Toshikazu Sakano

Fast optical channel recovery in field demonstration of 100-Gbit/s Ethernet over OTN using real-time DSP

A field trial of 100-Gbit/s Ethernet over an optical transport network (OTN) is conducted using a real-time digital coherent signal processor. Error free operation with the Q-margin of 3.2 dB is confirmed at a 100 Gbit/s Ethernet analyzer by concatenating a low-density parity-check code with a OTN framer forward error correction, after 80-ch WDM transmission through 6 spans x 70 km of dispersion shifted fiber without inline-dispersion compensation. Also, the recovery time of 12 msec is observed in an optical route switching experiment, which is achieved through fast chromatic dispersion estimation functionality.

Disaster-resilient networking: a new vision based on movable and deployable resource units

During the great east Japan earthquake on March 11, 2011, a lot of ICT resources - telecom switching offices, optical fiber links, and so forth - were completely or partially damaged due to the tremor and the resultant tsunami. As a consequence, the demand for ICT services explosively increased, mainly because the people of the affected areas were trying desperately to communicate with the outside world that led to a phenomenal rise in the network traffic. In the Nippon Telegraph and Telephone (NTT) East Corporation alone, 385 telephone offices stopped operating immediately following the earthquake because of power outages and disruption of facilities. Approximately 1.5 million users were cut off from using fixed-line telephone service. The demand for fixed-line and mobile telephone services jumped up to 10-50 times the usual. This gave rise to serious traffic congestion, and the emergency ICT networks and services could not deal with this issue sufficiently. This article proposes a network architecture that is resilient even through devastating disasters by effectively exploiting specially designed movable and deployable resource units, which we refer to as MDRUs. An MDRU having the ability to accommodate communication and information processing functions can be rapidly transported or moved to the disaster zone, and can be deployed within a reasonably short time to establish the network at the disaster site and launch ICT services. The concept and configuration of the network architecture based on the MDRU and its features are described in this article. Some preliminary simulation results are also reported to evaluate the performance of our adopted MDRU-based disaster resilient network.

Three-dimensional board-to-board free-space optical interconnects and their application to the prototype multiprocessor system: COSINE-III

A prototype multiprocessor system using three-dimensional board-to-board free-space optical interconnects is constructed for the first time to our knowledge. In the system, 64 processing units form a three-dimensional mesh processor network with the help of bidirectional board-to-board free-space optical interconnects. A theoretical analysis shows that the three-dimensional board-to-board freespace optical interconnects effectively solve common interconnection problems such as wiring congestion, signal delay, and clock skew. The prototype system, COSINE-III, is confirmed to work well as a multiprocessor system. The system is also shown to be easy to extend to a larger and more flexible system.

Large-dispersion-tolerance optical signal transmission system based on temporal imaging

A novel optical signal transmission system, which is highly tolerant of dispersion of the transmission fiber, is proposed. The system employs a dispersion fiber and a phase modulator in both the transmitter and the receiver. We analyzed the characteristics of the system, using the temporal imaging concept, and found that the output optical pulse is insensitive to dispersion of the transmission fiber if the parameters of the system are set to hold a specific condition. We report simulation results that confirm these characteristics of the proposed system.

A Spectrum- and Energy-Efficient Scheme for Improving the Utilization of MDRU-Based Disaster Resilient Networks

The movable and deployable resource unit (MDRU)-based network provides communication services in disaster-struck areas where the lack of spectrum and energy resources is intensified due to the high demand from users and the power outages after a disaster. The MDRU-based network attempts to apply spectrum- and energy-efficient methods to provide communications services to users. However, existing works in this field only consider spectrum efficiency or energy efficiency separately, in spite of the tradeoff relationship between them. Thus, we propose a scheme to improve the utilization of both spectrum and energy resources for better system performance. The considered MDRU-based network is composed of gateways deployed in the disaster area, which can replenish their energy by using solar panels. Our proposed scheme constructs a topology based on the top k spectrum-efficient paths from each sender and applies a max flow algorithm with vertex capacities, which are the number of transmissions that each gateway can send, which is referred to as transmission capability. The transmission capability of each gateway is determined by its energy resource and distances to its neighbors. Furthermore, we show that the proposal can be used for multisender-multireceiver topologies. A new metric named spectrum-energy efficiency to measure both spectrum efficiency and energy efficiency of the network is defined. Through analyses, we prove that a value of k exists such that the spectrum-energy efficiency of a given topology is maximized. Furthermore, our simulation results show that, by dynamically selecting appropriate value of k, the proposed scheme can provide better spectrum-energy efficiency than existing approaches. Moreover, our experimental results verify the findings of our analysis.

Bringing movable and deployable networks to disaster areas: development and field test of MDRU

Communication demand is paramount for disaster-affected people to confirm safety, seek help, and gather evacuation information. However, the communication infrastructure is likely to be crippled due to a natural disaster, which makes disaster response excruciatingly difficult. Although traditional approaches can partially fulfill the most important requirements from the user perspective, including prompt deployment, high capacity, large coverage, useful disaster-time application, and carrier-free usability, a complete solution that provides all those features is still required. Our collaborative research and development group has developed the Movable and Deployable Resource Unit, which is referred to as the MDRU and has been proven to have all those required features. Via extensive field tests using a compact version of an MDRU (i.e., the van-type MDRU), we verify the effectiveness of the MDRU-based disaster recovery network. Moreover, we demonstrate the further improvement of the MDRUs performance when it is complemented by other technologies such as relay-by-smartphone or satellites.

Throughput and Delay Tradeoffs for Mobile Ad Hoc Networks With Reference Point Group Mobility

In this paper, we explore the throughput-delay tradeoff in a mobile ad hoc network (MANET) operating under the practical reference point group mobility model and also a general setting of node moving speed. In particular, we consider a MANET with unit area and n nodes being divided evenly into Θ(n^α) groups, α ∈ [0, 1], where the center of each group moves according to a random direction model with speed of no more than υ ∈ [0, 1]. We determine the regions of per-node throughput and average delay and their tradeoffs that can be achieved (in order sense) in such a network. For the regime of v = 0, we first prove that the per-node throughput capacity is Θ(n^-α/2) and then develop a routing scheme to achieve this capacity, resulting in an average delay of Θ(max{n^1/2, n^1-α}) for any α ∈ [0, 1]. Regarding the regime of v > 0, we prove that the per-node throughput capacity can be improved to Θ(1), which is achievable by adopting a new routing scheme with an average delay of Θ(max{n^1-α, n^α/2/v}) for υ = o(1) and Θ(n) for v = Θ(1). The results in this paper help us to have a deep understanding on the fundamental performance scaling laws and also enable an efficient throughput-delay tradeoff to be achieved in MANETs with correlated mobility.

A Performance Evaluation of Multiple MDRUs Based Wireless Mesh Networks

Since communications services become much more demanded after disaster strikes, it is necessary to promptly set up a temporary communications infrastructure to provide services to those in need. The Movable and Deployable Resource Unit (MDRU) based Wireless Mesh Network (WMN) is an attractive candidate to achieving this goal. In MDRU based WMN, the MDRU is transported to the disaster affected area by either ground or air transportation like truck or helicopter. After arriving at the disaster area, it configures any remaining wireless Access Points (AP) in the area to provide connectivity services. This work provides an insight on the performance of MDRU based WMN under the situation where multiple MDRUs are deployed within close region to increase the overall coverage and performance of the network. A simulation is conducted to estimate the performance of the network under both scenarios where mesh tier operates under a single channel and where a channel assignment scheme is applied. We show that the performance of MDRU based WMN can be greatly enhanced by deploying multiple MDRUs to the area. However, since the number of available MDRUs is limited, using more MDRUs than necessary is not efficient. We discuss some of the factors that should be taken into account when selecting an appropriate number of MDRUs for a single area.

Inter-hospital PACS designed for teleradiology and teleconference using a secured high-speed network

We developed a tele-radiology and tele-conference system between our related hospitals. This system consisted of the image database, the WWW server, WWW browsers, high resolution CRT displays, the videoconference system and an asynchronous transfer mode (ATM) network. In advance X-ray images were stored into the Image Save And Carry magneto- optical (MO) disks, then images on the MO were transferred to the image database. The image database was created from MO disks. Total amount of images reached 100 GB and the number of the image was 65,000. The ATM network connected the hospitals each other. The ATM network device provided the permanent virtual circuit function. The transmission speed was from 6Mbit/second to 155 Mbit/second. The client station consisted of the WWW browser and the super high definition CRT display which had the 2k X 2k full color frame memory and 54 X 54 cm square display area. The result of the query was transformed to a hypertext markup language. Then a browser on a client machine displayed the result. The server could retrieve some images in about ten seconds and transmit an image from a server to a client in 2-10 seconds that depend on the network speed. At the tele- radiology, both terminals could display same image and physicians could talk each other by the videoconference system. We solved the security problems by the PVC methods and the on time password device. The ATM network showed the high transmission performance and good security. Physicians were able to use this system with no special training and this system brought us an effective utilization of the image.

A Failure-Tolerant and Spectrum-Efficient Wireless Data Center Network Design for Improving Performance of Big Data Mining

Wireless Data Center Network (Wi-DCN) is considered one of the most promising future data center architectures due to its low installation and management cost and high flexibility of network design. However, the existing Wi-DCN is, still, not capable of providing an efficient big data mining service such as MapReduce because its topology (i.e., Cayley graph with same degree) cannot achieve enough connectivity on the breakdown of servers and spectrum efficiency, which are important factors to improve the performance of big data mining. Therefore, in order to modify the existing Wi-DCN for big data mining, this paper proposes a spherical rack architecture based on a bimodal degree distribution that improves both failure tolerance and spectrum efficiency. Extensive computer simulations demonstrate the effectiveness of our proposed rack architecture in terms of data transmission time required for MapReduce under a failure-prone environment.