Browsing by Author "Liu, WC"
Now showing 1 - 12 of 12
- Results Per Page
- Sort Options
Item type:Item, An experience of using FDOOA method to develop a telecommunications network management system(1998) ;Hung, YS ;Liu, WC ;Chung, CG ;資訊工程學系Department of Computer ScienceMost OOA methods ignore the importance of system functions that are the focus of users, and require an extra effort to remove the redundant classes introduced in identifying candidate classes from the norms and verbs of problem description To remedy these drawbacks, we proposed the FDOOA (Functional-Decomposition OOA) method. The system function model is introduced; the class model is constructed through a parallel decomposition of function and data and bottom-up composition to formulate a class hierarchy. We also design an executable specification language - AOORPL and a supporting environment to construct the rapid prototype automatically so that the user can validate the specification visually. To evaluate feasibility of the FDOOA method on large-scale software systems, we use a telecommunications network management system as example to define its system specification. This system includes 44 services in total, and its system specification has 112 system functions/subfunctions and 183 classes. After this evaluation, we indeed demonstrate the advantage of facilitating the communication,vith the users, reducing redundant classes, decreasing the cost of modifications, and simplifying the verification of requirement specifications.Item type:Item, APER-2: A developer-centered, object-oriented process language(2000) ;Chen, JYJ ;Chou, SC ;Liu, WC ;資訊工程學系Department of Computer ScienceThis paper presents a developer-centered, object-oriented process language called APER-2. The language offers following features 1) it is developer-centered meaning that process programs map to developers which presumably is more realistic than commonly seen activity-centered or product-centered modeling, 2) it is object-oriented which models all process components as classes. This is expected to offer object orientation benefits such as reusability and maintainability.Item type:Item, Cross-layer goodput analysis for rate adaptive IEEE 802.11 a WLAN in the generalized Nakagami fading channel(2004) ;Wang, LC ;Lin, YW ;Liu, WC ;交大名義發表National Chiao Tung UniversityThis paper aims to evaluate the goodput performance of the IEEE 802.11a wireless local area network (WLAN) from both the media access control (MAC) layer and physical (PHY) layer perspectives. From the physical layer perspective, we analyze the packet error rate performance of the orthogonal frequency division multiplexing (OFDM) based WLANs under the generalized Nakagami fading channel. According to the Chernoff bound analysis in the Nakagand fading channel, we derive the approximate packet error rate for the IEEE 802.11a WLAN with different data rates and fading parameters. Furthermore, we propose an efficient channel driven rate adaptation (CDRA) scheme. Through a PHY/MAC cross-layer analysis, we demonstrate that the goodput of the CDRA scheme indeed approaches to that of the optimal dynamic programming based rate adaptation method, while avoiding the complex calculations in selecting transmission parameters as the optimal method.Item type:Item, iNetwork - a GSM compliant IF-based communication platform(2000) ;Cheng, CW ;Liu, WC ;Chung, CG ;資訊工程學系Department of Computer ScienceEnterprise communication systems require data, voice, and mobile communication services. However, the cost to maintain many disparate networks is high. We propose iNetwork, an IF-based communication system complied with GSM to provide these services at an integrated platform Furthermore, an iNetwork system can cooperate with other iNetwork, Internet, PSTN or GSM systems to provide global services.Item type:Item, Locating faults on cables by applying the principle of the transmission line oscillator(2003-05-01) ;Liu, WC ;Lu, MH ;光電工程學系Department of PhotonicsA new method for locating the break point of a cable has been developed. It operates by application of the principle of the transmission line oscillator. Coaxial cables of four different lengths (1.5 m, 5 m, 10 m, 20 m) were measured in accordance with this method. The results show high precision (deviation <0.5 cm) and accuracy (error &LE;3 cm).Item type:Item, On the performance of using multiple transmit and receive antennas in pulse-based ultrawideband systems(2005-11-01) ;Wang, LC ;Liu, WC ;Shieh, KJ ;電信工程研究所Institute of Communications EngineeringThis paper presents an analytical expression for the signal-to-noise ratio (SNR) of the pulse position modulated (PPM) signal in an ultrawideband (UWB) channel with multiple transmit and receive antennas. A generalized fading channel model that can capture the cluster property and the highly dense multipath effect of the UWB channel is considered. Through simulations, it is demonstrated that the derived analytical model can accurately estimate the mean and variance properties of the pulse-based UWB signals in a frequency-selective fading channel. Furthermore, the authors investigate to what extent the performance of the PPM-based UWB system can be further enhanced by exploiting the advantage of multiple transmit antennas or receive antennas. Numerical results show that using multiple transmit antennas in the UWB channel can improve the system performance in the manner of reducing signal variations. However, because of already possessing rich diversity inherently in the UWB channel, using multiple transmit antennas does not provide diversity gain in the strict sense [i.e., improving the slope of bit error rate (BER) versus SNR] but can possibly reduce the required fingers of the RAKE receiver for the UWB channel. Furthermore, because multiple receive antennas can provide higher antenna array combining gain, the multiple receive antennas technique can be used to improve the coverage performance for the UWB system, which is crucial for a UWB system due to the low transmission power operation.Item type:Item, Path-based protocol verification approach(2000-03-01) ;Liu, WC ;Chung, CG ;資訊工程學系Department of Computer ScienceProtocol verification is one of the most challenging tasks in the design of protocols. Among the various proposed approaches, the one based on reachability analysis (or known as state enumeration approach) is of the most: simple, automatic and effective. However, the state explosion problem is a principle obstacle toward successful and complete verifications of complex protocols. To overcome this problem, we proposed a new approach, the "path-based approach." The idea is to divide a protocol into a collection of individual execution record, denoted as: concurrent paths, a partial order representation recording the execution paths of individual entities. Then, the verification of the protocol is, thus, replaced by that of individual concurrent paths. Since concurrent paths can be automatically generated through Cartesian product:of the execution paths of all modules, and verified independently, the memory requirement is limited to the complexity of individual concurrent path rather than the whole protocol. Thus, the state explosion problem is alleviated from such "divide and conquer" approach. Furthermore, we propose an algorithm, making the trade-off on memory requirement to generate the concurrent paths more efficiently; and utilize the technique of symmetric verification, parallel computing to improve the efficiency of verification. Eventually, our experience of verifying real protocols shows that our approach uses much less memory:and time than reachability analysis. (C) 2000 Elsevier Science B.V. All rights reserved.Item type:Item, Performance of pulse position modulated signals over the ultra-wideband channel with multiple transmit and receive antennas(2004) ;Wang, LC ;Liu, WC ;電信工程研究所Institute of Communications EngineeringThis paper presents a closed-form expression for the signal to noise ratio (SNR) of the pulse position modulated (PPM) signal in the ultra-wideband (UWB) channel with multiple transmit and receive antennas. We consider a generalized fading channel model that can capture the cluster property and the highly dense multipath effect. Through simulations, we demonstrate that the derived analytical model can accurately estimate the PPM signals of the UWB system in such a frequency selective fading channel. Furthermore, we investigate to what extend the performance of the PPM based UWB system can be further enhanced by exploiting the advantage of the multiple antenna systems. Our results show that the synergy of combining the time-switched transmit diversity (TSTD) scheme with a template-based detection can be further exploited to improve the UWB system performance.Item type:Item, Physical layer effects on the MAC goodput performance for the rate adaptive IEEE 802.11 a/g WLAN(2004) ;Liu, WC ;Wang, LC ;Lin, YW ;交大名義發表National Chiao Tung UniversityAccording to the IEEE 802.11n, the next generation wireless local area networks (WLAN) aim to support data rates at least 100 Mbps, as measured at the medium access control (MAC) layer. Thus, it is important to develop a cross-layer analytical model to evaluate the MAC layer goodput performance with consideration of the physical layer effects. This paper suggests such an analytical model to calculate the MAC layer goodput for the IEEE 802.11a WLAN in the Nakagami fading channel, while incorporating the effects of channel estimation, delay spread and signal detection scheme in the physical layer. Furthermore, we develop a simple and efficient channel-driven rate adaptation (CDRA) scheme to dynamically adjust the transmission parameters to maximize the MAC layer goodput. From out numerical results, we find that some physical (PHY) modes in the IEEE 802.11a are unnecessary. We also find that as rms delay spread increases, the goodput in the MAC layer can be improved due to frequency diversity. For the case of rms delay spread changing from 50 to 200 nsec with 15 dB E-b/N-0, the goodput is improved from 12 Mbps to 15 Mbps.Item type:Item, Symbolic path-based protocol verification(2000-03-01) ;Liu, WC ;Chung, CG ;資訊工程學系Department of Computer ScienceThe principal problem in protocol verification is state explosion problem. In our work (W.C. Liu, C.G. Chung, Path-based Protocol Verification Approach, Technical Report, Department of Computer Science and Information Engineering, National Chiao-Tung University, Hsin-Chu, Taiwan, ROC, 1998), we have proposed a "divide and conquer" approach to alleviate this problem, the path-based approach. This approach-separates the protocol into a set of concurrent paths, each of which can be generated and verified independently of the others. However, reachability analysis is used to identify the concurrent paths from the Cartesian product of unit paths, and it is time-consuming. Therefore,:in this paper, we propose a simple and efficient checking algorithm-to identify the concurrent paths from the Cartesian product, using only Boolean and simple arithmetic operations. (C) 2000 Elsevier Science B.V; All rights reserved.Item type:Item, The design of a signaling system in an enterprise network system(2003) ;Cheng, CW ;Liu, WC ;Chung, CG ;資訊工程學系Department of Computer ScienceAn enterprise communication system has the requirements of data, voice, and mobile communications. We have proposed iNetwork as an enterprise-based communication system that integrates GSM services and IP network resources to establish a low cost communication environment. In this paper we discuss the design of signaling system in iNetwork system, and study the performance and efficiency of iNetwork by investigating the call setup time and mis-routing events. The resultant design has less routing stages than the original design of GSM and it can avoid bandwidth wastage and enhances the utilization of network resources.Item type:Item, The measurement of propagation delay in multimode optical fibre with pulse-reflection-oscillation method(2004-02-01) ;Liu, WC ;Lu, MH ;光電工程學系Department of PhotonicsUsing the pulse-reflection-oscillation method, a high precision optical fibre propagation delay measurement system is made and applied to measure multimode optical patcheords of four different lengths (10, 20, 50 and 100 m). The result shows that the standard deviation is less than I I ps. (C) 2003 Elsevier Ltd. All rights reserved.