Search

Content sharing in Delay Tolerant Networks is largely contingent on the cooperation of participating nodes as message dissemination is executed in a store-carry-forward approach. The scope of the term cooperation has largely been limited to content forwarding in prior work. This work considers a universe of distributed content- sources that generate disjoint contents. A population of users who are capable of moving from one source to another are interested only in some selected subset of the localized contents. Interest of users is assumed to vary from one another; being rational entities, they are solely interested in collecting their preferred content(s) at the expense of minimum cost. With solutions like Direct Reciprocity been proposed earlier to force a user to be altruistic in message forwarding, our work starts in this backdrop where preference-based content caching and deletion strategy comes into play even if a user is forced to share her contents with all other encountered users. Alongside, we introduce mobility as a user strategy to characterize the non-cooperation of a user along with deletion of non-preferred cached contents. Under this premise, we explore the optimal strategies that a user can exercise to maximize her benefit corresponding to the strategies adopted by other users. More specifically, we investigate the joint impact of user's mobility and caching strategy on self and overall network performance using a Public Goods game in the Social Dilemma framework evincing the condition when non-cooperation is costlier than cooperation.

This study demonstrates a new possibility of estimating intelligibility of speech in informational maskers. The temporal and spectral properties of sound maskers are investigated to achieve acoustic privacy in public spaces. Speech intelligibility (SI) tests were conducted using Japanese sentences in daily use for energy (white noise) or informational (reversed speech) maskers. We found that the masking effects including informational masking on SI might not be estimated by analyzing the narrow-band temporal envelopes, which is a common way of predicting SI under noisy conditions. The masking effects might instead be visualized by spectral auto-correlation analysis on a frame-by-frame basis, for the series of dominant-spectral peaks of the masked target in the frequency domain. Consequently, we found that dissimilarity in frame-based spectral-auto-correlation sequences between the original and masked targets was the key to evaluating maskers including informational masking effects on SI.

The primary challenges faced by wireless sensor networks are how to construct the shortest spanning tree and how to determine the optimal sink node position in terms of minimizing the data transmission times and their variances for data gathering from all sensor nodes to a sink node. To solve these two problems, we propose a novel algorithm that uses the polygonal affine shortening algorithm with flow aggregation. This algorithm enables a wireless sensor network that has movable sensor nodes and one movable sink node to self-organize the shortest spanning tree and self-determine the optimal sink node position in a fully distributed manner. We also show that our algorithm is faster than the existing shortest path algorithm in terms of computational complexity.

A high-capacitance, compact, low-cost, and convenient optical data-link subsystem for parallel computers is developed using an 8 Gb/s/spl times/1.1 Gb/s multimode fiber (MMF) array optical module and newly developed data-link IC. Although the subsystem uses an MMF ribbon, parallel data is successfully transmitted over 1 km due to /spl plusmn/15-ns deskew operation of the IC. The subsystem operated stably in a processor networking testbed system of the Real World Computer-1, a newly constructed parallel computer. Random packets are transmitted without any error for over 17.5 h, corresponding to a bit-error rate (BER) of less than 10/sup -15/.

A 400 Gbps backplane switch was developed with low-cost, small-size, 8-channels 10 Gbps/port optical I/O and a SiGe Bi-CMOS switch LSI on a 60x60-mm 2 BGA package. It indicates the applicability of backplane switch for high throughput backplane interconnections.

Summary form only given. In conclusion, a low crosstalk 8-channel optical transceiver was developed for star topology access networks for the first time. A receiver dynamic range of 21.6 dB and an allowable interchannel receiving power difference of 16.6 dB, both large enough for fiber in the loop (FITL) system with a serving area of 7.2 km, were obtained at 155 Mbit/s.

A 400 Gbit/s backplane switch was developed with a low-cost, small-size, 8-channel 10 Gbit/s/port optical I/O and a SiGe Bi-CMOS switch LSI on a 60/spl times/60-mm-square BGA package. It indicates the applicability of OIP for high throughput backplane interconnections.

A new-type 10-Gbps VCSEL-TOSA (vertical cavity surface emitting laser transmitter optical sub-assembly) was demonstrated with single-mode-fiber direct coupling optics and a flexible substrate platform. The TOSA showed high fiber coupling efficiency (-1.6 dB) and clear 10-Gbps eye-opening with -2.6 dBm average optical power and 6 dB extinction ratio.

There is a dilemma in choosing between high density packaging and low loss transmission line. Though the LSI circuit and packaging become high density, the wiring on the printed wire board (PWB) becomes wider over 10 Gbit transmission. This paper introduces opto-electronics packaging techniques which aims to solve the problem of optical interconnection dilemma.

A data-format-free 622-Mbit/s/ch 12-channel optical transmitter and receiver have been developed. They provide a skewless PECL interface parallel link for 100-m transmission operating in the ambient temperature range from 0-70/spl deg/C with a 3.3 V power supply.

A novel opto-electronic hybrid integration platform using a flexible resin interposer on a silicon substrate has achieved low-cost, compact-size optical sub-assemblies. A 10 km error-free-transmission has been demonstrated with the optical sub-assemblies for 10-Gigabit Ethernet transceivers.

Eight-channel 1-Gbit/s array optical transmitter and receiver modules with 1.3-/spl mu/m wavelength have been developed using a passive-alignment assembly technology and one-chip transmitter and receiver LSIs, to achieve short-haul optical fiber interconnection cost effectively. It features a large dynamic range receiver that can tolerate the power fluctuation in low-cost adjustment-free transmitters and the loss deviation in cascaded optical connectors.

This paper describes 48 x 100 Mbit/s data link subsystem using a 6 x 1.1 Gbit/s arrayed optical transceiver module and 15-ns deskew function ICs. This link will be applied to the real world computing (RWC)-1, which is one of the massively parallel computer with 1,000-class processor elements.

This paper describes a 3.125 Gbps 4-channel parallel optical transmitter and a receiver module with MT-RJ receptacle designed for a very short reach OC-192 interface between a DWDM transponder and a core router. The discussion is addressed on the following key features: (1) 4-channel 1310 nm FP-LDs with their anodes and cathodes separated channel-by-channel, (2) 4-channel passive optical alignment technology, (3) inter-channel crosstalk reduction technology in a Si optical bench, (4) low power consumption high speed LD driver and receiver LSIs. The applicability of newly developed 3.125 Gbps/ch 4-channel parallel optical transmitter and receiver module to very short reach application has also been demonstrated through a transmission experiment over a 500 m length of GI-50 MMF (graded-index multi mode fiber).

An optical interface is expected to solve the problems the electrical counterpart has. However, the conventional optical interface system using optical transceiver modules separately located from LSI packages on a printed circuit board (PCB), electrical interfaces among optical transceivers and LSI packages would limit interconnection speed. In addition, the electrical interface consumes a lot of power to drive the I/O buffers and signal lines on the PCB. To solve these problems, we have proposed a packaging technology to which OIP (Optical-interconnection as Intellectual Property) concept is applied. We developed a 3.125-Gbit/s/port crosspoint switch with 16 optical I/O ports on a 35x35-mm BGA package. It incorporates four small optical I/O modules called PETIT (Photonic/Electronic Tied InTerface).The PETIT consists of a 4-channel VCSEL array, a 4-channel pin-PD array, and a GaAs trans-impedance amplifier on an 8x8-mm ceramic substrate, and a newly developed optical connector interface is attached on the substrate. The sensitivity is 11.9 dBm under 12.5-Gbit/s full-duplex operation of the PETIT.

Hybrid integration of optical and electrical components (opt-electronic hybrid integration) leads to great progress in low cost and high volume optical packaging. We have developed various types of opt-electronic hybrid integration Tx/Rx module. Recently, demand for transmission speed is increasing up to 10 Gbps in Metro, LAN networks. A new packaging technique is now requested to realize lower cost and high speed. This paper presents the advance of the opt-electronic hybrid integration technology. A new packaging platform is proposed after revealing the present opt-electronic hybrid integration Tx/Rx module, and the prototype 10G-Ether modules are introduced as a trial

This paper describes a sound field reproduction method based on the spatial covariances. Wave surface control and convoluting the head related transfer function are major techniques for sound field reproduction. On the contrary, Takahashi and Tohyama, coauthors of this report, considered that the spatial impressions of a sound field, which we perceive through our ears, are reproduced by preserving the relative relationship between the observation points even if the wave surface is not completely controlled. Moreover, they proposed a new sound field reproduction method that can control the point‐to‐point covariance in a sound field [19th ICA, RBA15‐012]. In this work, we demonstrated about the differences between the covariance based method and the conventional methods, such as stereophonic playback and boundary surface control based on the Kirchhoff–Helmholtz equation. Then, we discussed the performances of several methods on the point of view of reproductivity and device cost.

Signal analysis based on spectral accumulation has great potential for monitoring the condition of structures using random vibration records excited by natural forces. This article describes cumulative harmonic analysis (CHA) by introducing a spectral accumulation function into Berman and Fincham’s conventional cumulative analysis, thus exploring a potential new area in cumulative analysis. CHA effectively visualizes the feedback process from the flow of air against an edge. In this work, we visualized the harmonic starting process for understanding of physical experiment of musical instrument study especially the flute.

Perception of reverberant sound field changes with a sound source distance (SSD). This article describes SSD perception in a coherent region close to the sound source in the reverberation field. We performed listening tests for SSD perception and loudness of speech and carried out transfer functions analysis using a reverberant room. We confirm that both SSD perception and loudness are correlated to the standard deviation of the magnitude frequency response of the transfer function in the coherent region. That is SSD erception and loudness increase as SSD becomes long in the coherent region. However loudness decreases in an incoherent region. Consequently, we surmise SSD perception in the coherent region might be due to increase in loudness.

This paper describes the transfer function (TF) including the coherent sound field as the distance between source and observation points increases, from a point of view of the number of zeros on a complex frequency plane. The propagation phase could be estimated according to the linear‐regression analysis for narrow frequency‐band frequency characteristics of the minimum‐phase component [Inst. Electron. Inf. Commun. Eng. Jpn., J89‐A(4) 291–297 (2006) (in Japanese with English figures)]. This paper analyzes zeros of the TFs for the coherent and reverberant fields based on the modal and random sound field theory. Consequently the distribution of zeros could be estimated in the transition from the coherent to the reverberant conditions.

We have evaluated the exposure latitude and the mask linearity of the lines and spaces (L/S), the isolated lines and the cell patterns used for DRAMs for various sizes down to 0.12 µ m. The exposure latitude was larger than 10% for all types of patterns with sizes larger than 0.12 µ m, using a mask of contrast of 3.7 in a mask-to-wafer gap of 20 µ m. We found that the mask linearity is insufficient and that mask pattern bias is required. The image shortening effect in the cell pattern was also evaluated. The results showed that the value of the shortening is higher when the design rule is smaller, the pattern density is sparser and the mask-to-wafer gap is larger. The value of the shortening in a 0.12-µ m- design- rule cell pattern was 13 nm, which is small enough to be corrected by simply giving a bias in the mask pattern.

This article discusses the resolution of synchrotron radiation lithography in the sub‐100 nm region, taking into consideration the mass production of large‐scale integrated circuits, under attainable conditions for the x‐ray mask, proximity gap, and resist processes. Resolution and exposure latitude for line‐and‐space patterns are markedly improved by using a mask with a contrast of only 2.5. Resolutions of 90, 80, 70, and 60 nm can be achieved with proximity gaps of 30, 20, 15, and 10 μm if a high‐contrast resist and a low‐surface tension developer are used. The latitude will be 10% for pattern sizes as small as 70 nm when the proximity gap is narrower than 15 μm. The effects of mask duty [which is defined to be the ratio of the absorber (line) width to the pattern pitch, i.e., duty cycle] on the optimum exposure dose and mask linearity are also evaluated.

A detailed analysis is performed to investigate the effects of mask contrast, mask-pattern size error and proximity gap on resolution and exposure dose margin for replicating 0.1-µm-region line-and-space (L&S) patterns using synchrotron radiation lithography. The analysis is performed by comparing Fresnel diffraction simulation and pattern replication experiments. When a conventional-contrast (7) mask is used, designing the pattern size of a Ta mask absorber to be slightly narrower than the specified value is a great advantage for faithful pattern replication of 0.1-µm L&S patterns. A large dose margin is obtained for proximity gaps of 15 µm or less. On the other hand, a low-contrast (2.5) mask has an advantage over the conventional mask in that it allows the use of a larger proximity gap for replicating 0.1-µm L&S patterns. A relatively large margin is obtained by limiting the Ta size error to the range of -0.05 µm to +0.03 µm for proximity gaps smaller than 25 µm, and ±0.02 µm for a 30-µm gap.

This paper reports the results of analyzing the proximity effect on the patterning characteristics for plural neighboring hole patterns in synchrotron radiation lithography. Fresnel diffraction simulation was used and pattern replication experiments were performed with pattern pitch, proximity gap, and mask contrast as parameters. Even when the pattern pitch (hole:space) is 1:1, pattern sizes down to 0.2 µ m can be replicated with a large dose margin under a large proximity gap condition up to 40 µ m, irrespective of the mask contrast. A low-contrast (2.5) mask has an advantage over the conventional-contrast (7) mask in that it allows the use of a larger proximity gap when replicating hole patterns with a size of 0.1–0.2 µ m. Moreover, the phase-shifting mask we previously proposed improves the exposure latitude and widens the proximity gap, so that it is possible to use a 20-µ m gap to replicate 0.1-µ m hole patterns for a pitch of 1:1 and to use a 30-µ m gap for a pitch of 1:2.

This paper analyzes the influence of Fresnel diffraction on the patterning characteristics of isolated hole patterns in synchrotron radiation lithography by comparing Fresnel diffraction simulation extended to 2-dimensional patterns and pattern replication experiments with mask contrast as a parameter. It is possible to replicate hole patterns down to 0.1 µm at a large proximity gap of 30 µm by keeping the mask contrast at 2.5-7. However, the exposure latitude for fine patterns below 0.1 µm diminishes abruptly. It also becomes clear that the phase-shifting mask which we previously proposed offers the possibility of replicating ultra-fine hole patterns (

The effects of Fresnel diffraction on the patterning characteristics in synchrotron radiation lithography are greatly influenced by X-ray phase shift in the absorber, especially when a low-contrast X-ray mask is used. We show that the resolution in sub-quarter-micron region can be markedly improved without reducing the proximity gap by utilizing the phase-shifting effects in the absorber. By controlling the absorber thickness within a suitable range (1.5≤mask contrast≤4.0, -30°≥phase shift φ≥-120°), fine patterns as small as 0.1 µm can be replicated with a large exposure latitude at a large proximity gap of 30 µm. Based on these results, we propose a novel X-ray phase-shifting mask with shallowly etched patterns (L in width) at the absorber edge as a phase shifter. This mask can be fabricated by a simple self-alignment process and also offers the possibility of 0.1-µm pattern replication at the 30-µm proximity gap by using 0.02≤L≤0.1 µm.

The applicability of synchrotron radiation x‐ray lithography to future ultralarge scale integrated circuit fabrication processes is demonstrated by the test fabrication of subquartermicron bipolar‐ complementary metaloxide semiconductor devices (SRAM, gate arrays, and several test element groups) with a total size of two‐million transistors. Synchrotron radiation lithography is used at four critical levels: gate poly, first metal, via hole, and second metal. Both negative and positive chemically amplified resists are used with a single‐layer resist system to simplify the resist process. An overview of the lithography process is presented with emphasis on patterning and overlay performance.

To explore the applicability of synchrotron radiation X-ray lithography for fabricating sub-quartermicron devices, we investigate the patterning characteristics of the chemically-amplified negative resist SAL601-ER7. Since these characteristics depend strongly on the conditions of the chemical amplification process, the effects of post-exposure baking and developing conditions on sensitivity and resolution are examined. The resolution-limiting factors are investigated, revealing that pattern collapse during the development process and fog caused by Fresnel diffraction, photo-electron scattering, and acid diffusion in the resist determine the resolution and the maximum aspect ratio of the lines and spaces pattern. Using the model of a swaying beam supported at one end, it is shown that pattern collapse depends on the resist pattern's flexural stiffness. Patterning stability, which depends on the delay time between exposure and baking, is also discussed.

An expression for the configuration integral for three overlapping convex bodies, which is a generalization of Hadwiger-Isihara's formula for two convex bodies, has been found. As an application of this expression, two- and three-molecule cluster integrals (or second and third virial coefficients) for polyatomic molecules in gases are discussed on the basis of a squarewell potential with convex cores.