the increase of the number of users connecting to the Internet, degradation of communication quality becomes one of the important problems. Such quality degradation is caused by competition of network resources between users/services, or interference between traffic with different properties. To overcome this problem, DiffServ and IntServ have been proposed as fundamental techniques. Although these approaches guarantee metrics such as end-to-end delay, packet loss rate, and transfer bit rate etc., feedbacks from the users are not considered. Moreover, recent studies have considered the feedbacks from users in controlling the QoS in real-time applications. However, evaluation of the quality is performed by an absolute assessment Therefore, we investigated a threshold which most of users do not feel dissatisfaction with the QoS, by performing experiment of user's subjective assessment using a comparative evaluation. However, we did not consider diversity of user's network environment. In this paper, we investigate environment-independent parameter in bulk data transmission by performing experiment regarding diversity of user's network environment.

CiNii Article

Publishing type：Research paper (scientific journal)  

In network management, identification of traffic type is important. However encrypted traffic has been increasing (in recent year), and so it is becoming difficult to identify the traffic type. The identification of application from the encrypted traffic is generally difficult, and so it is expected that existing method to identify the application can't operate accurately. One of method to overcome this problem is to detect whether traffic is encrypted or not, and identification is operated for traffic remove encryption overhead from encrypted traffic. In this paper, we study the method to detect encrypted traffic early. We analyze statistical information of encrypted traffic and plaintext traffic. We show statistical information which has remarkable characteristics in each traffic. We then propose the method to detect encrypted traffic by using the information.

CiNii Article

Publishing type：Research paper (scientific journal)  

Recently cognitive radios attract much attention, because it leads to flexible and high capacity wireless systems. Modulation identification is the first step and a key issue for cognitive radios. In this paper, a new modulation classification method based on amplitude and cosine moments is proposed. Our proposed method employs two dimensional decision variable composed of amplitude and cosine moments. Numerical results of the classification error probability show that the method presents superior identification performance for QAM compared with amplitude moments or cosine moments approach.

CiNii Article

Publishing type：Research paper (scientific journal)  

We propose Unified Multiplex Communication Architecture that can provide secure communication by utilizing huge addressing space of IPv6. The advantage of Unified Multiplex is that it does not require any modifications of existing application to support exclusive use of IP address to each service. This system has a big feature regarding the use of IP addresses, i.e., a client assigns an Ephemeral Address (EA) for every session as an identifier of the session. Also, a server assigns Specific Service Address (SSA) for exclusive use for the connection from client. These types of addresses are completely new compared to legacy type of IP addresses that only identify the communication node. Therefore the characteristics of these types of addresses are still unclear to secure the correct use of them. In this paper, we analyze the fundamental characteristics of such service-oriented addresses (EA and SSA), especially in terms of usage, lifetime, and management. We then clarify what kinds of functionalities are required to realize these characteristics, and describe its implementation method.

CiNii Article

Publishing type：Research paper (scientific journal)  

Next-generation network architectures will be governed by the need for flexibility. Therefore, Information access by not only IP addresses but also variable-length address information such as text or device-name will be required. In addition, the current Internet provides a number of the routing functions. For Instance, various service utilizes intelligent routing in overlay network. This reason is because the current routers is specialized in IP, and the routers can process only IP addresses. However, the routing process in these routing functions differ only slightly. The similar functions are realized with different layers. The decentralization of such an equivalent function causes a lot of problems like the inefficiency of processing and the mutual interference between layers, etc. in the network architecture. In this paper, we design and implement a new routing protocol, which can process the generalized addresses and support various protocol as advanced functions.

CiNii Article

Publishing type：Research paper (scientific journal)  

Recently, in P2P systems such as the file-sharing system, traffic in the network is increasing enormously, so that the necessity for an improvement have been widely recognized. For the improvement of the P2P systems, constructing an overlay network based on a single attribute of the node, e.g., similarity of contents and geometric location of nodes, has previously been studied. When an overlay network is constructed in consideration of only a single attribute, however, there exists a mismatch between overlay topologies that are constructed based on different attributes, search efficiency falls as a result. Therefore, we have proposed the technique to optimize the overlay network with the support of multiple attributes. But in our previous technique, the overhead to support multiple attributes in overlay networks is large, and the search hit ratio may slightly fall. In this paper, we propose the new technique to realize the overlay network, namely SMART, that can support multiple attributes with lower overhead. SMART is based on two key ideas: 1) similarity of overlay topologies that are constructed based on different attributes, and 2) exchange of indices to merge multiple overlay networks. Through the performance evaluation using a simulation, we show that SMART supports multiple attributes with lower overhead and increases search efficiency.

CiNii Article

Publishing type：Research paper (scientific journal)  

When multiple flows using different versions of TCP exist together on the network, they are influenced seriously between each other. One of methods to overcome this problem is to control such flows individually by identifying the TCP versions of each flow. Thus, this process needs to identify online which TCP version is used in each flow. In this paper, we first focus on the method for controlling the congestion window in different versions of TCP. We find out the characteristics value from the method, and then propose a method to identify flow by using the characteristics. We show that any two out of fourteen TCP versions which are implemented in Linux can be identified by using our proposed method.

CiNii Article

Publishing type：Research paper (scientific journal)  

We have proposed a new communication architecture that is called "Unified Multiplex Communication Architecture." By introducing new types of addresses that are assigned to and consumed at each communication session, the architecture can provide secure communication services without requiring special operations to end users. The architecture has been designed to coexist with current communication environments and to be migrated to gradually. In the architecture that innovates address dealing methods drastically but can coexist with the current, communication infrastructure (e.g., DNS) and communication control tables (e.g., Routing Table, Neighbor Cache) themselves are not changed. However, their contents information dealing methods are changed. It is very important to evaluate influence to the current architecture that is caused by introducing the new architecture. This paper analyzes and evaluates the influence from various viewpoints with assuming the Unified Multiplex Communication Architecture deployment situations.

CiNii Article

Publishing type：Research paper (scientific journal)  

The IPv4 (Internet Protocol version 4) addressing scheme has been the standard for Internet communication since its establishment in the 1960s. However, due to the technological advances in router technology and the enormous increase in Internet traffic usage, this addressing can no longer cope with various issues. In the past some proposals for partially solving these problems have been presented, but the proposals show limitations when it comes to establishing the foundations of future generation networks for which more sophisticated routing protocols, like content-based routing, are required. Furthermore, those previous approaches were not conceived to fully utilize the advantages of Ternary Content Addressable Memory (TCAM), which is a type of memory capable of performing high-speed lookups and is already implemented in high-end routers. In this thesis, we show that routing based on domain names in Network layer, which has been proposed as substitute for IP routing in the future, is already now a feasible technology and we evaluate the necessary hardware and network resources to implement name-based routing strategies. We present a routing scheme and propose three methods for equally balancing the routing information in the TCAM of multiple routers.

CiNii Article

Publishing type：Research paper (scientific journal)  

We propose and develop new communication system that realizes both anonymity of client nodes and distinction of client nodes, by using Ephemeral Address of Unified Multiplex communication architecture. In general, node distinction is much harder if anonymity of node improves. However, proposed communication system achieves both anonymity of node and distinction of node, by using dynamic address generation and address pool mechanism that Unified Multiplex communication architecture has. In this paper, we describe the design and the building method of proposed system, and then, we evaluate the effectivity of the system.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

The end-to-end round trip time (RTT) is one of the most important communication characteristics for Internet applications. From the viewpoint of network operators, RTT may also become one of the important metrics to understand the network conditions. Given this background, we should know how a factor such as a network incident influences RTTs. It is obvious that two or more factors may interfere in the observed delay characteristics, because packet transmission delays in the Internet are strongly dependent on the time-variant condition of the network. In this paper, we propose a modeling method by using mixed distribution which enables us to express delay characteristic more accurately where two or more factors exist together. And, we also propose an inferring method of network behavior by decomposition of the mixed distribution based on modeling results. Furthermore, in experiments we investigate the influence caused by each network impact factor independently. Our proposed method can presume the events that occur in a network from the measurements of RTTs by using the decomposition of the mixed distribution.

DOI： 10.1587/transcom.e92.b.1233

CiNii Article

Publishing type：Research paper (scientific journal)  

The increased demands for wireless communications lead to the development of cognitive radio, which realizes an effective utilization of channel resources. In the cognitive radio, the cognition takes an important role, and a modulation should be classified correctly at receivers. In this paper, an identification of block orthogonal modulations is discussed using general orthogonal modulations. We propose clustering method to estimate basis vectors of orthogonal modulations, and symbol estimation method by the estimated basis vectors. Bit error rate of the estimated modulation is evaluated in an additive white Gaussian noise channel by computer simulations.

CiNii Article

Publishing type：Research paper (scientific journal)  

When multiple flows using different versions of TCP exist together on the network, they are influenced seriously between each other. One of methods to overcome this problem is to control such flows individually by identifying the TCP versions of each flow. Thus, this process needs to identify online which TCP version is used in each flow. In this paper, we first focus on the method for controlling the congestion window in two different versions of TCP (TCP Reno and one of the other versions). We then propose a method to identify flow by using the characteristics. We show that TCP Reno and other TCP versions can be identified by using our proposed method.

CiNii Article

Publishing type：Research paper (scientific journal)  

When a carrier frequency and a phase are not recovered with the sufficient accuracy in a receiver, a signal point rotates in a vector diagram. The signal point rotation causes a phase offset. In this paper, the optimal modulation classification method based on cosine moments with the phase offset is considered. At the front end of modulation classifier, the small signals are rejected, and only the large signals are fed into the classifier to improve the performance. Numerical results of the classification error probability show the effectiveness of proposed methods with phase offset for QAM classification.

CiNii Article

Publishing type：Research paper (scientific journal)  

Distributed denial-of-service attacks on public servers have recently become more serious. Most of them are SYN flood attacks, since the malicious attackers can easily exploit the TCP specification to generate traffic making public servers unavailable. We need a defense method which can protect legitimate traffic so that end users can connect the target servers during such attacks. In this paper, we propose a new framework, in which all of the TCP connections to the victim servers from a domain are maintained at the gateways of the domain (i. e., near the clients). We call the nodes maintaining the TCP connection <i>defense nodes</i>. The defense nodes check whether arriving packets are legitimate or not by maintaining the TCP connection. That is, the defense nodes delegate reply packets to the received connection request packets and identify the legitimate packets by checking whether the clients reply to the reply packets. Then, only identified traffic are relayed via overlay networks. As a result, by deploying the defense nodes at the gateways of a domain, the legitimate packets from the domain are relayed apart from other packets including attack packets and protected. Our simulation results show that our method can protect legitimate traffic from the domain deploying our method. We also describe the deployment scenario of our defense mechanism.

DOI： 10.1093/ietcom/e91-b.8.2618

CiNii Article

Publishing type：Research paper (scientific journal)  

In P2P file sharing systems, caching of contents is one of practical methods to improve the search efficiency and the load balancing. Since the resources (e.g., disk spaces) for caching are limited, the contents to be cached may vary according to the cache replacement algorithm. However, typical cache replacement algorithms do not consider the time-dependent trends of query keywords, many of caches are not utilized effectively, which limits the improvement of performance by caching. In this paper, we propose a new cache replacement algorithm by taking the time-dependent trends of query words into consideration. Specifically, we categolize most frequent search words in Google into four types according to time-dependent variance of frequency obtained by Google Trends, and propose a newcache replacement algorithm based on categolization. Through numerical examples we show that our method can improve 24% of cache utilization compared to current replacement algorithms.

CiNii Article

Publishing type：Research paper (scientific journal)  

Cognitive radio is expected to realize an efficient use of the radio channel. In the cognitive radio, the cognition takes an important role, and a modulation should be classified correctly at receivers. In this paper an identification of block orthogonal modulations is discussed using general orthogonal modulations. It is necessary to reconstruct a multidimensional hypercube which is composed of transmitted signals to estimate the modulations. We propose clustering and labeling methods to reconstruct any orthogonal modulations by using feature of general orthogonal modulations.

CiNii Article

Publishing type：Research paper (scientific journal)  

In this paper, a new modulation classification method based on cosine moments without small signals is proposed. At the front end of modulation classifier, the small signals are rejected, and only the large signals are fed into the classifier. As the method of small signals rejection, the circular rejection and the rectungular rejection are considered. Numerical results of the classification error probability show the superiority of cosine moment method for QAM classification.

CiNii Article

Publishing type：Research paper (scientific journal)  

Recently, Peer-to-Peer (P2P) networks have been receiving increasing attention. Many studies on information have done so far, in order to achieve the search efficiency. Especially, it is important for P2P Networks to be optimized based multiple attributes; e.g., similarity of contents, and geometric location of nodes. However, most of literatures have focused on the optimization of overlay network topology based on single attribute, and it is not easy to optimize the topology for another attribute with keeping optimized for the attribute. Since the topology optimized for one attribute is not likely to be optimized for another attribute, there exists a mismatch between overlay topologies optimized by different attribute. In this study, we propose a method to optimize the overlay network with the support of multiple attributes. We focus on the similarity of the property of overlay topologies optimized by different attributes. Based on the similarity we introduce the exchange of indices which supports additional attributes in the overlay topology optimized by single attribute. Through the performance evaluations, we show that the overlay network optimized based on multiple attributes using our proposed method has both of characteristics of the overlay networks optimized based on single attributes, separately. Furthermore, the results show that the amount of the search messages and the message forwarding delay in information retrieval can be reduced by our method.

CiNii Article

Publishing type：Research paper (scientific journal)  

TCAM (Ternary Content Addressable Memory) is widely used for high speed address lookup. However, TCAM has a potential problem on hardware cost and energy consumption, which limits to deploy large amount of capacity in IP routers. In this paper, we propose a new TCAM-based hardware architecture which reduces significantly power consumption. The proposed architecture consists of multiple sub TCAMs and a indexing table. We also propose a method to select appropriate bit posisions used for indexing based on the statistical characteristics of routing table. Our numerical results show that the proposed method can divide the whole routing table neally equal to sub TCAMs.

CiNii Article

Publishing type：Research paper (scientific journal)  

In this paper, a new modulation classification method based on cosine moments with a constellation limiter is proposed. The cosine moments are derived in a form of the infinite series of elementary functions to analyze classification error probability. Numerical results of the classification error probability show the superiority of the proposed method. Especially, the proposed method is demonstrated to yield better performance than the method based on the amplitude moments.

CiNii Article

Publishing type：Research paper (scientific journal)  

In recent years, various network services are provided by ISPs, as end users increase because of diffusion of the Internet. ISPs try to provide QoS of network service, however, increase of end users may be cause degradation of quality of network. To ensure user's satisfaction, it is necessary to prevent increase of end-to-end delay and narrowing of bandwidth. However, satisfaction of each user is subjective assessment. It is difficult to understand user's satisfaction from bandwidth and delay. Therefore, QoE (Quality of Experience) is focused recently as a indicator of user's satisfaction. In this paper, we investigate user level QoE by using contant method and verbal rating method which are psychological measuring methods in data downloading. Moreover, we investigate acceptable level of user in degradation of bandwidth of communication link.

CiNii Article

Publishing type：Research paper (scientific journal)  

Recent studies have shown that the topology of the Internet is categorized as a scale-free network. However, previous studies have not discussed whether the current Internet is optimized in terms of attack tolerance and communication efficiency. Therefore, we showed the topology of the Internet can be reorganized to more appropriate topology for communication by changing individual's connecting strategy. In our previous work, we randomly selected target nodes that are changed connecting strategy. The effect of reorganization by changing strategy on randomly selected nodes may not reflect the effect of reorganization on the current Internet correctly. In this paper, we evaluate the impact of changing strategy to network topology by adopting generic diffusion model instead of selecting target nodes randomly. Moreover, we discuss the feasibility that our strategy can apply to the current Internet.

CiNii Article

Publishing type：Research paper (scientific journal)  

We propose Unified Multiplex Communication Architecture which can be considered security by using disposable address for services. In this paper, we first analyze the architecture by itself from the technical viewpoint, and describe the detailed design including new concepts needed to realize. We then present the technical descriptions of kernel/userland behavior in the Unified architecture.

CiNii Article

Publishing type：Research paper (scientific journal)  

We propose Unified Multiplex Communication Architecture which can be considered security by using disposable Specific Service Addresses (SSAs). In this architecture, generation of SSA is one of important function-alities to improve the security and the privacy protection. This paper describes a variety of generation techniques for SSA and discusses what kind of SSA generation is suitable for privacy protection.

CiNii Article

Publishing type：Research paper (scientific journal)  

We propose Unified Multiplex Communication Architecture that can be considered security by using disposable address for services. In this document, we test existing major applications performance on Unified Multiplex Communication Architecture kernel, check whether applications operate or not, and clear up problems on this kernel. Especially, we present legacy nodes and Unified Multiplex nodes can communicate with each other under mixed nodes.

CiNii Article

Publishing type：Research paper (scientific journal)  

We propose a new communication architecture that can provide secure services by introducing disposable SSA (Service Specific Address)s. In the new architecture, a large number of dynamically generated addresses are used. In use of dynamically generated addresses, we face new problems (e.g., newly generated addresses can not be used soon after they are generated and addresses can not be reserved before they are used.) that are not appeared in the current statically generated address usages. This paper proposes a new address state called "Uncertain State" that solves above problems comprehensively without causing bad effects to existing implementations.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

DOI： 10.1007/s10922-007-9089-2

Publishing type：Research paper (scientific journal)  

A binary symmetric channel(BSC)is considered in a binary tree network. End-nodes of the network are Connected to a fusion center, which sums up decision data of the end-nodes to make a decision variable When the decision variable is applied to a hard decision, the error probability of the fusion qenter was shown to be almost the same on the BSC crossover probability[1]. In this paper, a probability density of the decision variable is demonstrated, and convolutiollal coding / soft decision Viterbi decoding is applied to the network. The error ptobability is analyzed to show the error probability improvelnent by coding.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

TCAM (Ternary Content Addressable Memory) is a special type of memory used in routers in order to achieve high speed packet classification. Port numbers are one of the five typical fields in an ACL (Access Control List). Those that are expressed in ranges have been the reason for raising the cost of hardware due to the requirement of multiple entries. In this paper, we suggest a method to reduce the number of entries when expressing ranges in TCAM, using prefix expansion and hardware modification.

CiNii Article

Publishing type：Research paper (scientific journal)  

Network coding realizes the maximum flow in a network by an appropriate coding technique. In this paper, the low density parity check (LDPC) codes are applied to network coding for a multicast network composed of a tree graph. Parts of LDPC codeword are relayed in the network. At each node, received parts of codeword are regarded as punctured codes and restore the data. Numerical results show that a node recovers the data correctly when 63% of codeword is received in a certain network condition. The effects of noise on the block correct rate are also discussed.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

When multiple flows using different versions of TCP exist together on the network, they are influenced seriously between each other. One of method to overcome this problem is to control such flows individually by identifying the TCP versions of each flow. Thus, this process needs to identify online which TCP version is used in each flow. In this paper, we first focus on the increasing rate of the congestion window in two defferent versions of TCP (TCP Reno and HSTCP (High Speed TCP)). We also show that increasing rate of congestion window has remarkable characteristics in each TCP version. We then propose a method to identify flow by using the characteristics. We show that TCP Reno and HSTCP can be completely identified by using our proposed method.

CiNii Article

Publishing type：Research paper (scientific journal)  

Revent works have shown that topologies of various networks are categorized as scale-free networks. However, previous works have focused on investigation of scale-free nature of the Internet. Moreover, they have proposed more appropriate network services and protocols that take advantages of scale-free network. Network topologies which are more suitable for the Internet have not been considered. We previously investigated how node attacks affect scale-free networks with different characteristics. We also showed a possibility that a more appropriate topology for the Internet exists, which topology satisfies in terms of attack tolerance and communication efficiency. In order to change the current Internet to more appropriate topology, no strategies have clarified. If the characteristics of the Internet can be changed by modifying strategies and policies of indivisual ISP for connecting with the Internet, the strategy for more effective network formation can be derived. In this paper, we discuss strategies for obtaining the network which is more appropriate to achive communication efficiency and attack tolerance.

CiNii Article

Publishing type：Research paper (scientific journal)  

In the current Internet, there is a real need for the online real-time traffic identification technique to provide different services for real-time and bulk applications. Previously, it is easy to identify real-time traffic by checking the protocol/port number in IP header, however, it becomes more difficult due to the existence of real-time traffic over TCP connection, P2P and VPN. To treat these problems, we proposed the identification method. However, the method isn't resistant to changing network environment due to it uses static threshold for identification. In this paper, we propose a new identification method for real-time traffic based on characteristics of packet arrivals without checking the protocol/port number. Our approach is fast, accurate and adaptable to variety of environment compared to previous techniques. Through experiments by using the monitored data, we show that our proposed method cannot only identify real-time traffic accurately.

CiNii Article

Publishing type：Research paper (scientific journal)  

Packet sampling is one of effective ways to reduce the overhead of the traffic measurement in high speed routers, in which the router collects only a part of whole traffic based on the probabilistic theory. Since the statistic information of uncollected packets may be lost, the accuracy of the statistic information of sampled traffic becomes degraded as the increase of the sampling interval. In this paper, we analyze the difference information from statistics obtained by multiple sub-samplings with different sampling interval in order to model the change in the number of packets of flow by packet sampling. We then propose a new method which improves the accuracy of the estimation of flow length distribution based on derived model. We then apply our method to show that our method can estimate original flow length distributions accurately.

CiNii Article

Publishing type：Research paper (scientific journal)  

It is important for realistic performance evaluation of address lookup algorithms to consider both routing table and trace data obtained from target network. However, most of trace data of public domains are published after anonymizing personal information. Thus, the published trace data can't be applied directly to estimate performance of routers. Previously, performance of routers have been evaluated through simulations which use random address genereation, or small amount of trace data. In such way, performance evaluation may be different from actual performance of routers. In this papaer, we propose a methos for predicting more realistic router performance based on anonymized trace data.For our motivation, we analyze correlations of address space between real trace data and routing table.Based on analystic results, we also proposed a mapping method between anonymized trace data and routing table. Moreover, we show that our method can predict near real actual performance with a trace-driven simulator.

CiNii Article

Publishing type：Research paper (scientific journal)  

Coding and route diversity in networks are attractive for reliable communications in wireless sensor networks. In this paper, we target to demonstrate the fundamental characteristics of route diversity. The error probability is derived for a majority decision in tree networks composed of binary symmetric channel (BSC).

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

General orthogonal modulations produce new flexible modulations by changing modulation parameters which are rotation plane and rotation angle of multidimensional space. The modulation parameters should be transmitted to reciever at the change of modulations. In this paper, we propose a method that we estimate a modulation without side information on the modulation employed. We evaluate the proposed method by simulations in AWGN channel and show the availability of the proposed method.

CiNii Article

Publishing type：Research paper (scientific journal)  

With the rapid growth of broadband subscribers and data transmissions, it is predicted the congestion in the access network for the limited trunk bandwidth to the ISPs. In order to provide the QoS control for the applications, the session-based flow control on the edge router is needed. The ISP also faces a challenge of keeping operating costs. In this paper, by using the existing equipments of the ISP fully, we propose a new session-based packet classification algorithm with high granularity for the edge router. Our proposal uses the session information in the SIP proxy to build the session tables on the edge router. We also show the proposal is possible to function effectively by using the limited hardware resource of the router.

CiNii Article

Publishing type：Research paper (scientific journal)  

We have proposed a VNT (virtual network topology) reconfiguration method with traffic matrix estimation considerations. In this method, by dividing the whole VNT transition sequence into multiple transitions, estimation errors are calibrated at each stage by using network state information of prior stages. However, when some of traffics suddenly change, this method cannot accurately estimate the traffic matrix, because this method uses also network state information before the sudden change of traffics. Therefore, in this paper, we propose an estimation method which can estimate accurately the traffic matrix even in the case of sudden change of traffic. In our proposed method, we eliminate the information of the traffics changing suddenly from the network state information of prior stages. And then, by using the information after the elimination, we can estimate the traffic matrix accurately even in the case of sudden change of traffic. In this paper, we simulate our estimation method and verify that our estimation method can accurately estimate the traffic matrix even in the case of sudden change of traffic.

CiNii Article

Publishing type：Research paper (scientific journal)  

Packet sampling is one of effective ways to reduce the overhead of the traffic measurement in high speed routers, in which the router collects only a part of whole traffic based on the probabilistic theory. Since the statistic information of uncollected packets may be lost, however, the accuracy of the statistic information of sampled traffic becomes degraded as the increase of the sampling interval. In this paper. A new method which improves the accuracy of the estimation of flow length distribution by utilizing the difference information from statistics obtained by multiple sub-samplings with different sampling interval. Our method use difference information between flow statistics obtained at a different sampling period to estimate original flows. We then apply our method to show that our method can estimate original flow distributions with high accuracy.

CiNii Article

Publishing type：Research paper (scientific journal)  

Recent works have shown that the topology of the Internet and the link structure of WWW are scale-free networks. In scale-free networks, the degree distribution follows power law expressed by P (κ) ∝ κ^<-γ>, where the degree κ is number of edges. The network which exhibits power law is more advantageous than the random network in the connectivity between two ends and information retrieval, because the network distance is short. Moreover, scale-free networks show a high degree of error tolerance, because most of the nodes in network have small number of edges. However, these networks are extremely vulnerable to node attacks. In previous works, they have investigated AS-level topology of the Internet and influences of error and node attack. However, no strategies to reduce vulnerability to node attacks have shown and no evaluations whether the network structure such as the Internet is appropriate to communicate have shown. In this paper, we propose metrics to search for the appropriate network structure in aspects of communication efficiency and attack tolerance, and investigate what the network structure such as the Internet should be.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

It is thought that address info of network layer and port info of transport layer are used to discriminate IP communication sessions. Since this method multiplexes sessions at different two layers, it becomes inefficient and causes problems in security aspect. This paper describes a new communication architecture called "Unified Multiplex" that fix above problems comprehensively with epoch-making ideas. IP communication sessions are multiplexed only by IP address info of network layer in this architecture.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

In the current Internet, the online real-time traffic identification is a key technology to achieve differentiated service for real-time and bluk applications. Previously it is easy to identify real-time traffic by checking the protocol/port number in IP header, however, it becomes more difficult due to the existence of streaming traffic over TCP connection. Moreover, detection of real-time traffic over a kind of overlay network such as P2P or VPN is neither easy. In this paper, we propose a new identification method for real-time traffic based on not checking the protocol number, but analyzing the statistical characteristics of packet arrivals. Our approach is fast, accurate and lightweight compared to conventional techniques. Through experiments by using the monitored data, we show that our proposed method can not only identify real-time traffic accurately, but also find real-time over HTTP traffics which would not be detected by other methods.

CiNii Article

Publishing type：Research paper (scientific journal)  

トラヒックを効率的に収容する方法として、IP/光ネットワーク上で、光パスによって構築される論理トポロジ(VNT)を動的に再構成する手法の研究が進められている。しかしながら、VNTを適切に再構成するためには、対地間のトラヒッタ量を把握することが必須であるが、ネットワークの規模が大きくなるとともに、すべての対地間トラヒック量を測定することは困難となる。そのため、リンク負荷などの一部の測定情報から対地間トラヒック量を推定するトラヒックマトリクス推定手法の適用が望まれるが、トラヒックマトリクス推定を考慮に入れていない従来のVNT再構成手法では、推定誤差の影響を大きく受けてしまう。そこで、本稿では、トラヒックマトリクス推定を考慮に入れた新しいVNT再構成の手法を提案する。提案手法では、VNT再構成を複数ステージに分け、前のステージでの測定情報を推定に反映させることにより、推定誤差を削減しつつVNT再構成を行う。また、提案手法では、各ステージで追加・削除される光パスの本数に制約をもうけることにより、推定誤差の影響を受ける範囲を制限したVNT再構成を行う。本稿では、シミュレーションを用い、提案手法が誤差を削減し、トラヒックエンジニアリングへの誤差の影響を緩和できることを示す。

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

トラヒックを効率的に収容する方法として、パケットレイヤ網とオプティカルレイヤ網で協調してトラヒック制御をおこなうマルチレイヤトラヒックエンジニアリングの研究が進められている。トラヒック収容効率を高めるためには、対地間トラヒック量を測定することでネットワーク全体のトラヒック発生状況を把握し、その上で最適なトラヒック制御を行うことが必須となる。しかし、ネットワークの規模が大きくなるとともに、すべての対地間トラヒック量を測定することは困難となる。そのため、リンク負荷などの一部の測定情報から対地間トラヒック量を推定するトラヒックマトリクス推定手法の適用が望まれるが、トラヒックマトリクス推定により対地間トラヒック量に推定誤差が生じ、その結果、効率的なトラヒック収容が達成できない可能性がある。本稿では、推定手法により得たトラヒックマトリクスを用いてトラヒックエンジニアリングを行った際に、推定誤差がネットワーク性能にあたえる影響の評価を行う。評価の結果、マルチレイヤトラヒックエンジニアリングを適用した場合、シングルレイヤによるトラヒックエンジニアリングよりも良い性能を示すものの、推定誤差による性能低下が大きくなることが明らかとなった。そこで、推定誤差を減少させることを目的として、トラヒックの経路の変更に基づいたトラヒックマトリクスの再推定手法を提案した。そして提案手法を用いて各対地間トラヒツクの40%の精度を向上することによりネットワークに収容可能なトラヒック量が増加することがわかった。

CiNii Article

Publishing type：Research paper (scientific journal)  

In general, the modulation can be classified into the amplitude modulation(AM), the frequency modulation(FM), and the phase modulation(PM). The adaptive modulation system selects an appropriate modulation from these modulations against channel impairments. However, these modulations have a low degree of freedom or a small number of parameters to be optimized. The General Orthogonal Modulation(GOM) defines more flexible modulations [1],[2], with the modulation parameters of rotation plane and angle of a multi-dimentional space. In this paper, the error probability of GOM with Low-Density Parity-Check(LDPC) codes [3] is evaluated in the intersymbol interference(ISI) channel, and the modulation parameters are optimized in the channel.

CiNii Article

Publishing type：Research paper (scientific journal)  

In the adaptive modulation systems,a receiver identifies the modulation employed at the transmitter. In this paper, a new modulation classification method based on joint moments and then linear transform is proposed. In the method, the amplitude and phase of the received signal are transformed linearly, and joint moments are measured from the amplitude and phase. The moments depend on the linear transform, and the classification performance improves by some transforms. The joint moments are derived in an infinite series of elementary functions. Using the moments, the classification error probability for BPSK and QPSK is analyzed to discuss the effective linear transform.

CiNii Article

Publishing type：Research paper (scientific journal)  

In this paper, modulation parameters in multiple access systems are discussed to develop new modulations which can be optimized for user-dependent channel. The general orthogonal modulations are used as modulations to be optimized. The general orthogonal modulations are based on orthogonal vectors, which are obtained by multidimensional rotations and have parameters of rotation planes and rotation angles. Each user's transfer matrix of the proposed system is expressed by a Kronecker product of a transfer matrix and a code vector which maintain orthogonality among different users. We find parameters which maximize minimum squared distance for intersymbol interference (ISI).

CiNii Article

Publishing type：Research paper (scientific journal)  

Recently, streaming services to deliver video and voice are increasing rapidly. As for traffic of the Internet, two kinds (bulk/real-time) exist together in the same network now. Therefore, it is difficult to satisfy the communication quality of each traffic individually. One approach is expected that the edge router classifies the kind of traffic and provides different service for each type of traffic. Previously it is easy to identify real-time traffic by checking the protocol/port number in IP header, however, it becomes more difficult due to the existence of streaming traffic over TCP traffic connections. Moreover, detection of real-time traffic over a kind of overlay network such as P2P or VPN is neither easy. In this paper, we propose a new identification method for real-time traffic based on not checking the protocol number, but analyzing the statistical characteristics of packet arrivals. Through experiments by using the monitored data, our proposed method can effectively identify real-time traffics from all of TCP flows.

CiNii Article

Publishing type：Research paper (scientific journal)  

In the content distribution architecture not only improving the throughput and load balancing but also providing a stable performance independent from location of the end user is important. In this paper, we propose a new content placement method which offers a steady performance regardless of the physical location of the client. More specifically, we approximate physical network topology by using base pointers, and determine the servers which should be hold the content by taking the link's degree of sharing by edge-to-edge paths into account. Through simulation experiments on the dynamic parallel downloading, we show that our proposed method can decrease the number of content placements with achieving a stable performance regardless the location of the user.

CiNii Article

Publishing type：Research paper (scientific journal)  

Probabilistic Packet Marking (PPM) is one of the methods on IP traceback. In this method, intermediate router embeds (i.e., marks) the address of itself into the header of the sampled packet. However, because the PPM does not have a mechanism to avoid overrides of marking information, there is a problem that it needs a lot of packets to reconstruct attack route. We previously proposed two marking schmes based on the history of marking. They can reconstruct the attack route with fewer packets than the original PPM, however, they still require some parameter tunings for obtaining the optimal performance. In this paper we analyze factors which cause override of marking information, and propose a new packet marking scheme to minimize the impact of packet overrides. Through simulation results, we show that our proposed method can reduce the required number of packets significantly to reconstruct the attack route without tuning of parameters compared with the original and history-based PPMs.

CiNii Article

Publishing type：Research paper (scientific journal)  

An end-to-end round trip time (RTT) is one of the most important communication characteristics for Internet applications. From the viewpoint of network operators, RTT may also become one of the important variables to understand the network characteristics. For this case, we should know how a factor such as network incident influences RTTs. However, it is obvious that two or more factors may interfere in observed delay characteristics, because characteristics of packet transmission delays in the Internet are strongly dependent on the time-dependent condition of the network. In the previous work we proposed modeling method by using mixed distribution as a method of more accurately expressible to express delay characteristic more accurately in an environment where two or more impact factors exist together. However, no investigations have been shown to infer the network behavior based on modeling results. In this paper, we propose an estimation method of network behavior by decomposition of the mixed distribution. Furthermore, in experimental environment, we investigate the influence that cause by each network impact factor independently which affects the characteristics of the RTT distribution. By applying the proposed method, we also presume the events that occur in network from the measurements of RTTs and its result of decomposition.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

When a carrier frequency and a phase are not recovered with the sufficient accuracy in a receiver, a signal point rotates in a vector diagram. The signal point rotation causes a phase offset. In this paper, the optimal modulation classification with the phase offset is considered. Two new modulation classification methods based on joint moments of amplitude and phase are proposed with the application to the phase offset. One of the methods uses a difference of a measured moment, and another method uses a shift of signal points. The classification error probability is analyzed to show the superiority of proposed methods.

CiNii Article

Publishing type：Research paper (scientific journal)  

In this paper, modulation parameters in a multiple access system are discussed to develop new modulations which realize dual-optimization for two channel impairments. The general orthogonal modulations are used as modulations to be optimized. They are based on orthogonal vectors, which are obtained by multidimensional rotations and have parameters of rotation planes and rotation angles. The transfer matrix in a multiple access system is expressed by a kronecker product of two transfer matrices. We find parameters which maximize minimum squared distance for two channel impairments, rayleigh fading and intersymbol interference (ISI) by the dual-optimization approach.

CiNii Article

Publishing type：Research paper (scientific journal)  

An end-to-end round trip time (RTT) is one of the most important communication characteristics for Internet applications. In order to meet the quality of service (QoS) for applications such as streaming services, it is important to understand and predict the delay characteristic accurately. Because characteristics of the packet transmission delay in the Internet varies are strongly related to the time-dependent the condition of network. Especially, it is obvious that two or more factors may interfere in observed delay characteristics. In this paper, we express these delay characteristics by using the mixture of two or more probability distributions (especially normal distribution), not by single distribution. By investivation the factor which infects the delay distribution, we presume the events that occur in network from the delay characteristic obtained by the observation in the end host.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

When implementing optical satellite communication systems, restriction of transmission power is a major issue, and in order to get high reliability of the channel, using error correcting codes is indispensable. In earlier literatures much attention has been devoted to applying Reed-Solomon codes, BCH codes and convolutional codes, etc. to optical communication systems [1], [2]. In this paper, we focus on the characteristic of Low-Density Parity-Check (LDPC) codes which have the powerful error correcting capability, in optical communication systems. We apply Gaussian approximation of avalanche photodiode (APD) output modeling and three (3, 6)-regular LDPC codes whose codelength N=18 ; 996 ; 4998, respectively, and code rate is 0.5, into the direct detection optical communication system using On-Off-Keying modulation (OOK) and an APD detector, and investigate the relation between error probability performance and system parameters by Monte Carlo simulations.

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

An end-to-end round trip time (RTT) is one of the most important communication characteristics for Internet applications. In order to meet the quality of service (QoS) for applications such as streaming services, it is important to understand and predict the delay characteristic accurately. Because characteristics of the packet transmission delay in the Internet varies are strongly related to the time-dependent the condition of network. Especially, it is obvious that two or more factors may interfere in observed delay characteristics. In this paper, we express these delay characteristics by using the mixture of two or more probability distributions (especially normal distribution), not by single distribution. By investivation the factor which infects the delay distribution, we presume the events that occur in network from the delay characteristic obtained by the observation in the end host.

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

CiNii Article

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Publishing type：Research paper (scientific journal)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Symposium, workshop panel (nominated)  

Presentation type：Oral presentation (invited, special)  

Presentation type：Public lecture, seminar, tutorial, course, or other speech  

Presentation type：Symposium, workshop panel (nominated)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)  

Presentation type：Oral presentation (general)