• Journal of the Korea Society of Computer and Information
• /
• v.23 no.10
• /
• pp.1-9
• /
• 2018

Modern graphics processing units (GPUs) have become one of the most attractive platforms in exploiting high thread level parallelism with the support of new programming tools such as CUDA and OpenCL. Recent GPUs has applied cache hierarchy to support irregular memory access patterns; however, L1 data cache (L1D) exhibits poor efficiency in the GPU. This paper shows that the L1D does not always positively affect the applications in terms of performance and energy efficiency for the GPU. The performance of the GPU is even harmed by using the L1D for lots of applications. Our proposed technique exploits the characteristics of the currently-executed applications to predict the performance impact of the L1D on the GPU and then decides whether to continuously use the cache for the application or not. Our experimental results show that the proposed technique improves the GPU performance by 9.4% and saves up to 52.1% of the power consumption in the L1D.

• Korean Journal of Optics and Photonics
• /
• v.24 no.6
• /
• pp.297-310
• /
• 2013

With the explosive growth of internet data traffic due to the FTTH penetration, prevalence of smart devices and cloud network service, the demand for higher bandwidth is ever increasing with the pace of more than 40% annual growth. To accommodate ultra high bandwidth traffic, optical components in each hierarchy have progressed rapidly. WDM has begun to be deployed along with higher bandwidth service in the access network. Next-generation ROADM is under development for efficient network management in the metro network. For long-haul transmission, an advanced modulation scheme based on coherent transmission technology has been adopted to enhance spectral efficiency. In this paper, core components to meet the demands of high speed, high efficiency and low power consumption will be reviewed.

• Journal of Information Processing Systems
• /
• v.14 no.3
• /
• pp.771-789
• /
• 2018

In heterogeneous wireless networks supporting multi-access services, selecting the best network from among the possible heterogeneous connections and providing seamless service during handover for a higher Quality of Services (QoSs) is a big challenge. Thus, we need an intelligent vertical handover (VHO) decision using suitable network parameters. In the conventional VHOs, various network parameters (i.e., signal strength, bandwidth, dropping probability, monetary cost of service, and power consumption) have been used to measure network status and select the preferred network. Because of various parameter features defined in each wireless/mobile network, the parameter conversion between different networks is required for a handover decision. Therefore, the handover process is highly complex and the selection of parameters is always an issue. In this paper, we present how to maximize network utilization as more than one target network exists during VHO. Also, we show how network parameters can be imbedded into IEEE 802.21-based signaling procedures to provide seamless connectivity during a handover. The network simulation showed that QoS-effective target network selection could be achieved by choosing the suitable parameters from Layers 1 and 2 in each candidate network.

• Journal of KIISE:Computer Systems and Theory
• /
• v.32 no.9
• /
• pp.483-490
• /
• 2005

Recently, flash memory is widely used in embedded applications since it has strong points: non-volatility, fast access speed, shock resistance, and low power consumption. However, due to its hardware characteristics, it requires a software layer called FTL(flash translation layer). The main functionality of FTL is to convert logical addresses from the host to physical addresses of flash memory We present a new FTL algorithm called STAFF(State Transition Applied Fast Flash Translation Layer). Compared to the previous FTL algorithms, STAFF shows five times higher performance than basic block mapping scheme and requires less memory. We provide performance results based on our implementation of STAFF and previous FTL algorithms.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.6
• /
• pp.2492-2512
• /
• 2018

Multicast communication can effectively reduce network resources consumption in contrast with unicast. With the advent of SDN, current researches on multicast traffic are mainly conducted in the SDN scenario, thus to mitigate the problems of IP multicast such as the unavoidable difficulty in traffic engineering and high security risk. However, migration to SDN cannot be achieved in one step, hybrid SDN emerges as a transitional networking form for ISP network. In hybrid SDN, for acquiring similar TE and security performance as in SDN multicast, we redirect every multicast traffic to an appropriate SDN node before reaching the destinations of the multicast group, thus to build up a core-based multicast tree substantially which is first introduced in CBT. Based on the core SDN node, it is possible to realize dynamic control over the routing paths to benefit traffic engineering (TE), while multicast traffic manageability can also be obtained, e.g., access control and middlebox-supported network services. On top of that, multiple core-based multicast trees are constructed for each multicast group by fully taking advantage of the routing flexibility of SDN nodes, in order to further enhance the TE performance. The multicast routing and splitting (MRS) algorithm is proposed whereby we jointly and efficiently determine an appropriate core SDN node for each group, as well as optimizing the traffic splitting fractions for the corresponding multiple core-based trees to minimize the maximum link utilization. We conduct simulations with different SDN deployment rate in real network topologies. The results indicate that, when 40% of the SDN switches are deployed in HSDN as well as calculating 2 trees for each group, HSDN multicast adopting MRS algorithm can obtain a comparable TE performance to SDN multicast.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.12
• /
• pp.4997-5006
• /
• 2010

A flash-memory-based SSDs(Solid State Disks) are one of the best media to support portable and desktop computers' storage devices. Their features include non-volatility, low power consumption, and fast access time for read operations, which are sufficient to present flash memories as major database storage components for desktop and server computers. However, we need to improve traditional storage management schemes based on HDD(Hard Disk Drive) and RAID(Redundant array of independent disks) due to the relatively slow or freezing characteristics of write operations of SSDs, as compared to fast read operations. In order to achieve this goal, we propose a new storage management scheme called Hetero-Mirroring based on traditional HDD mirroring scheme. Hetero-Mirroring-based scheme improves RAID-1 operation performance by balancing write-workloads and delaying write operations to avoid SSD freezing. Our test results show that our scheme significantly reduces the write operation overheads and freezing overheads, and improves the performance of traditional SSD-RAID-1 scheme by 18 percent, and the response time of the scheme by 38 percent.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.11B
• /
• pp.1245-1253
• /
• 2009

Since 2000, the demand for multimedia content consumption has been continuously increasing. The research and standardization for the mobile broadcast service has been done to deliver multimedia contents to the huge number of users with the minimum cost. A few commercial services started after the related standards were released. We call this mobile broadcast service the 1st generation service. Now, many experts are researching and developing the technologies for 3Screens(TV, PC, and Mobile) Services that an user can enjoy multimedia contents at every place with the various networks and devices. To enable this service, two main technical areas exist. The first is a core network technology that can support the various access network and the second is a common application layer technology. To resolve these technical areas, we propose IMS(IP Multimedia Subsystem) based OMA BCAST(Open Mobile Alliance Mobile Broadcast Service Enabler Suite) as the technical solution for 3 Screen Services in the paper. In addition, we propose application layer multicast that is enable to implement IMS based OMA BCAST to the existing network with the minimum cost. Finally, we show how we can save the implementation cost of IMS based OMA BCAST with the simulation results.

• Journal of Internet Computing and Services
• /
• v.8 no.5
• /
• pp.133-140
• /
• 2007

Currently, as the requirement for high quality Internet access from anywhere at anytime is consistently increasing, the interconnection of pure ad hoc networks to fixed IP networks becomes increasingly important. Such integrated network, referred to as hybrid ad hoc networks, can be extended to many applications, including Sensor Networks, Home Networks, Telematics, and so on. We focus on some data communication problems of hybrid ad hoc networks, such as broadcasting and routing. In particular. power failure of mobile terminals is the most important factor since it affects the overall network lifetime. We propose an energy-efficient routing protocol based on clustering for hybrid ad hoc networks. By applying the index-based data broadcasting and selective tuning methods, the infra system performs the major operations related to clustering and routing on behalf of ad hoc nodes. The proposed scheme reduces power consumption as well as the cost of path discovery and maintenance, and the delay required to configure the route.

• The Journal of Korean Association of Computer Education
• /
• v.13 no.1
• /
• pp.53-64
• /
• 2010

In this paper, we tried to enable a mobile device as a resource to access to mobile grid networks. By advanced Internet techniques, the use of mobile devices has been rapidly increased. Some researches in mobile grid computing tried to combine grid computing with mobile devices. However, according to intrinsic properties of mobile environments, mobile devices have many considerations, such as mobility management, disconnected operation, device heterogeneity, service discovery, resource sharing, security, and so on. To solve these problems, there are two trends for mobile grid computing: a proxy-based mobile grid architecture and an agent-based mobile grid architecture. We focus on a proxy-based mobile grid architecture with IP-paging, which can easily manage idle mobile devices and grid resource status information. Also, we use SIP(Session Initiation Protocol)to support mobility management, mobile grid services. We manage variation of mobile device state and power by paging cache. Finally, using the candidate set and the reservation set of resources, we perform task migration. The performance evaluation by simulation, shows improvement of efficiency and stability during execution.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.1
• /
• pp.166-174
• /
• 2010

SSDs are one of the best media to support portable and desktop computers' storage devices. Their features include non-volatility, low power consumption, and fast access time for read operations, which are sufficient to present flash memories as major database storage components for desktop and server computers. However, we need to improve traditional index management schemes based on B-Tree due to the relatively slow characteristics of flash memory operations, as compared to RAM memory. In order to achieve this goal, we propose a new index management scheme based on a compressed hot-cold clustering called CHC-Tree. CHC-Tree-based index management improves index operation performance by dividing index nodes into hot or cold segments and compressing pointers and keys in the index nodes and clustering the hot or cold segments. The offset compression techniques using unused free area in cold index node lead to reduce the number of slow erase operations in index node insert/delete processes. Simulation results show that our scheme significantly reduces the write and erase operation overheads, improving the index search performance of B-Tree by up to 26 percent, and the index update performance by up to 23 percent.