Published papers

Large parallel applications become sensitive to communication latencies, suggesting the need for low-latency networks in high-performance computer systems. Switch delay dominates network latencies, especially for a large number of small transfer data. To reduce the network latencies, we exploit routing cache on a switch. Routing decision based on off-chip CAM(Content Addressable Memory)-based table lookup imposes a significant delay, however, using on-chip small routing cache can bypass it when it hits. Our simulation results showed that 1,024-entry routing cache improves not only up to 13% of packet latency but also up to 18% of network throughput compared with no-cache switches. © 2012 IEICE Institute of Electronics Informati.

© 2016 IEEE. The internet of things (IOT) continues to transform the world, and in many countries is now an integral part of our everyday lives-influencing everything from the way that we intercommunicate to how we conduct business. Innovators continue to find ways to integrate IOT into uses as far flung as fashion to medicine. This survey looks at how IOT is currently being integrated into the built environment for the purpose of saving energy and improving occupants' livelihoods. In particular, it reviews three technologies that have received a lot of attention in the literature as the future of an IOT mediated built environment. Based on this literature, predictions are made of the likely trends in regards to the future scholarship on these technologies.

The bi-directionality of electric vehicle (EV) charging has proved to be important for load leveling purposes, amongst others. Thus, Vehicle-to-Grid (V2G) communication protocols are a necessity which ensures that the information exchange that takes place during this process, is efficient, secure, and reliable. Several research articles regarding different V2G communication protocols have been published over the years. This paper summarizes current relevant research in this area in order to provide a reference guide for researchers interested in this subject.

There are many challenges for Internet of Things (IoT) sensor networks including the lack of robust standards, diverse wireline and wireless connectivity, interoperability, security, and privacy. Addressing these challenges, the Institute of Electrical and Electronics Engineers (IEEE) P1451.0 standard defines network services, transducer services, transducer electronic data sheets (TEDS) format, and a security framework to achieve sensor data security and interoperability for IoT applications. This paper proposes a security solution for IEEE P1451.1.6-based sensor networks for IoT applications utilizing the security framework defined in IEEE P1451.0. The proposed solution includes an architecture, a security policy with six security levels, security standards, and security TEDS. Further, this paper introduces a new service to update access control lists (ACLs) to regulate the access for topic names by the applications and provides an implementation of the security TEDS for IEEE P1451.1.6-based sensor networks. The paper also illustrates how to access security TEDS that contain metadata on security standards to achieve sensor data security and interoperability.

We developed the RHiNET-2/SW high-speed network switch for a high-throughput parallel computing system consisting of optically interconnected distributed personal computers (PCs). RHiNET-2/SW enables high-throughput network switching (64 Gbit/s) by means of a switching LSI and eight pairs of synchronized parallel optical interconnection modules. To confirm the reliability of RHiNET-2/SW, we measured its bit error rate (BER), which was less than 10-11 (estimated value: 10-20). Five RHiNET-2/SWs and sixteen PC nodes can perform parallel computing. This result demonstrates the feasibility of optical interconnection in a high-speed, flexible network in a parallel computing system.

© 2016 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc. Internet communication message protocol (ICMP)-based traceability methods are widely used to trace packets over the Internet; however, in their attacks, adversaries likewise use ICMP packets. Furthermore, the lack of security in ICMP-based traceability results in failures for many current traceability methods. Moreover, current routers are unable to provide extended services to Internet users and applications. To address this need, our laboratory has introduced the service-oriented router (SoR). SoR is middleware that can be implemented on a Cisco AXP and Juniper JunosV App Engine. In this paper, we propose a secure method of providing packet traceability over public networks using SoR features. We implemented a secure packet traceability service prototype on the ns-3 simulator. The test results conclude that there is a maximum additional cost of 48.69 and 123.91 μs of processing overhead per packet in each hop when the proposed secured traceability service is used with a 128-bit key space in AES-GCM and AES-CTR modes, respectively. Moreover, for 256-bit key spaces, AES-GCM and AES-CTR modes consumed additional 47.18 and 123.25 μs, respectively, over the plain traceability. Yet, from the evaluations on the test bed topology, it clearly shown that the proposed method was 79% faster than the conventional trace route method in providing the secured end-to-end traceability. © 2016 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

We propose a technique to reduce compulsory misses of packet processing cache (PPC), which largely affects both throughput and energy of core routers. Rather than prefetching data, our technique called response prediction cache (RPC) speculatively stores predicted data in PPC without additional access to the low-throughput and power-consuming memory (i.e., TCAM). RPC predicts the data related to a response flow at the arrival of the corresponding request flow, based on the request-response model of internet communications. Our experimental results with 11 real-network traces show that RPC can reduce the PPC miss rate by 13.4% in upstream and 47.6% in downstream on average when we suppose three-layer PPC. Moreover, we extend RPC to adaptive RPC (A-RPC) that selects the use of RPC in each direction within a core router for further improvement in PPC misses. Finally, we show that A-RPC can achieve 1.38x table-lookup throughput with 74% energy consumption per packet, when compared to conventional PPC.

© 2014 IEEE. It is a well-known fact that the Internet traffic travels through public networks and these networks are vulnerable and congested. Whenever an error in a packet is detected, the packet should be retransmitted back to the receiver to reduce/overcome the data loss. Automatic repeat request (ARQ) method is a well-known data retransmission method used by protocols such as Transmission Control Protocol (TCP). This method will retransmit data all the way from the sender resulting higher end-to-end retransmission delays in the packet transmission. Nevertheless this method is widely used and operational, it will also introduce unnecessarily duplicate packets further congesting the networks when it retransmit the entire packet stream from an unacknowledged packet after detecting a retransmission requirement. Encryption and public key cryptography are important technologies that are used to preserve data security and integrity, and to reduce information theft on the public networks. These existing technologies have less flexibility over the current security requirements and the conventional routers are unable to provide end services to the users and applications. To address this need, our laboratory introduced the per-hop data encryption protocol for service-oriented router (SoR) based public networks which allows more secured and flexible method to transfer data over public networks [15]. SoRs are introduced to maintain rich information for the next-generation networks by shifting the current Internet infrastructure to an information-based and an open-innovation platform [6]. SoR is a middleware and can be implemented on a Cisco AXP and Juniper JunosV App Engine where it can analyse all packet stream transactions on its interfaces. Using the features of the SoR, in this paper, we propose a previous hop data retransmission service for public networks. This proposing service can provide higher data availability and reliability to the data that traverses through public networks, reducing the end-to-end data retransmission delays. We implemented a prototype of previous hop data retransmission service on the ns-3 simulator. Furthermore, from the evaluations on the test bed topology, it clearly showed that the proposed method was 80.43% faster than the conventional ARQ based retransmission method.

© 2015 IEEE. Homomorphic encryption for smart grids has been investigated in many studies. It is possible to estimate the total power consumption in an area without knowing the consumption data of individual households. In the case of demand response (DR), it is important to calculate the total electric power consumption in an area because DR reports are published accordingly to reduce peak power consumption when the demand is high. However, the published data may reveal private information about residents, such as the timings of specific activities (leaving from and returning home), and device details. To overcome this problem, we propose a method specialized to enable energy providers to securely share electric power consumption data. The proposed method uses a self-organizing map (SOM), which is an unsupervised learning method. In order to share power consumption data while preserving privacy, the SOM is shared without the raw data. In this framework, a target accuracy of nearly 3% is achieved, while actual data are not published by any company.

© 2015 IEEE. Heating, ventilation, and air conditioning (HVAC) control is used to effectively reduce energy consumption in a home energy management system (HEMS). For the HVAC control, it is necessary to evaluate the indoor comfort level because the HVAC operation is significantly related to indoor environmental conditions. Although there are some indices for indoor comfort, individuals have different comfort ranges. A questionnaire is one way to gather data on individual comfort ranges. However, it requires contributions from users. Therefore, we proposed a new survey method that varied the intervals of the questions addressed to users to reduce the number of responses needed. Additionally, a comfort estimation method was proposed that did not require a questionnaire. This method analyzes the user's operation of the HVAC system, such as touching its on/off switch. An HEMS with the capability of utilizing these two methods was installed in a laboratory and 16 houses in Miyagi prefecture, Japan. An experiment using this HEMS was conducted, and it was able to reduce the number of feedbacks needed to maintain each person's comfort by an average of 32.2%. The HEMS provided the same accuracy as hourly questionnaires. Namely, it confirmed that the estimation based on the air conditioner operations was almost the same as that based on an hourly questionnaire.

Federated Learning (FL) is a distributed machine learning paradigm that enables multiple data holders to collaborate on building machine learning models while preserving the privacy of their data. FL can be categorized as horizontal or vertical, depending on the distribution characteristics of the data. Specifically, horizontal FL uses data partitioned in the sample space, whereas vertical FL uses data partitioned in the feature space. Traditional vertical FL methods aim to facilitate collaboration among clients to infer a single global target. However, these methods may be impractical because each client often has a unique target to be inferred. In this paper, we propose a novel vertical FL method, called personalized vertical federated learning (pvFed), which addresses this limitation by allowing each client to perform inferences specific to their individual task. To the best of our knowledge, no existing method currently addresses this limitation. The objective of pvFed is to construct a global model that generates a representation vector to support client inference. The global model, constructed using distillation and dimensionality reduction, takes a sample ID common to all clients as input and outputs a sample-specific representation vector. Clients utilize the intermediate representation of their own model and the representation vectors output by the global model for inference. Because these vectors are not dependent on client-specific tasks, clients can repurpose them for any additional tasks. Our experiments, conducted on two distinct data types—image and tabular data sets, under a vertical partitioning where each client had its own specific task, demonstrated the efficacy of vectors generated by the global model in pvFed for client inference. © 2024 Institute of Electrical Engineer of Japan and Wiley Periodicals LLC.

In recent years, instead of wired communication, wireless communication technology has been utilized in many applications. Particularly, sensor networks are used in cases such as environmental monitoring, security service, space exploration etc. In this paper, motion planning and keeping connectivity of multiple robots distributed in a ZigBee based sensor networks are discussed in order to sense their unknown surroundings. The multiple robots distributed in a ZigBee based sensor network exchange information each other over the wireless sensor network and accomplish cooperative tasks in a closed area. In a rescue task, it changes dynamically according to the continuously shifting situation. In this case, the robots need to keep connectivity between other robots in order to accomplish the tasks in the real world. The proposed method enables mobile robot to be controlled under non-holonomic constraint and it is controlled based on the indicator of communication connectivity, which is related to the quality of communication as a real-time and real world application.

© 2014 Technical University of Munich (TUM). Many network services including intrusion detection and recommendation provide their services by analyzing information acquired from network transactions. A careful analysis of these data can reveal valuable information when deep packet inspection is performed. Since these packet analyses generate sensitive information from enormous volumes of transmitted data, the requirement for data anonymization has been discussed. There have been many studies of anonymization techniques and their implementation in software applications. However, limited research has been undertaken regarding hardware-based anonymizers. This paper proposes and evaluates a RAM-based anonymization architecture that maintains both high throughput and a low information-loss ratio.

In the regulation of heating, ventilation, and air conditioning (HVAC) for a peak cut or peak leveling control of electric power consumption, the amenity of the living environment should be considered. In the existing method, the metabolic equivalent (MET) value is not considered because of the substantial measurement costs and because a special infrastructure is required for both measuring the value by the MET meter and for communicating with the HVAC control server in the home energy management system (HEMS) or building EMS (BEMS). This paper proposes a method for measuring the MET value on a mobile device for accurately calculating the predicted mean vote (PMV). This method is designed as an iPod touch or iPhone application. The application has sufficiently reduced energy consumption for use as a mobile device application and is sufficiently accurate for use as an HVAC control. The availability and usefulness of both the designed application and the proposed MET measuring method is experimentally confirmed, and the reduction in energy consumption is estimated using an actual HEMS. © 2013 IEEE.

The increase in the amount of digital data transferred between contents Servers and Clients in a network causes problems created by congestion when downloading big web files, streaming media, etc. Content Delivery Networks (CDNs) are introduced to overcome such problems by redirecting the Clients request to the nearest Server. However, as the number of people accessing Internet increases rapidly, the performance of CDN is reduced, resulting in network congestion at times. To solve these problems, we propose an architecture in which a new management method is employed to redirect the Clients request using a content-based Router called the Service-oriented Router (SoR). This method is more effective than the currently implemented, traditional DNS-based request redirection (RR). The SoR is a special router that can perform a deep packet inspection of the packet streams. Our experiments were based on analytical modeling as well as test bed experiments, and they confirmed that the response time in the Router-based RR was less than that in the DNS-based RR by 23.3%. In addition, our test bed experiments showed that the RTT in the Router-based RR is less than that in the DNS-based RR by 7.7%. © 2013 IEEE.

Preserving the privacy of individuals while publishing their relevant data has been an important issue. K-anonymization, which is one of the most useful privacy models, is valid for static data; however, it does not support dynamic data updating. In this paper, we propose a method for inserting and deleting data into and from k-anonymous tabular data while maintaining k-anonymity. This method ensures k-anonymity after multiple publications by retaining the raw data and the latest published data. In the deletion process, the specified record is first deleted from the data table, and the records remaining in the q∗-block to which the specified record belonged are inserted into the data table excluding the remaining records. This insertion of multiple records is performed by repeating the above addition process in sequence.

IoT applications require secure communication methods that protect personal information contained in communication data. This study focuses on MQTT, a low-cost protocol used for IoT communication, and proposes a mechanism to anonymize communication data between IoT and clients. MQTT is a publish-subscribe model of communication where a broker handles many-to-many communications among clients. Due to the concentration of communications on the broker, it is efficient to anonymize data there. Therefore, the proposed mechanism performs differential privacy anonymization of communication data on the MQTT broker. We also propose a mechanism to anonymize data according to anonymization criteria required by senders and receivers using topic names and user properties, which are features of MQTT. We implemented the proposed mechanism in an FPGA-based MQTT broker and confirmed that it achieves the same throughput and low latency as regular MQTT communication and satisfies IoT applications such as power control and automated driving that require sub-millisecond latency.