Mariko SAKAMOTO Akira KATSUNO Go SUGIZAKI Toshio YOSHIDA Aiichiro INOUE Koji INOUE Kazuaki MURAKAMI
Broadcast and synchronization techniques are used for cache coherence control in conventional larger scale snoop-based SMP systems. The penalty for synchronization is directly proportional to system size. Meanwhile, advances in LSI technology now enable placing a memory controller on a CPU die. The latency to access directly linked memory is drastically reduced by an on-die controller. Developing an enterprise server system with these CPUs allows us an opportunity to achieve higher performance. Though the penalty of synchronization is counted whenever a cache miss occurs, it is necessary to improve the coherence method to receive the full benefit of this effect. In this paper, we demonstrate a coherence directory organization that fits into DSM enterprise server systems. Originally, a directory-based method was adopted in high performance computing systems because of its huge scalability in comparison with snoop-based method. Though directory capacity miss and long directory access latency are the major problems of this method, the relaxed scalability requirement of enterprise servers is advantageous to us to solve these problems along with an advanced LSI technology. Our proposed directory solves both problems by implementing a full bit vector level map of the coherence directory on an LSI chip. Our experimental results validate that a system controlled by our proposed directory can surpass a snoop-based system in performance even without applying data localization optimization to an online transaction processing (OLTP) workload.
Junichi FUNASAKA Atsushi KAWANO Kenji ISHIDA
Parallel downloading retrieves different pieces of a file from different servers simultaneously and so is expected to greatly shorten file fetch times. A key requirement is that the different servers must hold the same file. We have already proposed a proxy system that can ensure file freshness and concordance. In this paper, we combine parallel downloading with the proxy server technology in order to download a file quickly and ensure that it is the latest version. Our previous paper on parallel downloading took neither the downloading order of file fragments nor the buffer space requirements into account; this paper corrects those omissions. In order to provide the user with the required file in correct order as a byte stream, the proxy server must reorder the pieces fetched from multiple servers and shuffle in the delayed blocks as soon as possible. Thus, "substitution download" is newly introduced, which requests delayed blocks from other servers to complete downloading earlier. Experiments on substitution download across the Internet clarify the tradeoff between the buffering time and the redundant traffic generated by duplicate requests to multiple servers. As a result, the pseudo-optimum balance is discovered and our method is shown both not to increase downloading time and to limit the buffer space. This network software can be applied to download files smoothly absorbing the difference in performance characteristics among heterogeneous networks.
Jae Sam YOON Gil Ho LEE Hong Kook KIM
Existing standard speech coders can provide high quality speech communication. However, they tend to degrade the performance of automatic speech recognition (ASR) systems that use the reconstructed speech. The main cause of the degradation is in that the linear predictive coefficients (LPCs), which are typical spectral envelope parameters in speech coding, are optimized to speech quality rather than to the performance of speech recognition. In this paper, we propose a speech coder using mel-frequency cepstral coefficients (MFCCs) instead of LPCs to improve the performance of a server-based speech recognition system in network environments. To develop the proposed speech coder with a low-bit rate, we first explore the interframe correlation of MFCCs, which results in the predictive quantization of MFCC. Second, a safety-net scheme is proposed to make the MFCC-based speech coder robust to channel errors. As a result, we propose an 8.7 kbps MFCC-based CELP coder. It is shown that the proposed speech coder has a comparable speech quality to 8 kbps G.729 and the ASR system using the proposed speech coder gives the relative word error rate reduction by 6.8% as compared to the ASR system using G.729 on a large vocabulary task (AURORA4).
Kuang-Yow LIAN Hui-Wen TU Chi-Wang HONG
In this paper, we propose an integral-type T-S fuzzy control scheme to deal with the regulation problem of buck converters without current sensors. This current sensorless control of converters provides the output voltage to achieve zero steady-state error and is with high robust performance. The stability of the overall closed-loop system is rigorously analyzed by using Lyapunov's method. Based on an appropriate assumption, the separation principle can still succeed in the control problems. Hence, the controller and observer gains can be separately obtained by solving LMIs via Matlab's toolbox. The observer-based controller is realized with Simulink and digital signal processors (DSPs). The simulation and experimental results verify the feasibility of the proposed schemes and show the satisfactory performance for the power converters.
Junya SHIMIZU Yixin DIAO Maheswaran SURENDRA
One of the system greatly affecting the performance of a database server is the size-division of buffer pools. This letter proposes an adaptive control method of the buffer pool sizes. This method obtains the nearly optimal division using only observed response times in a comparatively short duration.
Zhonghua QUAN Soohee HAN Wook Hyun KWON
We propose a stability-guaranteed horizon size (SgHS) for stabilizing receding horizon control (RHC). It is shown that the proposed SgHS can be represented explicitly in terms of the known parameters of the given system model and is independent of the terminal weighting matrix in the cost function. The proposed SgHS is validated via a numerical example.
Hidetoshi OYA Kojiro HAGINO Masaki MATSUOKA
This paper deals with a design problem of an observer-based robust preview control system for uncertain discrete-time systems. In this approach, we adopt 2-stage design scheme and we derive an observer-based robust controller with integral and preview actions such that a disturbance attenuation level is satisfactorily small for allowable uncertainties.
Hiroshi MATSUURA Naotaka MORITA Tatsuro MURAKAMI Kazumasa TAKAMI
Multilayered network interaction among various networks such as IP/MPLS packet networks and optical fiber networks are now achieved using generalized multiprotocol label switching (GMPLS) technology. One unique feature of GMPLS networks is that GMPLS packet-layer label switching paths (LSPs), such as IP/MPLS LSPs, sometimes tunnel through GMPLS lower layer LSPs such as optical fiber/lambda LSPs. One problem that occurs in this situation is protecting an important primary packet LSP by using a protection LSP that is physically separated from the primary LSP. The packet router has difficulty recognizing lower layer LSPs that are totally disjointed from the primary LSP. This is because, in a GMPLS's packet layer, a source router only differentiates one lower layer LSP from another, and does not check the disjointedness of segments through which the lower layer path passes. Sometimes, different lower LSPs pass through the same optical fiber, and a malfunction of one optical fiber sometimes causes many lower layer LSPs to malfunction at the same time. To solve this problem, a shared risk link group (SRLG) is introduced. Network links that belong to the same SRLG share a common physical resource. We apply this SRLG to the proposed hierarchically distributed path computation elements (HDPCEs) and achieve effective disjointed SRLG protection for important primary GMPLS packet paths.
Seunglak CHOI Jinwon LEE Su Myeon KIM Junehwa SONG Yoon-Joon LEE
Most commercial Web sites dynamically generate their contents through a three-tier server architecture composed of a Web server, an application server, and a database server. In such an architecture, the database server easily becomes a bottleneck to the overall performance. In this paper, we propose WDBAccel, a high-performance database server accelerator that significantly improves the throughput of database processing. WDBAccel eliminates costly, complex query processing needed to obtain query results by reusing the results from previous queries for subsequent queries. This differentiates WDBAccel from other database cache systems, which employ traditional query processing. WDBAccel further improves its performance by fully utilizing main memory as the primary storage. This paper presents the design and implementation of the WDBAccel as well as the results of performance evaluation with a prototype.
Cho-chin LIN Da-wei WANG Tsan-sheng HSU
We discuss the problem of finding a dominant sequence for sending input data items from a low-end client to a server for computational intensive tasks under the realistic assumption of unpredictable communication behavior. Under this assumption, the client has to send the input data items using a specified sequence to maximize the number of computations performed by the server at any time. The sequence-finding problem is NP-hard for the general case. In this paper, we address three fundamental and useful applications: the product of two polynomials, matrices multiplication and Fast Fourier Transform. We show that the sequence-finding problems of the three applications can be solved optimally in linear time. However, we also show counter examples to rule out any possibility of finding a dominant sequence for sparse cases of the three applications. Finally, a simulation is conducted to show the usefulness of our method.
Sung-Kwan Youm Meejoung KIM Chul-Hee KANG
This paper considers the reliable multicast transport protocols used in hybrid networks that include wired and wireless networks and transparent proxy servers. We present four analytic performance models of two extreme reliable multicast transport protocols, sender-initiated and receiver-initiated, and supported and unsupported by transparent proxy servers are considered in each reliable multicast protocol. We analyze the throughputs of these four different models mathematically. Numerical results show that transparent proxy servers give good effects to overall performance. Furthermore, the receiver-initiated reliable multicast supported by transparent proxy servers gives better performances of total throughput than sender-initiated reliable multicast supported by transparent proxy servers. We provide efficiency criterion of the optimal number of transparent proxy servers for each protocol under varying wireless loss probabilities. Numerical results are verified by simulations.
Kyungtak YU Nam-Hoon JO Jin Heon SEO
In this letter, an illustrative example is given, which shows that the number of integrators needed for the dynamic observer error linearization using integrators can not be bounded by a function of the dimension of the system and the number of outputs in contrast to dynamic feedback linearization results.
Popular Web sites form their Web servers into Web server clusters. The Web server cluster operates with a load-balancing algorithm to distribute Web requests evenly among Web servers. The load-balancing algorithms founded on conventional periodic load-information update mechanism are not scalable due to the synchronized update of load-information. We propose a load-balancing algorithm that the load-information update is not synchronized by exploiting variant execution times of executing scripts in dynamic Web pages. The load-information of each server is updated 'individually' by a new load-information update mechanism, and the proposed algorithm supports high scalability based on this individual update. Simulation results have proven the improvement in system performance through another aspect of high scalability. Furthermore, the proposed algorithm guarantees some level of QoS for Web clients by fairly distributing requests. A fundamental merit of the proposed algorithm is its simplicity, which supports higher throughput of the Web switch.
Hiroyuki OKAMURA Satoshi MIYAHARA Tadashi DOHI
Long running software systems are known to experience an aging phenomenon called software aging, one in which the accumulation of errors during the execution of software leads to performance degradation and eventually results in failure. To counteract this phenomenon a proactive fault management approach, called software rejuvenation, is particularly useful. It essentially involves gracefully terminating an application or a system and restarting it in a clean internal state. In this paper, we evaluate dependability performance of a communication network system with the software rejuvenation under the assumption that the requests arrive according to a Markov modulated Poisson process (MMPP). Three dependability measures, steady-state availability, loss probability of requests and mean response time on tasks, are derived through the hidden Markovian analysis based on the time-based software rejuvenation scheme. In numerical examples, we investigate the sensitivity of some model parameters to the dependability measures.
Wei-Chi KU Hsiu-Mei CHUANG Min-Hung CHIANG
Recently, Juang proposed an efficient password authenticated key agreement scheme using smart cards for the multi-server architecture. Juang's scheme was intended to provide mutual authentication and session key agreement. Herein, we show that Juang's scheme is vulnerable to a privileged insider's attack and is not easily reparable. Furthermore, it does not provide forward secrecy and the user eviction mechanism.
Satosi KATO Hidetosi OKAMOTO Toyofumi TAKENAKA
We propose a novel routing algorithm for reverse proxy servers, called load balancing content address hashing (LB-CAH), and evaluate the performance of the proposed routing algorithm compared with that of the content address hashing (CAH) and the hash and slide (HAS) routing algorithms. The proposed LB-CAH routing algorithm calculates the popularity of pages in the load balancer using an LFU caching technique and periodically makes a popularity list. Using this popularity list, the proposed routing algorithm selects a reverse proxy server as follows. When the requested page appears in the popularity list, the request is routed according to the round robin method; otherwise, it is routed according to the content address hashing method. We evaluate and compare the LB-CAH, CAH and HAS routing algorithms by simulation experiments from the viewpoints of load balancing, consumed cache space and cache hit rate. Simulation experiments show that the proposed LB-CAH routing algorithm achieves almost the same degree of load balancing as the HAS algorithm and the same cache hit rate as the CAH algorithm for reverse proxy servers in various web site environments.
Zhihui WANG Tohru KIRYU Mamoru IWAKI Keisuke SHIBAI
General exercise approaches are not convenient for some people in undertaking appropriate exercise due to the limited variety of present programs at existing exercise machines. Moreover, continuous support by one sports doctor is only available for a limited number of users. In this paper, therefore, we propose an Internet-based technical framework, which is designed on multi-tiered client/server architecture, for integrating and easily upgrading exercise programs. By applying the technical framework, a cycle ergometer health promotion system was developed for providing personally fitted. We also presented some facilities to assist sports doctors in quickly designing and remotely improving individual exercise protocols against cycle ergometer exercise based on a history database. Then we evaluated the Internet-based cycle ergometer system during two months of feasibility experiments for six elderly persons in terms of usability. As a result, the Internet-based cycle ergometer system was effective for continuously supporting the personal fitting procedure.
Hiroshi MATSUURA Tatsuro MURAKAMI Kazumasa TAKAMI
The demand for intra- and interdomain routing for multilayered networks such as those using generalized multiprotocol label switching (GMPLS) is strong. One of the features that is peculiar to GMPLS networks is that because several different domains, such as those of IP, ATM, and optical fiber, are combined with each other hierarchically, various routing policies, which are sometimes independent from underlying domains and sometimes taking the underlying domains' policies into consideration, are required. For example GMPLS's lower layer LSPs like lambda LSP are expected to be established independently before the upper-layer LSPs, like IP and MPLS LSPs, are established in the underlying domains. Another requirement for the GMPLS interdomain routing is lightening the burden for selecting the interdomain route, because there are a lot of demands to interconnect many GMPLS domains. In order to satisfy these demands, we propose a path computation server (PCS) that is special for the intra/interdomain routing of GMPLS networks. As a counterpart of the proposed interdomain routing, it is now becoming popular to apply OSPF to the GMPLS interdomain routing. Therefore, we compared the proposed interdomain routing with OSPF, and show the applicability of the routing to GMPLS networks.
Wei-Chi KU Shen-Tien CHANG Min-Hung CHIANG
Recently, Lin, Hwang, and Li proposed an efficient remote authentication scheme using smart cards for multi-server architecture based on the geometric property of the Euclidean plane. Herein, we show that their scheme is vulnerable to two forgery attacks and a password-guessing attack, and is not easily repairable. Furthermore, their scheme lacks a proper user eviction mechanism.
With the occurrence of new applications such as Voice over IP (VoIP) and multimedia conference, there is an ongoing discussion about realizing QoS in the Internet today. Because of its potential scalability in support of QoS guarantees, the Differentiated Service (DiffServ) architecture with aggregate packet scheduling has recently attracted much attention in the networking community as a feasible solution for providing Internet QoS. Thus, it is important to understand delay bound of an individual flow in the DiffServ architecture in order to provide delay-sensitive applications. In this paper, we study, via both analysis and simulation, the deterministic bound on edge-to-edge delay of a flow in a DiffServ network domain with FIFO aggregation and a class-based Latency Rate (LR) server that provides guaranteed performance with rate reservation for a traffic class. We derive edge-to-edge delay bound for a single flow as a function of allocated service rate for a traffic class, token bucket parameters adopted for flows at the network ingress, and information about joining and leaving flows. We compare the obtained delay bound with previous works using analytic results, and then conduct simulation to confirm the results. The derived bound is less than that of previous studies in all cases.