Feature Articles: Movable and Deployable ICT Resource Unit— Architecture for Quickly Responding to Unexpected ICT Demand A Media Storage and Transmission System Using MDRUs Takayuki Nakachi, Sun Yong Kim, Atsushi Okada, and Tatsuya Fujii Abstract Developing information and communication technology (ICT) networks and services that are resilient to disasters is a critical task. NTT Network Innovation Laboratories is promoting innovative research and development on a transportable ICT node called the Movable and Deployable ICT Resource Unit (MDRU), which is designed to promptly recover ICT services after a disaster. In this article, we intro- duce a media storage and transmission system based on the use of MDRUs. It efficiently provides infor- mation such as text and image/video data to people in disaster areas in circumstances when ICT resourc- es are constrained such as when network bandwidth is limited or storage failures occur. We demon- strated its feasibility in a field trial. Keywords: layered coding, distributed storage, image/video transmission 1. Introduction greatest extent possible within the capacity restric- tions posed by MDRUs. The proposed system is The Great East Japan Earthquake of March 11, implemented by the following steps. 20 damaged or completely destroyed many infor- () Transport and activation of stand-alone MDRUs mation and communication technology (ICT) to disaster areas resources, which demonstrated that the development An MDRU with a cache server is quickly trans- of resilient ICT networks and services is a critical ported to a disaster area. The MDRU holds a cache issue demanding urgent attention. Immediately after server that runs software for the media storage and a disaster, it is important to rapidly share accurate transmission system. Layered storage consisting of information. In addition to text data, image/video random-access memory, solid-state drive (SSD)*, data are important in disaster recovery. Image/video and hard disk drive (HDD)*2 devices is installed in content can instantly provide detailed information to the cache server. The MDRU forms a wireless access people in disaster-stricken areas. Examples of infor- network, with or without wired network connections, mation that may need to be transmitted or exchanged to reach user terminals such as laptop personal com- after a disaster are shown in Fig. 1. puters (PCs), smartphones, and tablet devices. People A portable media storage and transmission system within 500 m or so from the MDRU can view infor- can provide such information to people working in mation such as text and image/video data stored in the disaster relief offices and to evacuees at evacuation cache server and/or upload data to the cache server. centers and in their own homes. Our media storage and transmission system is based on the specially * SSD: A data storage device that uses integrated circuit assemblies designed Movable and Deployable ICT Resource as memory to store data persistently. Compared with electrome- Unit, which we refer to as an MDRU. MDRUs will be chanical disks such as HDDs, SSDs are typically more resistant to transported immediately to a disaster area to provide physical shock and have lower access time and less latency. *2 HDD: A data storage device that uses rapidly rotating disks coated recovery of ICT services. The media storage and with magnetic material. HDDs remain the dominant medium due transmission system must transmit information to the to advantages in price per unit of storage and recording capacity. NTT Technical Review Feature Articles (2) Evacuation lists handwritten information Images/video can stored as images convey information instantly. (1) Photos and videos from individuals taken by smartphones, tablet computers, etc. MDRU Evacuation center MDRU coverage area (500-m radius) (3) Surveillance camera images and videos used for observing infrastructure such as arterial roads, airports, and Image/video sharing harbor facilities and viewing Fig. 1. Information needed in disaster areas. MDRU (equipped with cache server) MDRU Transmission based on bandwidth Distribution and sharing of images/video Upload Desktop computer (Viewing) MDRU MDRU MDRU MDRUs Smartphone (equipped with distributed storage) (image/video capturing, viewing) Fig. 2. Operation of multiple MDRUs in local disaster areas. (2) Operation of multiple MDRUs in local disaster area can watch news coverage originating outside the areas disaster area and share information about other As many MDRUs as needed are deployed to the areas. disaster area. Each MDRU has interworking storage The media storage and transmission system is devices that form a distributed storage system along based on two technologies. One is disaster resilient with the cache server. Data uploaded to each MDRU layered data transmission [], and the other is net- can be backed up to and shared among the multiple worked distributed storage [2]. The former transfers MDRUs (Fig. 2). information efficiently to work within the MDRU’s (3) Operation of multiple MDRUs connected to limited network bandwidth and power supply. The wide area network latter recovers data lost by ) network congestion or MDRUs will be connected to a local datacenter and disconnection or 2) storage failures caused by MDRU wide area network. Information data can be widely accidental power-off or MDRU transport. More shared via the local datacenter. People in the disaster details of these technologies are given below. Vol. 13 No. 5 May 2015 2 Feature Articles MDRU Layered storage Layered data (3) Transmission of layered code stream Low Large Low HDD g PL2 High resolution SSD PL1 Priority Mappin Memory Capacity Reliability PL0 High Small High Low resolution (2) Mapping layered code (1) Layered stream to layered storage image/video coding Fig. 3. Layered data transmission. 2. Disaster resilient layered data transmission (2) Mapping layered code-stream to layered stor- age The proposed disaster resilient layered data trans- The layered storage system consists of random- mission system offers image/video services to people access memory, SSD, and HDD devices. The JPEG in disaster-affected areas through MDRUs. The disas- 2000 bit streams are mapped to the layered storage as ter resilient image/video transmission scheme is shown in Fig. 3. PL0, PL1, and PL2 are priority-map- designed around three technologies: () layered ping layers and are mapped to random-access memo- image/video coding, (2) mapping of the layered code ry, SSD, and HDD, respectively. The progression stream to layered storage systems installed in the order of the JPEG 2000 bit stream is determined in MDRUs, and (3) transmission of layered code advance. Different scalability levels are achieved by streams according to network bandwidth (Fig. 3). ordering packets and can be arbitrarily assigned to () Layered image/video coding each form of storage. If MDRU power consumption We utilize the JPEG 2000 standard [3] created by becomes a problem, we can turn off the HDD devices, the Joint Photographic Experts Group (JPEG) for and if additional energy saving measures are neces- layered image/video coding. Input images are decom- sary, the SSD devices can be deactivated. posed into layered data in a priority manner; that is, (3) Transmission of layered code stream JPEG 2000 bit streams are created. JPEG 2000 offers JPEG 2000 is attractive as it can truncate lower four basic scalability dimensions in each JPEG 2000 priority bit streams from partial ones. This allows for bit stream: resolution (R), SNR (signal-to-noise ratio) simple responses to changes in network bandwidth quality (L), spatial location (P), and component (C). and MDRU power consumption. The truncation pro- Different scalability levels are achieved by ordering cess consists simply of extracting the required layers packets within the JPEG 2000 bit stream. We assume from the encoded bit stream; there is no additional that text data consist of handwritten notes on a bul- letin board set up in an evacuation center. The hand- written information is captured by smartphones and/ *3 H.264/MPEG-4 AVC (Advanced Video Coding): A video coding standard that was developed by the JVT (Joint Video Team) of the or tablet PCs. We use Motion JPEG 2000 (MJ2) for ITU-T (International Telecommunications Union-Telecommuni- the layered video coding. It specifies the use of the cation Standardization Sector) and the ISO/IEC (International JPEG 2000 format for timed sequences of images Organization for Standardization / International Electrotechnical (video). Other layered video codecs such as Scalable Commission) in 2003. It covers common video applications rang- ing from mobile services and video conferencing to IPTV (Inter- Video Coding (SVC), which is a scalable extension of net protocol television), high-definition (HD) TV, and HD video. H.264/MPEG-4 AVC*3, or Scalable High Efficiency *4 H.265/HEVC (High Efficiency Video Coding): The most recent Video Coding (SHVC), a scalable extension of standardized video compression standard developed by the JCT- H.265/HEVC*4, can also be used. We selected MJ2 VC (Joint Collaborative Team on Video Coding) of the ITU-T and the ISO/IEC. It reduces the bit rate by around 50% while maintain- because of its transmission robustness and low-com- ing the same subjective video quality relative to its predecessor plexity compared with SVC and SHVC. H.264/ MPEG-4 AVC. 3 NTT Technical Review Feature Articles processing of the stream itself. The MDRU forms a Layered storage wireless access network around itself to reach user- Layered LDGM held equipment. The truncation consists of extracting Layered data the required layers based on the bandwidth between Low Redundant data the MDRU and the user. An example of layered trans- PL2 mission is shown in Fig. 3 when the priority of the Low PL1 Mid resolution is selected. One usage example is that the Priority High user can see a specific full-resolution image after High PL0 overviewing multiple thumbnail-sized images. Send- ing only the higher resolution components of an image specified by a user results in more efficient use Distributed storage of network bandwidth resources.