Une architecture massivement parallèle intégrée

This paper presents a new massively parallel MIMD architecture, halfway between the Connection Machine and hypercubes based on 32-bit processors. It is built from specific 8-bit processors arranged in a 2-D grid and communicating by message transfers. We discuss the communication problems, the instruction set of the basic processing unit, the programmation of the whole array and the use of a high level data-flow language.     

Multihour engineering and traffic routing optimization in hierarchical telephone networks

Classical hierarchical routing in telephone networks is extended to a wider class called out-of-chain routing in such a way that some useful properties of hierarchical routing are retained. This new routing pattern offers more potential paths than the fixed hierarchical one and can be introduced as a dynamic routing where the fixed alternate sequences change at some predetermined instants during the day. The effect of this new routing pattern on the network performances is examined. The main topic of this paper is to present heuristic methods used to optimise such routings in large networks. We show on artificial networks that the throughput of a given network can be significantly improved by suitable routing choices. We demonstrate that the integration of routing changes within a multihour dimensioning process is possible but the lack of realistic data does not permit at this time to quantify the value of routing optimization on real networks.     

RCS reduction of a conducting cone-cylinder using an active method

This paper presents an active method for reducing the radar cross section (rcs) of a perfectly conducting cone-cylinder. The active elements are four microstrip patch antennas located symmetrically along the conecylinder axis. These elements radiate a field which permit the total scattered field in the direction of the receiver to be close to zero. In each time step, the hoarded system should determine some parameters of the incident wave: the incident direction, its frequency, amplitude, and phase. Then when antennas are fed with currents of suitable amplitude and phase, they radiate an electromagnetic field in the direction of the receiver that has the same amplitude but is opposite in phase compared to the scattered field, without feeding patches. Then the total field vanishes in the direction of the receiver.     

石英振梁加速度计：一种精密低成本的小型敏感器

当考虑到标度因数精度要求的时候，振梁加速度计是一种很有吸引力的概念。这种概念基于当梁受到因加速度引起的拉力或压力时，振梁谐振频率发生变化。     

Time-domain surface impedances of lossless Debye media

In a previous paper, a surface impedance formalism was given. Its application to interfaces modelling between homogeneous and frequency dependent media, was of great interest in the finite difference timedomain (fdtd) codes. In this paper, an extension of the method to dispersive media is presented. Applying this formalism to lossless Debye medium, the analytical expressions of the time- domain surface impedances are given. The implementation in a fdtd code permits then a numerical verification of the results in relation to the Fresnel method.     

Speech transmission using rate-compatible trellis codes andembedded source coding

This paper presents bandwidth-efficient speech transmission systems using rate-compatible channel coders and variable bitrate embedded source coders. Rate-compatible punctured convolutional codes (RCPC) are often used to provide unequal error protection (UEP) via progressive bit puncturing. RCPC codes are well suited for constellations for which Euclidean and Hamming distances are equivalent (BPSK and 4-PSK). This paper introduces rate-compatible punctured trellis codes (RCPT) where rate compatibility and UEP are provided via progressive puncturing of symbols in a trellis. RCPT codes constitute a special class of codes designed to maximize residual Euclidean distances (RED) after symbol puncturing. They can be designed for any constellation, allowing for higher throughput than when restricted to using 4-PSK. We apply RCPC and RCPT to two embedded source coders: a perceptual subband coder and the ITU embedded ADPCM G.727 standard. Different operating modes with distinct source/channel bit allocation and UEP are defined. Each mode is optimal for a certain range of AWGN channel SNRs. Performance results using an 8-PSK constellation clearly illustrate the wide range of channel conditions at which the adaptive scheme using RCPT can operate. For an 8-PSK constellation, RCPT codes are compared to RCPC with bit interleaved coded modulation codes (RCPC-BICM). We also compare performance to RCPC codes used with a 4-PSK constellation     

Overview of Cu contamination during integration in a dual damascene architecture for sub-quarter micron technology

A detailed study of copper contaminating steps performed during integration of multilevel Cu metallisation in dual damascene architecture has been performed. Contamination at the wafer back and the bevel edge should make it difficult to use the same equipment for conventional technology and new copper based technology. Several barrier materials have been claimed as efficient against copper diffusion. However, during process integration, contamination issues will be faced before deposition of the barrier layers. Heavy contamination can occur either during Cu chemical mechanical polishing (CMP) or during dielectric etching and via opening on top of contacted copper lines. These residues, concentrated at the dielectric surface, could result in current leakage and shorts between interconnection lines. Several cleaning solutions to remove metal contamination are reviewed and their efficiencies are compared.     

Survey of ICIC techniques in LTE networks under various mobile environment parameters

LTE networks’ main challenge is to efficiently use the available spectrum, and to provide satisfying quality of service for mobile users. However, using the same bandwidth among adjacent cells leads to occurrence of Inter-cell Interference especially at the cell-edge. Basic interference mitigation approaches consider bandwidth partitioning techniques between adjacent cells, such as frequency reuse of factor m schemes, to minimize cell-edge interference. Although SINR values are improved, such techniques lead to significant reduction in the maximum achievable data rate. Several improvements have been proposed to enhance the performance of frequency reuse schemes, where restrictions are made on resource blocks usage, power allocation, or both. Nevertheless, bandwidth partitioning methods still affect the maximum achievable throughput. In this proposal, we intend to perform a comprehensive survey on Inter-Cell Interference Coordination (ICIC) techniques, and we study their performance while putting into consideration various design parameters. This study is implemented throughout intensive system level simulations under several parameters such as different network loads, radio conditions, and user distributions. Simulation results show the advantages and the limitations of each technique compared to frequency reuse-1 model. Thus, we are able to identify the most suitable ICIC technique for each network scenario.     

Myoelectric signal classification for phoneme-based speech recognition

Traditional acoustic speech recognition accuracies have been shown to deteriorate in highly noisy environments. A secondary information source is exploited using surface myoelectric signals (MES) collected from facial articulatory muscles during speech. Words are classified at the phoneme level using a hidden Markov model (HMM) classifier. Acoustic and MES data was collected while the words "zero" through "nine" were spoken. An acoustic expert classified the 18 formative phonemes in low noise levels [signal-to-noise ratio (SNR) of 17.5 dB] with an accuracy of 99%, but deteriorated to approximately 38% under simulations with SNR approaching 0 dB. A fused acoustic-myoelectric multiexpert system, without knowledge of SNR, improved on acoustic classification results at all noise levels. A multiexpert system, incorporating SNR information, obtained accuracies of 99% at low noise levels while maintaining accuracies above 94% during low SNR (0 dB) simulations. Results improve on previous full word MES speech recognition accuracies by almost 10%.