Bell Syst Research Articles

On information, entropy, and early stone tools

Humans affect the biosphere by producing entropy as living beings –as explained by Schröedinger [What is Life? The physical aspect of the living cell (Cambridge University Press, New York, 1992)] – but mostly as a consequence of our ever increasing economic interventions. However, humans also create new forms and arrangements of matter. Is it possible that they carry less entropy than equivalent object subject to natural forces? Art and tech works tell us something, thus carry information. The connection between probability and information was introduced by Shannon 75 years ago [Bell Syst. Tech. J. 27, 379–423 (1948)]. In this work we calculate the probability that a stone tool could have been constructed by random, natural causes. We do so by considering the earliest tools created by our hominin ancestors, the stone tools of the Paleolithic. By analysing a sample of tools, we show that they are highly unlikely constructs by random means, and that therefore carry a great deal of information, as generations of archaeologists and anthropologists have eloquently exhibited. Further, we quantify this information, and following Shannon, we ascribe an entropy to the stone tools. We thus find that the stone tool carries less entropy than a stone hit by dispersed blows. This work presents a first simplified approach to these important issues.

Beyond the Csiszár–Körner Bound: Best-Possible Wiretap Coding via Obfuscation

A wiretap coding scheme (Wyner in Bell Syst Tech J 54(8):1355–1387, 1975) enables Alice to reliably communicate a message m to an honest Bob by sending an encoding c over a noisy channel ChB\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extsf{ChB}$$\\end{document}, while at the same time hiding m from Eve who receives c over another noisy channel ChE\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extsf{ChE}$$\\end{document}. Wiretap coding is clearly impossible when ChB\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extsf{ChB}$$\\end{document} is a degraded version of ChE\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extsf{ChE}$$\\end{document}, in the sense that the output of ChB\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extsf{ChB}$$\\end{document} can be simulated using only the output of ChE\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extsf{ChE}$$\\end{document}. A classic work of Csiszár and Korner (IEEE Trans Inf Theory 24(3):339–348, 1978) shows that the converse does not hold. This follows from their full characterization of the channel pairs (ChB,ChE)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$(\ extsf{ChB},\ extsf{ChE})$$\\end{document} that enable information-theoretic wiretap coding. In this work, we show that in fact the converse does hold when considering computational security; that is, wiretap coding against a computationally bounded Eve is possible if and only ifChB\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extsf{ChB}$$\\end{document} is not a degraded version of ChE\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extsf{ChE}$$\\end{document}. Our construction assumes the existence of virtual black-box obfuscation of specific classes of “evasive” functions that generalize fuzzy point functions and can be heuristically instantiated using indistinguishability obfuscation. Finally, our solution has the appealing feature of being universal in the sense that Alice’s algorithm depends only on ChB\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extsf{ChB}$$\\end{document} and not on ChE\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extsf{ChE}$$\\end{document}.

Makespan minimization with OR-precedence constraints

We consider a variant of the NP-hard problem of assigning jobs to machines to minimize the completion time of the last job. Usually, precedence constraints are given by a partial order on the set of jobs, and each job requires all its predecessors to be completed before it can start. In this paper, we consider a different type of precedence relation that has not been discussed as extensively and is called OR-precedence. In order for a job to start, we require that at least one of its predecessors is completed—in contrast to all its predecessors. Additionally, we assume that each job has a release date before which it must not start. We prove that a simple List Scheduling algorithm due to Graham (Bell Syst Tech J 45(9):1563–1581, 1966) has an approximation guarantee of 2 and show that obtaining an approximation factor of 4/3 - varepsilon is NP-hard. Further, we present a polynomial-time algorithm that solves the problem to optimality if preemptions are allowed. The latter result is in contrast to classical precedence constraints where the preemptive variant is already NP-hard. Our algorithm generalizes previous results for unit processing time jobs subject to OR-precedence constraints, but without release dates. The running time of our algorithm is O(n^2) for arbitrary processing times and it can be reduced to O(n) for unit processing times, where n is the number of jobs. The performance guarantees presented here match the best-known ones for special cases where classical precedence constraints and OR-precedence constraints coincide.

О ПРИМЕНИМОСТИ ФАЗОВОГО МЕТОДА СКАНИРОВАНИЯ В АНТЕННЫХ РЕШЕТКАХ ИМПУЛЬСНЫХ ЛЧМ-РАДАРОВ

О ПРИМЕНИМОСТИ ФАЗОВОГО МЕТОДА СКАНИРОВАНИЯ В АНТЕННЫХ РЕШЕТКАХ ИМПУЛЬСНЫХ ЛЧМ-РАДАРОВ

Rational over-specification in visually-situated comprehension and production

Contrary to the Gricean maxims of quantity (Grice, in: Cole, Morgan (eds) Syntax and semantics: speech acts, vol III, pp 41–58, Academic Press, New York, 1975), it has been repeatedly shown that speakers often include redundant information in their utterances (over-specifications). Previous research on referential communication has long debated whether this redundancy is the result of speaker-internal or addressee-oriented processes, while it is also unclear whether referential redundancy hinders or facilitates comprehension. We present an information-theoretic explanation for the use of over-specification in visually-situated communication, which quantifies the amount of uncertainty regarding the referent as entropy (Shannon in Bell Syst Tech J 5:10, https://doi.org/10.1002/j.1538-7305.1948.tb01338.x, 1948). Examining both the comprehension and production of over-specifications, we present evidence that (a) listeners’ processing is facilitated by the use of redundancy as well as by a greater reduction of uncertainty early on in the utterance, and (b) that at least for some speakers, listeners’ processing concerns influence their encoding of over-specifications: Speakers were more likely to use redundant adjectives when these adjectives reduced entropy to a higher degree than adjectives necessary for target identification.

(Invited) Investigating the Electronic Properties of Narrow Band Gap III-V Mos Systems

The electronic properties of the III-V/oxide MOS system play a central role in applications of compound semiconductors in RF, sensor and photonics based applications. The technique commonly used to probe the electronic properties of oxide/semiconductor interfaces in MOS structures is impedance spectroscopy, where the capacitance and conductance of the MOS system is measured as a function of bias, frequency and temperature [1, 2]. There is a growing interest in the integration of RF/mm wave [3] and photonic [4] components onto a large area silicon platform, to save on the overall use of III-V materials and also to allow more compact system integration. Considering the case of RF and mm wave devices in a MOSFET implementation, there is a need to understand and control electrically active defects at the interface between the III-V channel and the insulating gate oxide. In recent years there has been considerable progress in reducing the interface state density (Dit) [5-8], in narrow gap InGaAs and InGaSb [9] MOS structures to the point where genuine surface inversion can be observed for both n- and p-type InGaAs [8,9]. The research focus for III-V MOS systems is now turning to the issue of defects states in the oxide (border traps) which can communicate with the conduction and valence bands by elastic or inelastic tunnelling [10]. Analysis of these border traps to date has been based on a distributed capacitor/resistor model [10, 11]. In this presentation, we extend these studies by simulating the AC response of border traps (BTs) and interface states by employing a simulator [12, 13] that solves the Poisson, drift-diffusion and charge conservation equations, taking into account the elastic band-to-trap tunneling model [14] for BTs, yielding estimates for the trap volume as well as variation in border trap density with energy and depth into the oxide. The presentation will also provide an overview of the relationship between specific functions of the capacitance (C) conductance (G) for the case of narrow band gap III-V MOS structures which exhibit genuine surface inversion, where the capacitance and conductance of the MOS system as a function of ac angular frequency (ω) are related, and in particular the peak values of G/ω and −dC/dloge(ω) (≡− ωdC/dω) are equal, and that these equal peak magnitudes occur at the same value of ω [15]. The relationship is also confirmed by physics based ac simulations of MOS structures and through analysis of the equivalent circuit model in inversion. Results will be presented for InGaAs and InGaSb MOS structures (Al2O3 and Al2O3/HfO2 ALD oxides) where genuine inversion of the III-V/oxide surface is confirmed by the G/ω and −dC/dloge(ω) functions, which have peak values of Cox 2/2(Cox+Cd) when the surface is inverted (where Cox is the gate oxide capacitance and Cd is the maximum capacitance of the semiconductor surface in inversion). Extending on previous studies [15], this presentation will presents experimental and physics based simulations which indicate the relationship between G/ω and −dC/dloge(ω) holds true in all bias regions of the MOS response. Finally, it is noted that the relation between G/ω and −dC/dloge(ω) is not specific to III-V MOS systems, and is generally true for all MOS systems, and examples will be provides for MOS structures based on a 2 D dimensional semiconductor. All authors acknowledge the financial support of the EU through the INSIGHT project (688784). The authors from Tyndall acknowledge Science Foundation Ireland (15/IA/3131). [1] E. H. Nicollian and J. R. Brews, MOS Physics and Technology. Wiley, 1982 [2] E. H. Nicollian and A. Goetzberger, Bell Syst. Tech. J., 46, 6, pp. 1055–1133, 1967. [3] C. Zota, et. al., IEEE Electron Dev. Lett. vol 37, p1264 (2016) [4] S. Chen, et al., Optics Express Vol. 25, Issue 5, pp. 4632-4639 (2017) [5] É. O’Connor et al., J. Appl. Phys., vol. 109, no. 2, p. 024101, 2011. [6] H.-D. Trinh et al., Appl. Phys. Lett., vol. 97, no. 4, pp. 042903-1–042903-3, 2010. [7] T. D. Lin et al., Appl. Phys. Lett., vol. 100, no. 17, p. 172110, 2012. [8] E. O'Connor, et al. Appl. Phys. Lett 110, 032902 (2017) [9] D. Millar et al., Applied Materials & Interfaces (submitted) [10] Y. Yuan et al., IEEE EDL, 32, 485 (2011) [11] G. Sereni et al., IEEE TED 62, 705, (2015) [12] Synposys Inc., Sentaurus Device TM, v. L-2016.03-SP2 [13] S.E. Laux, IEEE Trans. on Computer-Aided Design of Integrated Circuits and Systems, 4, 478 (1985) [14] F. Jiménez-Molinos et al., JAP, 91 (2002), 5116 [15] Scott Monaghan et al., IEEE Transaction on Electron Devices, 61, 4176 (2014)

Weighted pooling for image recognition of deep convolutional neural networks

There are some traditional pooling methods in convolutional neural network, such as max-pooling, average pooling, stochastic pooling and so on, which determine the results of pooling based on the distribution of each activation in the pooling region. Zeiler and Fergus (Stochastic-pooling for regularization of deep convolutional neural networks, 2013) However, it is difficult for the feature mapping process to select a perfect activation representative of the pooling region, and can lead to the phenomenon of over-fitting. In this paper, the following theoretical basis comes out information theory (Shannon in Bell Syst. Tech. J. 27:379–423, 1948). First, we quantify the information entropy of each pooling region, and then propose an efficient pooling method by comparing the mutual information between activations and the pooling region which they are located in. Moreover, we assign different weights to different activations based on mutual information, and named it weighted-pooling. The main features of the weighted-pooling method are as follows: (1) The information quantity of the pooling region is quantified by information theory for the first time. (2) Also, each activation’s contribution was quantified for the first time and these contributions eliminate the uncertainty of the pooling region which it is located in. (3) For choosing a representative in this pooling region, the weight of each activation obviously superiors to the value of activation. In the experimental part, we respectively use MNIST and CIFAR-10 (Krizhevsky in Learning multiple layers of featurs from tiny images, University of Toronto, 2009; LeCun in The MNIST database, 2012) data sets to compare different pooling methods. The results show that the weighted-pooling method has higher recognition accuracy than other pooling methods and reaches a new state-of-the-art.

(Invited) The Inversion Behaviour of Narrow Band Gap Mos Systems: Experimental Observations, Physics Based Simulations and Applications

Impedance spectroscopy of the metal-oxide semiconductor (MOS) system has played a central role in the development of silicon-based complementary MOS (CMOS) technology over the past 50 years [1, 2]. With current research interest into alternative semiconductor channels to silicon for MOSFET and tunnel FET technologies, the measurement and interpretation of the overall impedance of the MOS structure requires detailed analysis to separate and quantify the contribution of interface states, and near interface traps (border traps), on the capacitance and conductance response, and to separate the contribution of these electrically active defect states from the ac response of minority carriers in the case of genuine inversion of the semiconductor/dielectric interface. There has been considerable progress in recent years in reducing the interface state density (Dit) [3], [4], [5], in narrow gap InxGa1-xAs MOS structures to the point where genuine surface inversion can be observed for both n- and p-type In0.53Ga0.47As MOS capacitors, which is confirmed by analysis of the minimum capacitance of n- and p-type In0.53Ga0.47As MOS structure with doping concentration ranging from approximately 1x1016 to 1x1018 cm-3 [6]. In this presentation we will provide an overview of the experimental relationship between specific functions of the capacitance (C) and conductance (G) for the case of narrow band gap III-V MOS structures which exhibit genuine surface inversion, where the capacitance and conductance of the MOS system as a function of ac angular frequency (ω) are related, and in particular, the peak values of G/ω and −dC/dloge(ω) (≡− ωdC/dω) are equal, and that these peak magnitudes occur at the same value of ω [7]. The relationship is also confirmed by physics based ac simulations of MOS structures and through analysis of the equivalent circuit model in inversion. Results will be presented for InGaAs and InGaSb MOS structures (Al2O3 and Al2O3/HfO2 ALD oxides) where genuine inversion of the III-V/oxide surface is confirmed by the G/ω and −dC/dloge(ω) functions, which have peak values of Cox 2/2(Cox+Cd) when the surface is inverted (where Cox is the gate oxide capacitance and Cd is the maximum capacitance of the semiconductor surface in inversion). In the case of p type InGaSb MOS structures it is also notable that the accumulation frequency dispersion is very low (<0.7%/decade) indicating a reduced density of border traps at energies in the Al2O3 gate oxide aligning with the In0.3Ga0.7Sb valence band edge, which is approximately aligned with the lowest energy in the In0.53Ga0.47As conduction band gamma valley. Finally, experimental results and physics based simulations will be presented, which indicate that the peak values of G/ω and −dC/dloge(ω) in inversion for any MOS system which demonstrates an inversion response within the typically range of temperatures and ac signal frequencies employed in experiments, opens a route to determine the gate oxide capacitance in inversion (where the gate oxide field and gate leakage are reduced), and that the angular frequency associated with the peak values of the G/ω and −dC/dloge(ω) functions allows the determination of the minority carrier generation rate in the semiconductor, which is relevant to the leakage currents in MOSFETs and the optical performance in photonic applications. [1] E. H. Nicollian and J. R. Brews, MOS Physics and Technology. New York, NY, USA: Wiley, 1982 [2] E. H. Nicollian and A. Goetzberger, “The Si-SiO2 interface—Electrical properties as determined by the metal-insulator-silicon conductance technique,” Bell Syst. Tech. J., vol. 46, no. 6, pp. 1055–1133, 1967. [3] É. O’Connor et al., J. Appl. Phys., vol. 109, no. 2, p. 024101, 2011. [4] H.-D. Trinh et al., Appl. Phys. Lett., vol. 97, no. 4, pp. 042903-1–042903-3, 2010. [5] T. D. Lin et al., “Realization of high-quality HfO2 on In0.53Ga0.47As by in-situ atomic-layer-deposition,” Appl. Phys. Lett., vol. 100, no. 17, p. 172110, 2012. [6] E. O'Connor, K. Cherkaoui, S. Monaghan, B. Sheehan, I. M. Povey, and P. K. Hurley, Appl. Phys. Lett 110, 032902 (2017) [7] Scott Monaghan, Éamon O’Connor, Rafael Rios, Fahmida Ferdousi, Liam Floyd, Eimear Ryan, Karim Cherkaoui, Ian M. Povey, Kelin J. Kuhn, and Paul K. Hurley, IEEE Transaction on Electron Devices, 61, 4176 (2014)

Information and Self-Organization

The process of “self-organization” takes place in open and complex systems that acquire spatio-temporal or functional structures without specific ordering instructions from the outside. [...]

A new structural approach to genomic discovery of disease: example of adult-onset diabetes

This paper reports on an investigation of disease discovery from genomic data, by methods which depart substantially from customary practices found in the investigation of genome-wide association studies. Such data in general are composed of the genomic content from two contrasting phenotypes, e.g., disease versus control populations, and the analysis proceeds under the hypothesis that populational dissimilarities might reveal disease risk alleles. The proposed suite of new methods is in part based on information theory (Shannon in Bell Syst Tech J 27:379-423, 1948a; Bell Syst Tech J 27:623-656, 1948b; Jaynes in Phys Rev 106:620-630, 1957), and strong evidence will be given of the effectiveness of this new approach. The methodology extends naturally and successfully to predicting genomic disposition to disease arising from large collections of weakly contributing genomic loci. Evidence will be advanced that the example of adult-onset diabetes ("type 2 diabetes") is such a candidate disease, and in this case, probably for the first time, it can be demonstrated that disease prediction is possible. Another novel element of this study is the search and identification of potential beneficial genomic loci that may counter a disease. The generality of the methodology suggests that it might extend to other diseases.

Codes from multipartite graphs and minimal permutation decoding sets

Permutation decoding method developed by MacWilliams and described in [Permutation decoding of systematic codes, Bell Syst. Tech. J. 43 (1964) 485–505] is a decoding technique that uses a subset of the automorphism group of the code called a PD-set. The complexity of the permutation decoding algorithm depends on the size of the PD-set and finding a minimal PD-set for an error correcting code is a hard problem. In this paper we examine binary codes from the complete-multipartite graph [Formula: see text] and find PD-sets for all values of [Formula: see text] and [Formula: see text]. Further we show that these PD-sets are minimal when [Formula: see text] is odd and [Formula: see text].

Existence of bounded discrete steady state solutions of the van Roosbroeck system with monotone Fermi---Dirac statistic functions

If in the classic van Roosbroeck system (Bell Syst Tech J 29:560---607, 1950) the statistic function is modified, the equations can be derived by a variational formulation or just using a generaliz...

Existence of bounded discrete steady state solutions of the van Roosbroeck system with monotone Fermi–Dirac statistic functions

If in the classic van Roosbroeck system (Bell Syst Tech J 29:560---607, 1950) the statistic function is modified, the equations can be derived by a variational formulation or just using a generalized Einstein relation. In both cases a dissipative generalization of the Scharfetter---Gummel scheme (IEEE Trans Electr Dev 16, 64---77, 1969), understood as a one-dimensional constant current approximation, is derived for strictly monotone coefficient functions in the elliptic operator $$\nabla \cdot { {f}(v)} \nabla $$?·f(v)?, v chemical potential, while the hole density is defined by $$p={\mathcal {F}}(v)\le e^v.$$p=F(v)≤ev. A closed form integration of the governing equation would simplify the practical use, but mean value theorem based results are sufficient to prove existence of bounded discrete steady state solutions on any boundary conforming Delaunay grid. These results hold for any piecewise, continuous, and monotone approximation of $${ {f}(v)}$$f(v) and $${\mathcal {F}}(v)$$F(v). Hence an implementation based on this discretization will inherit the same stability properties as the Boltzmann case based on the Scharfetter---Gummel scheme. Large chemical potentials and and related degeneracy effects in semiconductors can be approximated. A proven, stability focused blueprint for the discretization of a fairly general, steady state Fermi---Dirac like drift---diffusion setting for semiconductors using mainly new results to extend classic ideas is the main goal.

Oxide Structure-Dependent Interfacial Layer Defects of HfAlO/SiO2/Si Stack Analyzed by Conductance Method

Aggressive downscaling of effective oxide thickness (EOT) requires improvement in gate leakage current and interfacial characteristics. Recently other materials like Al, Zr, and La are introduced into HfO2 to further reduce the EOT and improve the reliability. Since the structure of the HfO2-based dielectric changes with addition such materials it is required that the impact of structural change on the interface needs to be investigated. In addition, the effect series resistance（R s）associated with the substrate and contact needs to be excluded. In this work we have used metal oxide semiconductor capacitor (MOS-C) devices of 3nm HfAlO/ 0.5nm SiO2/Si gate stack by different preparation methods (different Al doping, post annealing temperatures, deposition stacks) to characterize the quality of interface. Interface defect densities were evaluated by conductance method (1) while excluding any series resistance effect. 13 samples were prepared with 5 angstrom of interfacial layer SiOx, intentionally deposited on the Si substrate by vapor ozone treatment prior to depositing the high-k layer by atomic layer deposition (ALD) in a 300 mm TEL Trias™ cleanroom tool at TEL Technology Center. The metal gate was prepared by 5nm ALD TiN followed by 50nm physical vapor deposition (PVD) TiN. The substrate was of p-type Si (001) orientation with a.doping concentration of approximately 1015cm-3. Small signal capacitance and conductance measurements were investigated by HP4284 LCR analyzer. It is known that depending on the post-deposition annealing (PDA) temperatures Al doped HfO2 changes its structure from amorphous to tetragonal reducing its EOT (2). It was observed that D it is dependent on the EOT of hafnium aluminum oxide. The amorphous structure has the lowest D it (2.76) whereas tetragonal HfO2 (t-HfO2) has the highest D it (1.27). The Dit values of other EOT are within the range of observed highest and lowest values as EOT, as shown in Fig.1. It is expected that D it value are affected both by the oxide phase structure and interfacial layer thickness since Aluminum doping could impede the growth of the interfacial layer. The decrease in interfacial layer will significantly impact the interface defects as they are closer to interface, this was observed when annealing temperature was 800°C (3). However, in the low annealing temperature with addition of Al the dielectric constant of oxide decreases insignificantly to changed EOT without reducing Dit . E. H. Nicollian and A. Goetzberger, Bell Syst. Tech. J. 46, 1055 (1967).K. Tapily, S. Consiglio, R. Clark, R. Vasic, C. Wajda, J. Jordan-Sweet, G. Leusink, and A.C Diebold, ECS Trans., 64 (9), 123 (2014).Y.Sugimotoa, H.Adachia, K.Yamamotoa, D.Wangb,H. Nakashimaa, H. Nakashima, Material Sci. in Semicon. Process.,9, 1031 (2006). Figure 1

Entropy, complexity, and spatial information.

We pose the central problem of defining a measure of complexity, specifically for spatial systems in general, city systems in particular. The measures we adopt are based on Shannon’s (in Bell Syst Tech J 27:379–423, 623–656, 1948) definition of information. We introduce this measure and argue that increasing information is equivalent to increasing complexity, and we show that for spatial distributions, this involves a trade-off between the density of the distribution and the number of events that characterize it; as cities get bigger and are characterized by more events—more places or locations, information increases, all other things being equal. But sometimes the distribution changes at a faster rate than the number of events and thus information can decrease even if a city grows. We develop these ideas using various information measures. We first demonstrate their applicability to various distributions of population in London over the last 100 years, then to a wider region of London which is divided into bands of zones at increasing distances from the core, and finally to the evolution of the street system that characterizes the built-up area of London from 1786 to the present day. We conclude by arguing that we need to relate these measures to other measures of complexity, to choose a wider array of examples, and to extend the analysis to two-dimensional spatial systems.

Real-time market microstructure analysis: online transaction cost analysis

Motivated by the practical challenge in monitoring the performance of a large number of algorithmic trading orders, this paper provides a methodology that leads to automatic discovery of causes that lie behind poor trading performance. It also gives theoretical foundations to a generic framework for real-time trading analysis. The common acronym for investigating the causes of bad and good performance of trading is transaction cost analysis Rosenthal [Performance Metrics for Algorithmic Traders, 2009]). Automated algorithms take care of most of the traded flows on electronic markets (more than 70% in the US, 45% in Europe and 35% in Japan in 2012). Academic literature provides different ways to formalize these algorithms and show how optimal they can be from a mean-variance (like in Almgren and Chriss [J. Risk, 2000, 3(2), 5–39]), a stochastic control (e.g. Guéant et al. [Math. Financ. Econ., 2013, 7(4), 477–507]), an impulse control (see Bouchard et al. [SIAM J. Financ. Math., 2011, 2(1), 404–438]) or a statistical learning (as used in Laruelle et al. [Math. Financ. Econ., 2013, 7(3), 359–403]) viewpoint. This paper is agnostic about the way the algorithm has been built and provides a theoretical formalism to identify in real-time the market conditions that influenced its efficiency or inefficiency. For a given set of characteristics describing the market context, selected by a practitioner, we first show how a set of additional derived explanatory factors, called anomaly detectors, can be created for each market order (following for instance Cristianini and Shawe-Taylor [An Introduction to Support Vector Machines and Other Kernel-based Learning Methods, 2000]). We then will present an online methodology to quantify how this extended set of factors, at any given time, predicts (i.e. have influence, in the sense of predictive power or information defined in Basseville and Nikiforov [Detection of Abrupt Changes: Theory and Application, 1993], Shannon [Bell Syst. Tech. J., 1948, 27, 379–423] and Alkoot and Kittler [Pattern Recogn. Lett., 1999, 20(11), 1361–1369]) which of the orders are underperforming while calculating the predictive power of this explanatory factor set. Armed with this information, which we call influence analysis, we intend to empower the order monitoring user to take appropriate action on any affected orders by re-calibrating the trading algorithms working the order through new parameters, pausing their execution or taking over more direct trading control. Also we intend that use of this method can be taken advantage of to automatically adjust their trading action in the post trade analysis of algorithms.

Entropy, color, and color rendering

The Shannon entropy [Bell Syst. Tech J.27, 379 (1948)] of spectral distributions is applied to the problem of color rendering. With this novel approach, calculations for visual white entropy, spectral entropy, and color rendering are proposed, indices that are unreliant on the subjectivity inherent in reference spectra and color samples. The indices are tested against real lamp spectra, showing a simple and robust system for color rendering assessment. The discussion considers potential roles for white entropy in several areas of color theory and psychophysics and nonextensive entropy generalizations of the entropy indices in mathematical color spaces.

The invention and early history of the CCD

As the first practical solid state imaging device, the invention of the charge coupled device has profoundly affected image sensing technology. They are used in a wide range of applications both as area and linear imaging devices starting with the replacement of imaging tubes used in commercial TV cameras and cam-corders. The rapid rise of their use in digital cameras has initiated the demise of film photography and created vast new markets with great economic benefit for many. Other uses include a wide variety of scientific, surveillance, and scanning applications. The inception of the device at Bell Labs by W. S. Boyle and G. E. Smith, Bell Syst. Tech. J. 49, 587 (1970);G. F. Amelio, M. F. Tompsett, and G. E. Smith, Bell Syst. Tech. J. 49, 593 (1970); W. S. Boyle and G. E. Smith, U.S. patent 3,792,322 (12 February 1974) was strongly influenced by several unique factors existing both within Bell Labs and the current world state of technology. These factors and their relevance will be discussed along with the train of thought leading to the invention. Early experimental devices and their initial applications were vigorously pursued and will be described. Mention of current applications will be given.

Variant of the method of Fox and Li dedicated to intracavity laser beam shaping

We present a variant of the method of Fox and Li [Bell Syst. Tech. J. 40, 453 (1961); Proc. IEEE 51, 80 (1963)] dedicated to intracavity laser beam shaping for resonators containing an arbitrary number of amplitude and phase diffractive optics. Contrary to Fox and Li, the starting point is the desired field. The latter is injected into the usual sequence of lenses representing just a single round trip, and the optimization process iterates until the input and the output fields match as much as possible. We illustrate this technique by deriving a simple model for generating single cylindrical TEM(p0) modes, thanks to a π-phase plate placed inside a plano-concave cavity. The experimental validation attests an excellent agreement with numerical predictions.

On the mixing properties of operations in a finite field

In the context of the approach given in Shannon's research [Bell Syst. Tech. J. 28: 656–715, 1949], of interest are the mixing properties of operations in a finite field, in particular, the action of each operation on cosets with respect to the action of the second operation. The good mixing properties give a possibility to construct on this base good cryptographic algorithms.