In the circuit-switching literature, the Least Loaded Path Routing (LLR) concept has been shown to be very simple and efficient. However, it seems that there is no unique definition for the "least busy" path, i.e., how to measure the degree of "busy" of a path. In this paper, we examine six ways of defining the least busy path and a random policy. The performance of these policies is evaluated via both simulation and analysis. Our numerical results show that all policies, include the random policy, have almost the same performance under most of the network configurations. Only under extremely low traffic load conditions, the difference between the policies becomes significant. However, the magnitude of the difference is still very small (about 0.001). Therefore, we conclude that how to select the alternate path does not affect the performance of LLR-based routing algorithms significantly when the call blocking probability is not too small. Instead, we found that the trunk reservation level affects the performance of LLR-based routing algorithms significantly.

The Next Generation Internet Initiative was launched in the U.S. to advance key networking technologies that will enable a new wave of applications on the Internet. Now, in its third year, the program has launched and fostered over one hundred new research projects in partnership with academic, industrial and government laboratories. One key research area that has been emphasized within the program is the next-generation optical networking. Given the ever increasing demand for network bandwidth, and the recent phenomenal advances in WDM technologies, the Next Generation Internet is expected to be an IP-based optical WDM network. As IP over WDM networking technologies mature, a number of important architectural, management and control issues have surfaced. These issues need to be addressed before a true Next Generation Optical Internet can emerge. This paper provides a brief introduction to the overall goals and activities of DARPA's NGI program and describes the key architectural, management, and control issues for the Optical Internet. We review the different IP/WDM networking architectural models and their tradeoffs. We outline and discuss several management and control issues and possible solutions related to the configuration, fault, and performance management of IP over dynamic WDM networks. We present an analysis and supporting simulation results demonstrating the potential benefits of dynamic IP over WDM networks. We then discuss the issues related to IP/WDM traffic engineering in more detail, and present the approach taken in the NGI SuperNet Network Control and Management Project funded by DARPA. In particular, we motivate and present an innovative integrated traffic-engineering framework for re-configurable IP/WDM networks. It builds on the strength of Multi-Protocol Label Switching (MPLS) for fine-grain IP load balancing, and on the strength of Re-configurable WDM networking for reducing the IP network's weighted-hop-distance, and for expanding the bottleneck bandwidth.

We investigate traffic performance of CBWL schemes with multimedia services on non-homogeneous cellular network in which cut-off priority is given to handoff calls. Two generic routing schemes are analyzed: one is the randomized routing and the other is the least loaded routing. The performance measures that we focus on are the new call blocking probabilities and the handoff failure probabilities. To evaluate blocking probabilities of interest, we construct a generalized access network whose blocking probabilities are same as ones for CBWL systems. For analysis of generalized access network, we apply the reduced load approximation. The computational complexity and memory requirements of proposed algorithm are linear so that we can use this algorithm to approximate blocking probabilities of CBWL systems of large size. The proposed approximations are tested on a number of simple examples. Numerical results for 12 cells and 36 cells networks are given. The comparison between approximation and simulation results shows that the proposed approximation method is quite accurate.

Sticking is one of dominant characteristics of reliability in relays for medium current loads from several amperes to several dozen amperes, which are used for relays for automobiles, industrial control units or power supplies of household electrical appliances. Correlations between the release failures due to sticking and contact characteristics such as arc discharges, material parameters and design factors in relays have never been always made clear. This puts difficulty in the way of reasonable development of contact materials and rational design of relays. So, dependence of electrical load conditions on sticking characteristics are investigated, using the Ag-CdO contacts which have had high practical use to relays for medium current loads. Furthermore, relationship among the sticking characteristics, arc discharge characteristics and contact surface properties after operations are studied. Mechanism of sticking is considered on the basis of those data. The results are as follows: (1) Sticking phenomenon occurs intermittently from initial operations and lasts to the end. (2) The µ + 2 σ value (the sum of the mean value and the integral multiple of the standard deviation of sticking force) increases in proportion to the circuit current. On the other hand, it has the maximum value at a circuit voltage, slightly less than the minimum arc voltage. (3) Factors causing the sticking are considered to be divided into direct factors and its root factors. It is considered that a dominant direct factor is welding, and that its root factor is bridge or welding by Joule's heat. On the other hand, the sticking force becomes rather lower as the circuit voltage increases, in the circuit voltage range where regular arc discharge occurs.

This paper first examines the issues related to scheduling loop applications on a software distributed shared memory (DSM) system. Then, a dynamic scheduling scheme is developed based on the examined issues to enhance the performance of loop applications on DSM. Compared with previous works, the proposed scheme has several specialties. The first is that the workload of processors can be effectively balanced even when the computational capabilities of processors and the computational needs of threads are not identical. The second is it divides thread mapping into two phases, each with one consideration, i.e., load balance or communication cost, and adopts thread migration and exchange in the two phases, respectively. The third is the exploitation of data sharing among threads to reduce data-consistency communication, and the last is to attack the negative effect of the unnecessary inter-node sharing caused by thread re-mapping. The proposed scheme has been implemented on a page-based DSM system called Cohesion. Our experiments show that the proposed scheme is more effective to improve the performance of the test programs than related schemes.

Learning process is essential for good performance when a neural network is applied to a practical application. The backpropagation algorithm is a well-known learning method widely used in most neural networks. However, since the backpropagation algorithm is time-consuming, much research have been done to speed up the process. The block backpropagation algorithm, which seems to be more efficient than the backpropagation, is recently proposed by Coetzee in [2]. In this paper, we propose an efficient parallel algorithm for the block backpropagation method and its performance model in mesh-connected parallel computer systems. The proposed algorithm adopts master-slave model for weight broadcasting and data parallelism for computation of weights. In order to validate our performance model, a neural network is implemented for printed character recognition application in the TiME which is a prototype parallel machine consisting of 32 transputers connected in mesh topology. It is shown that speedup by our performance model is very close to that by experiments.

To provide a bounded-delay service with an end-to-end delay guarantee and utilize network efficiently, the admission control functions in a network require an adequate per-switch delay allocation policy to allocate end-to-end delay requirement to a set of local switches. This letter present an optimal delay allocation policy, according to a deterministic traffic model, and a rate-controlled scheduler. Empirical results indicate that the proposed policy perform better than EQ allocation policy, using the admission load region as index.

Demand for power in Japan has been increasing year by year, and steep demand is projected during daily peak load periods: particularly in summer, due to growing demand for air conditioning. This has resulted in a large gap between day and night demand for power. The daily and seasonal regularity of this demand gap is placing pressure on power utilities to reduce service costs and create a more dependable power supply. This study demonstrates the feasibility of an energy storage system for load leveling based on the electric double-layer capacitor (EDLC). This device is safer, has a longer service life and needs far less maintenance than the secondary cell. The system works to store surplus energy from a commercial AC line in an EDLC bank during the night, and release this energy for use during the daytime peak load period, using a novel interface circuit. This paper focuses in particular on the working principles and experimental results of the interface circuit, which comprises a voltage control oscillator (VCO), a bi-directional DC/DC converter, a bi- directional inverter, and a coupling inductor. The whole circuit is subjected to PLL control, so that automatic connection between DC from an EDLC bank and AC from a commercial power line may take place in a simpler, more reliable and less costly manner. The system allows for energy transfer on the basis of DC voltage as if electric charging and discharging had taken place in a full DC system.

It is generally known that the autonomic nervous system regulates the pupil. In this study, we attempted to assess mental workload on the basis of the fluctuation rhythm in the pupil area. Controlling the respiration interval, we measured the pupil area during mental tasking for one minute. We simultaneously measured the respiration curve to monitor the respiration interval. We required the subject to perform two mental tasks. One was a mathematical division task, the difficulty of which was set to two, three, four, and five dividends. The other was a Sternberg memory search task, which had four work levels defined by the number of memory sets. In the Sternberg memory search, the number of memory set changed from five to eight. In such a way, we changed the mental workload induced by mental loading. As a result of calculating an autoregressive (AR) power spectrum, we could observe two peaks which corresponded to the blood pressure variation and respiratory sinus arrhythmia under a low workload. With an increased workload, the spectral peak related to the respiratory sinus arrhythmia disappeared. The ratio of the power at the low frequency band, from 0.05-0.15Hz, to the power at the respiration frequency band, from 0.35-0.4Hz, increased with the work level. In conclusion, the fluctuation of the pupil area is a promising means for the evaluation of mental workload or autonomic nervous function.

An efficient scheme for establishing the multicast-path on ATM network is to configure the tree-shaped path via several intermediate copy nodes. This scheme needs the multicast routing control. Thus, restricting the number of copy nodes was proposed since it make this control fast and simple. However, if restricted number of copy nodes is fixed in the network design, multicast traffic will concentrate around copy nodes, and as a result, the possibility of occurring congestion will be higher. In our research, we permit restricted number of nodes as intermediate nodes which branch the tree-shaped path at routing, and name this permission "copy-token. " Namely, the node which has "copy-token" is defined as the copy node. Our research purpose is to distribute the multicast traffic by adaptively allocating "copy-token" to nodes which satisfy the conditions of both the priority for multicasting such as lower offered load and the geographical distribution at the same time. Finally, we evaluate the performance of our proposed routing scheme on the blocking probability through computer simulations.

This paper analyzes the effective radiation efficiency of a ground-penetrating radar (GPR) consisting of two resistor-loaded bow-tie antennas covered with a ferrite-coated conducting cavity by using finite-difference time-domain (FDTD) method. The GPR is located above a lossless or lossy ground surface. The relation between the radiation powers into the ground and air, dissipated powers at the loaded resistors and ferrite absorber, and the reflected power due to impedance mismatching, is clarified numerically.

The applications of reactance-loaded beam tilting dipole antennas have been reported by many researchers. The reactance elements loaded on the applications reported up to date have been used only for the purpose of beam tilting. This paper presents the basic characteristics of the beam tilting dipole antenna element in which one reactance element is used for the impedance matching at the feed point. The radiation pattern is tilted by the properly determined driving point position, and the loading reactance is used to obtain forced resonance without great changes in tilt angle. The numerical results demonstrate that the reactance element should be loaded in the region where the driving point is placed to obtain forced resonance of the antenna with little changes in beam tilt angle. In case the proposed forced resonant beam tilting antenna with 0.8λ length is driven at 0.2λ from the center, the main beam tilt angle of 57.7 degrees, the highest power gain of 8.6 dB, and VSWR=2.2 are obtained.

In this paper, we show two process allocation schemes to tolerate multiple faults when the primary-backup replication method is used. The first scheme, called multiple backup scheme, is running multiple backup processes for each process to tolerate multiple faults. The second scheme, called regenerative backup scheme, is running only one backup process for each process, but re-generates backup processes for processes that do not have a backup process after a fault occurrence to keep the primary-backup process pair available. In both schemes, we propose heuristic process allocation methods for balancing loads in spite of the occurrence of faults. Then we evaluate and compare the performance of the proposed heuristic process allocation methods using simulation. Next, we analyze the reliability of two schemes based on their fault-tolerance capability. For the analysis of fault-tolerance capability, we find the degree of fault tolerance for each scheme. Then we find the reliability of each scheme using Markov chains. The comparison results of two schemes indicate that the regenerative single backup process allocation scheme is more suitable than the multiple backup allocation scheme.

In this paper, we describe the PS3 load distribution scheme. A target service is a transaction service consisting of multiple processes that communicate with each other. A target system consists of workstations connected by a LAN. PS3 determines the process allocation by estimating response times and throughputs. It allows us to set an upper limit of a response time, and to set lower and upper limits for the throughput of each service. PS3 tries to find a process allocation that provides the minimum response time under conditions set by the user in advance. We measured the response times and throughputs and compared the values with the estimated ones. The results show that PS3 provides an appropriate process allocation, and that calculated results agree well with the measured ones.

A partial buffer sharing scheme is proposed as loss-priority control for a finite buffer with batch Poisson inputs under a whole batch acceptance rule. Customer and batch loss probabilities for high- and low-priority customers are derived under this batch acceptance rule using a supplementary variable method. A comparison of the partial buffer sharing scheme and a system without loss-priority control is made in terms of admissible offered load. Whole batch acceptance and partial batch acceptance rules are also compared in terms of admissible offered load.

An access attempt overload control method is proposed for CDMA cellular system to provide good communication quality to mobile stations under the traffic overload condition. Under the normal traffic condition, the effect of access attempts on the interference is negligible. But under the traffic overload condition, the interference due to access attempts could be large and deteriorate the communication quality (or reduce the capacity of a reverse link) if access attempts are not controlled properly. The numerical and simulation results show that our access attempt overload control method can maintain the interference caused by access attempts very low and the capacity of the reverse link is not reduced under the traffic overload condition.

A service control node in the Advanced Intelligent Network (AIN) allocates data for customers among multiple modules and performs distributed processing of multiple transactions. In such a node, load can vary among the modules due to dispersion in the amount of traffic for each customer. It is therefore important to balance out this load variation and raise the utilization of each module in order to achieve an efficient distributed processing system. We first propose a method for balancing the load among modules by dynamically transferring customer data in units of records from high-load modules to low-load modules. Then, based on this method, a method for selecting records to be transferred between modules is also proposed. And we clarify the processor overhead for transferring records. The effect of the reduction of number of modules by load balancing is also evaluated. Based on the these results, it is shown that dynamic transferring of records is an effective scheme for balancing load among modules in a service control node of the AIN.

This paper proposes a medium access control (MAC) protocol for single-hop WDM star-coupler networks, in which the number of stations is larger than the number of channels and the stations have arbitrary distances to the star coupler. The proposed protocol has one control channel for reserving the slots of data channels and several data channels which are used to transmit traffic. This paper also suggests a scheme that accomplishes load balancing among data channels for arbitrary traffic patterns between stations. Since this load balancing scheme diminishes an influence that traffic patterns have on the performance of the proposed MAC protocol, the proposed system is appropriate for systems which have asymmetric traffic intensity between stations. Throughput and mean message delay of the MAC protocol are analyzed using a discrete time Markov process and a D/G/1 queue with batch arrivals. The numerical results show that the performance is improved as the message length increases, the maximum round-trip propagation delay decreases, and the number of data channels increases.

Layout has strong influence on matching properties of a circuit. Current matching models, which characterize both local random non-uniformities and global systematic non-uniformities stochastically, are not adequate for the matching analysis taking the effect of layout realization into account. In order to consider topological information of layout into matching analysis, we propose a matching model which treats the random and systematic components separately. Also, we characterize the micro-loading effect, which modulates fabricated line-width according to the local density of layout patterns, into matching analysis. With these two techniques, we can perform matching analysis of CMOS circuits taking layout information into account.

Workers involved in software projects are unlike those working on a production line in a manufacturing field usually engaged in plural work (that is, not only main development work but also various other work), concurrently. Such other work might put pressure on the schedule of the whole project. Therefore, to manage the whole project, not only main development work but also various other work should be dealt with as management objects and workers' workload should be taken into consideration (that is, who is doing what work at what workload at what time). This paper proposes a framework for workload management facilities for managing software projects. This framework proposes to relate not only main development work but also various other work and each work step within cooperative work to the workers. This paper also shows the behavior of the facilities by using an example and shows its usefulness based on the application of a prototype system. Using this system, users can assign work to workers by simulating workers' workload. These facilities help managers grasp workers' workload as well as help workers grasp their assigned work.