• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.4 no.5
• /
• pp.1143-1150
• /
• 2000

A dual-mode multiplier (DMM) that performs single- and double-precision multiplications has been designed using a $0.25-\mum$ 5-metal CMOS technology. An algorithm for efficiently implementing double-precision multiplication with a single-precision multiplier was proposed, which is based on partitioning double-precision multiplication into four single-precision sub-multiplications and computing them with sequential accumulations. When compared with conventional double-precision multipliers, our approach reduces the hardware complexity by about one third resulting in small silicon area and low-power dissipation at the expense of increased latency and throughput cycles. The DMM consists of a $28-b\times28-b$ single-precision multiplier designed using radix-4 Booth receding and redundant binary (RB) arithmetic, an accumulator and a simple control logic for mode selection. It contains about 25,000 transistors on the area of about $0.77\times0.40-m^2$. The HSPICE simulation results show that the DMM core can safely operate with 200-MHZ clock at 2.5-V, and its estimated power dissipation is about 130-㎽ at double-precision mode.

• The KIPS Transactions:PartA
• /
• v.9A no.3
• /
• pp.379-386
• /
• 2002

This paper presents a scheduling approach for synthesizing pipelined datapaths under resource constraints. Our approach builds up a schedule based on gradual mobility reduction in contrast to other algorithms of previous researches, where an operation being scheduled is selected by using a priority function. The proposed method consists of a scheduling algorithm and a decision algorithm for detecting any violation against resource constraints. Our approach evaluates whether or not a scheduling solution can exist in case an operation temporarily is assigned to the earliest or latest control step among the assignable steps for the operation. If a solution cannot be found, it is impossible to assign the operation to that step due to a violation against resource constraints, and so we can eliminate that control step. This process is iterated until a reduction of mobility for all operations can not be obtained. Experiments on benchmarks show that this approach gains a considerable improvement over those by previous approaches.

• Journal of Korean Society of Transportation
• /
• v.24 no.2 s.88
• /
• pp.169-178
• /
• 2006

Traffic assignment has been used both for predicting travel demands and for evaluating the tools for alleviating congestion on road network in advance. Some assignment models have been proposed such as equivalent mathematical minimization method, variational inequality problem, nonlinear complementary problem and fixed point method, in following the principle of Wardrop (1952) that no driver can not Improve his travel cost by unilaterally changing his route. Recently Jin(2005a) presented a traffic assignment model based on dynamic process. This paper proposes a solution algorithm for the model of Jin and assesses the performances. Compared to the Frank-Wolfe method, which has been wildly used for solving the existing assignment models, the proposed algorithm is expected to be more efficient because it does not need to evaluate the objective function. Two numerical examples are used for assessing the algorithm, and they show that the algorithm converges to user equilibrium of Wardrop.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.6
• /
• pp.51-58
• /
• 2015

This paper suggests assignment-swap algorithm with time complexity O(mn) to obtain the optimal solution for large-scale of transportation problem (TP) with incomplete cost lists. Generally, the TP with complete cost lists can be solved with TSM (Transportation Simplex Method). But, we can't be solved for large-scale of TP with TSM. Especially. It is hard to solve for large-scale TP with incomplete cost lists using TSM. Therefore, experts simply using commercial linear programming package. Firstly, the proposed algorithm applies assignment strategy of transportation quantity to ascending order of transportation cost. Then, we reassign from surplus of supply to shortage of demand. Secondly, we perform the 2-opt and 1-opt swap optimization to obtain the optimal solution. Upon application to $31{\times}15$ incomplete cost matrix problem, the proposed assignment-swap algorithm more improves the solution than LINGO of commercial linear programming.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.1 s.39
• /
• pp.107-117
• /
• 2006

Java language creates class files through Java compiler. Class files include informations involved with achievement of program. We can do analysis and optimization for efficient codes by analyzing class files. This paper analyzes bytecodes using informations of Java class files. We translate stack-based Java bytecodes into 3-address codes. Then we translate into static single assignment form using the 3-address codes. Static single assignment form provides a compact representation of a variable's definition-use information. Static single assignment form is often used as an intermediate representation during code optimization. Static sing1e assignment form renames each occurrence of a variable such that each variable is defined only once.

• International Journal of Highway Engineering
• /
• v.11 no.1
• /
• pp.47-58
• /
• 2009

Calibrating toll roads of highway networks needs additional weights within volume delay functions not like other general highway lints. However, current methods assigning additional weights in the volume delay function of toll roads are not sufficiently enough to predict real toll road volumes measured, since it does not consider discounting rates and an extra charges. This study develops methods to improve relevant and reliable volume delay functions. Suggested ideas include a method of weighting volume delay functions considering a value of time of vehicle types, a method of weighting volume delay functions considering lane distributions of vehicles, and a method of weighting volume delay functions considering percentages of link lengths per a number of lanes of toll roads. It is found that the method of weighting volume delay functions considering lane distributions of vehicles show most reliable and appropriate results, while the first method shows overestimation and the third method does underestimation of highway link volumes. In terms of assignment methods, total OD equilibrium assignment shows better results than PCU based assignment.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.14 no.1
• /
• pp.31-45
• /
• 2015

Existing link-based Frank-Wolfe algorithm has been widely used, thanks to its ease of simulation and stable results; however, it comes with low convergence issue towards near the optimum value. Such issue was not considered as a major drawback in the past. However, in the present, some arguments have occurred over the method's stability, analysis time, and other limits as the size and details of the fundamental data for traffic analysis have vastly improved. Therefore, this paper compared the theoretical attributes and the pros and cons between the Frank-Wolfe algorithm and the Origin-based algorithm and Path-based algorithm newly being developed. As a result of this paper, there is possibility that a problem of stability may arise depending on the convergence and exit criteria. Thus, In practice, this effort to derive the appropriate level of convergence is required to secure and stable results.

• Journal of the Korean Society for Railway
• /
• v.11 no.6
• /
• pp.550-561
• /
• 2008

The purpose of this paper is to propose a new transit assignment model for intercity rail networks. The characteristics of intercity rail are different from that of public transit in urban area. Line selection probability on route section is introduced to include the characteristics of intercity rail into transit assignment model. Network expansion is more simplified by a assumption line selection probability is externally given. The generalized cost is used to decide the volume of each transit line in most of existing transit assignment models. But, many variables have influence on the volume of each line such as time schedule of transit lines, inter-station distance, passengers' income, seasonal variation of demand and regional characteristics. The influence of these variables can be considered to decide the volume of each line by introducing line selection probability on route section. The tests on a small scale network show that the model proposed in this paper is superior to existing models for predicting intercity rail demand. Proposed model is suitable to consider the complicated fare structure of intercity rail and to draw inter-station demand directly as a result of assignment procedure.

• Journal of Korean Society of Transportation
• /
• v.25 no.6
• /
• pp.19-31
• /
• 2007

When estimating the benefits from an investment project in the transportation sector, errors caused by many factors may exist. This study focuses on user equilibrium traffic assignment methods and stopping criteria. According to previous studies, when using a user equilibrium assignment model, the benefits of travel time savings can be effected by the relative gap value. As the stopping criteria decreases, the time needed for traffic assignment increases, so that lowering the criteria cannot be the best solution. Therefore, an effort is necessary to reduce this change rate and thus improve reliability. This paper considers three methods: reducing the links subject to benefit calculation, extracting sub-area O/D tables and networks, and applying the mean value of successive traffic assignment results. The results of the analysis show that the method using the mean value of five results is more proper than the other methods. Using the sub-area analysis method, if the study area is small the benefits of a project might be over- or under-estimated. This paper used a nationwide O/D table and network at peak time as a case study. The resulting patterns can differ according to basic data to be used in analysis. So further analysis using the data from metropolitan areas are needed.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.48 no.10
• /
• pp.72-81
• /
• 2011

In this paper, we proposed and designed a novel floating point converter which supports single and double precisions of IEEE754 standard. The proposed convertor supports conversions between floating point number single/double precision and signed fixed point number(32bits/64bits) as well as conversions between signed integer(32bits/64bits) and floating point number single/double precision and conversions between floating point number single and double precisions. We defined a new internal format to convert various input types into one type so that overflow checking could be conducted easily according to range of output types. The internal format is similar to the extended format of floating point double precision defined in IEEE754 2008 standard. This standard specifies that minimum exponent bit-width of the extended format of floating point double precision is 15bits, but 11bits are enough to implement the proposed converting unit. Also, we optimized rounding stage of the convertor unit so that we could make it possible to operate rounding and represent correct negative numbers using an incrementer instead an adder. We designed single cycle data path and 5 cycles data path. After describing the HDL model for two data paths of the convertor, we synthesized them with TSMC 180nm technology library using Synopsys design compiler. Cell area of synthesis result occupies 12,886 gates(2 input NAND gate), and maximum operating frequency is 411MHz.