GLOBAL REGISTRATION OF MULTIPLE POINT CLOUDS USING SEMIDEFINITE PROGRAMMING

Consider N points in R-d and M local coordinate systems that are related through unknown rigid transforms. For each point, we are given (possibly noisy) measurements of its local coordinates in some of the coordinate systems. Alternatively, for each coordinate system, we observe the coordinates of a subset of the points. The problem of estimating the global coordinates of the N points (up to a rigid transform) from such measurements comes up in distributed approaches to molecular conformation and sensor network localization, and also in computer vision and graphics. The least-squares formulation of this problem, although nonconvex, has a well-known closed-form solution when M = 2 (based on the singular value decomposition (SVD)). However, no closed-form solution is known for M >= 3. In this paper, we demonstrate how the least-squares formulation can be relaxed into a convex program, namely, a semidefinite program (SDP). By setting up connections between the uniqueness of this SDP and results from rigidity theory, we prove conditions for exact and stable recovery for the SDP relaxation. In particular, we prove that the SDP relaxation can guarantee recovery under more adversarial conditions compared to earlier proposed spectral relaxations, and we derive error bounds for the registration error incurred by the SDP relaxation. We also present results of numerical experiments on simulated data to confirm the theoretical findings. We empirically demonstrate that (a) unlike the spectral relaxation, the relaxation gap is mostly zero for the SDP (i.e., we are able to solve the original nonconvex least-squares problem) up to a certain noise threshold, and (b) the SDP performs significantly better than spectral and manifold-optimization methods, particularly at large noise levels.

THREE-LAYER FLUID FLOW OVER A SMALL OBSTRUCTION ON THE BOTTOM OF A CHANNEL

Many boundary value problems occur in a natural way while studying fluid flow problems in a channel. The solutions of two such boundary value problems are obtained and analysed in the context of flow problems involving three layers of fluids of different constant densities in a channel, associated with an impermeable bottom that has a small undulation. The top surface of the channel is either bounded by a rigid lid or free to the atmosphere. The fluid in each layer is assumed to be inviscid and incompressible, and the flow is irrotational and two-dimensional. Only waves that are stationary with respect to the bottom profile are considered in this paper. The effect of surface tension is neglected. In the process of obtaining solutions for both the problems, regular perturbation analysis along with a Fourier transform technique is employed to derive the first-order corrections of some important physical quantities. Two types of bottom topography, such as concave and convex, are considered to derive the profiles of the interfaces. We observe that the profiles are oscillatory in nature, representing waves of variable amplitude with distinct wave numbers propagating downstream and with no wave upstream. The observations are presented in tabular and graphical forms.

Optimal power allocation for protective jamming in wireless networks: A flow based model

We address the problem of passive eavesdroppers in multi-hop wireless networks using the technique of friendly jamming. The network is assumed to employ Decode and Forward (DF) relaying. Assuming the availability of perfect channel state information (CSI) of legitimate nodes and eavesdroppers, we consider a scheduling and power allocation (PA) problem for a multiple-source multiple-sink scenario so that eavesdroppers are jammed, and source-destination throughput targets are met while minimizing the overall transmitted power. We propose activation sets (AS-es) for scheduling, and formulate an optimization problem for PA. Several methods for finding AS-es are discussed and compared. We present an approximate linear program for the original nonlinear, non-convex PA optimization problem, and argue that under certain conditions, both the formulations produce identical results. In the absence of eavesdroppers' CSI, we utilize the notion of Vulnerability Region (VR), and formulate an optimization problem with the objective of minimizing the VR. Our results show that the proposed solution can achieve power-efficient operation while defeating eavesdroppers and achieving desired source-destination throughputs simultaneously. (C) 2015 Elsevier B.V. All rights reserved.

The Self-Association of Graphane Is Driven by London Dispersion and Enhanced Orbital Interactions

We investigated the nature of the cohesive energy between graphane sheets via multiple CH center dot center dot center dot HC interactions, using density functional theory (DFT) including dispersion correction (Grimmes D3 approach) computations of n]graphane sigma dimers (n = 6-73). For comparison, we also evaluated the binding between graphene sheets that display prototypical pi/pi interactions. The results were analyzed using the block-localized wave function (BLW) method, which is a variant of ab initio valence bond (VB) theory. BLW interprets the intermolecular interactions in terms of frozen interaction energy (Delta E-F) composed of electrostatic and Pauli repulsion interactions, polarization (Delta E-pol), charge-transfer interaction (Delta E-CT), and dispersion effects (Delta E-disp). The BLW analysis reveals that the cohesive energy between graphane sheets is dominated by two stabilizing effects, namely intermolecular London dispersion and two-way charge transfer energy due to the sigma CH -> sigma*(HC) interactions. The shift of the electron density around the nonpolar covalent C-H bonds involved in the intermolecular interaction decreases the C-H bond lengths uniformly by 0.001 angstrom. The Delta E-CT term, which accounts for similar to 15% of the total binding energy, results in the accumulation of electron density in the interface area between two layers. This accumulated electron density thus acts as an electronic glue for the graphane layers and constitutes an important driving force in the self-association and stability of graphane under ambient conditions. Similarly, the double faced adhesive tape style of charge transfer interactions was also observed among graphene sheets in which it accounts for similar to 18% of the total binding energy. The binding energy between graphane sheets is additive and can be expressed as a sum of CH center dot center dot center dot HC interactions, or as a function of the number of C-H bonds.

Making risk minimization tolerant to label noise

In many applications, the training data, from which one needs to learn a classifier, is corrupted with label noise. Many standard algorithms such as SVM perform poorly in the presence of label noise. In this paper we investigate the robustness of risk minimization to label noise. We prove a sufficient condition on a loss function for the risk minimization under that loss to be tolerant to uniform label noise. We show that the 0-1 loss, sigmoid loss, ramp loss and probit loss satisfy this condition though none of the standard convex loss functions satisfy it. We also prove that, by choosing a sufficiently large value of a parameter in the loss function, the sigmoid loss, ramp loss and probit loss can be made tolerant to nonuniform label noise also if we can assume the classes to be separable under noise-free data distribution. Through extensive empirical studies, we show that risk minimization under the 0-1 loss, the sigmoid loss and the ramp loss has much better robustness to label noise when compared to the SVM algorithm. (C) 2015 Elsevier B.V. All rights reserved.

The tarani mutation alters surface curvature in Arabidopsis leaves by perturbing the patterns of surface expansion and cell division

The leaf surface usually stays flat, maintained by coordinated growth. Growth perturbation can introduce overall surface curvature, which can be negative, giving a saddle-shaped leaf, or positive, giving a cup-like leaf. Little is known about the molecular mechanisms that underlie leaf flatness, primarily because only a few mutants with altered surface curvature have been isolated and studied. Characterization of mutants of the CINCINNATA-like TCP genes in Antirrhinum and Arabidopsis have revealed that their products help maintain flatness by balancing the pattern of cell proliferation and surface expansion between the margin and the central zone during leaf morphogenesis. On the other hand, deletion of two homologous PEAPOD genes causes cup-shaped leaves in Arabidopsis due to excess division of dispersed meristemoid cells. Here, we report the isolation and characterization of an Arabidopsis mutant, tarani (tni), with enlarged, cup-shaped leaves. Morphometric analyses showed that the positive curvature of the tni leaf is linked to excess growth at the centre compared to the margin. By monitoring the dynamic pattern of CYCLIN D3;2 expression, we show that the shape of the primary arrest front is strongly convex in growing tni leaves, leading to excess mitotic expansion synchronized with excess cell proliferation at the centre. Reduction of cell proliferation and of endogenous gibberellic acid levels rescued the tni phenotype. Genetic interactions demonstrated that TNI maintains leaf flatness independent of TCPs and PEAPODs.

A 0.5-2.0 GHz injection locked oscillator cascade for parallel wideband RF spectrum sensing

An area-efficient, wideband RF frequency synthesizer, which simultaneously generates multiple local oscillator (LO) signals, is designed. It is suitable for parallel wideband RF spectrum sensing in cognitive radios. The frequency synthesizer consists of an injection locked oscillator cascade (ILOC) where all the LO signals are derived from a single reference oscillator. The ILOC is implemented in a 130-nm technology with an active area of . It generates 4 uniformly spaced LO carrier frequencies from 500 MHz to 2 GHz. This design is the first known implementation of a CMOS based ILOC for wide-band RF spectrum sensing applications.

Novel Biobjective Clustering (BiGC) Based on Cooperative Game Theory

We propose a new approach to clustering. Our idea is to map cluster formation to coalition formation in cooperative games, and to use the Shapley value of the patterns to identify clusters and cluster representatives. We show that the underlying game is convex and this leads to an efficient biobjective clustering algorithm that we call BiGC. The algorithm yields high-quality clustering with respect to average point-to-center distance (potential) as well as average intracluster point-to-point distance (scatter). We demonstrate the superiority of BiGC over state-of-the-art clustering algorithms (including the center based and the multiobjective techniques) through a detailed experimentation using standard cluster validity criteria on several benchmark data sets. We also show that BiGC satisfies key clustering properties such as order independence, scale invariance, and richness.

Two player game variant of the Erdos-Szekeres problem

The classical Erdos-Szekeres theorem states that a convex k-gon exists in every sufficiently large point set. This problem has been well studied and finding tight asymptotic bounds is considered a challenging open problem. Several variants of the Erdos-Szekeres problem have been posed and studied in the last two decades. The well studied variants include the empty convex k-gon problem, convex k-gon with specified number of interior points and the chromatic variant. In this paper, we introduce the following two player game variant of the Erdos-Szekeres problem: Consider a two player game where each player playing in alternate turns, place points in the plane. The objective of the game is to avoid the formation of the convex k-gon among the placed points. The game ends when a convex k-gon is formed and the player who placed the last point loses the game. In our paper we show a winning strategy for the player who plays second in the convex 5-gon game and the empty convex 5-gon game by considering convex layer configurations at each step. We prove that the game always ends in the 9th step by showing that the game reaches a specific set of configurations.

In-doped multifilled n-type skutterudites with ZT=1.8

In this paper we maximize the thermoelectric (TE) figure of merit, ZT, of n-type skutterudites, (In,Sr,Ba,Yb)(y)Co4Sb12, via three different routes: (i) find the optimum fraction of In as fourth filler (ii) check the influence of powder particle, grain, and crystallite size on the TE properties and (iii) check thermal stability. Filled n-type (Sr, Ba, Yb)(y)Co4Sb12 was mixed in three different proportions with In0.4Co4Sb12, ball milled (regular or high-energy (HB) ball milling) and hot-pressed. Particle size analyses and SEM pictures of the broken surfaces of the hot pressed samples document that only HB produces uniform particles/grains with average crystallite sizes similar to 100 nm, proven by transmission electron microscopy. X-ray Rietveld refinements combined with EDX indicate that in all cases indium entered the icosahedral voids of the skutterudite. Temperature dependent physical properties of all three regularly ball-milled samples show that increasing In-content infers an increasing electrical resistivity, increasing Seebeck coefficient but a decreasing total thermal conductivity. Although ZT (823 K) is in the same range as for the sample without In, the ZT values in the whole temperature range are higher and consequently the TE-conversion efficiency, eta is at least 10% higher. Annealing the samples at 600 degrees C for three days shows minor changes in structure and thermoelectric properties, indicating TE stability. The HB sample, due to uniformly small particles, equally sized grains and crystallites, exhibits a high power factor (4.4 mW/m K-2 at 730 K) and a very low thermal conductivity leading to an outstanding high ZT = 1.8 at 823 K (eta(max) = 17.5%). (C) 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

L-infinity norms of holomorphic modular forms in the case of compact quotient

We prove a sub-convex estimate for the sup-norm of L-2-normalized holomorphic modular forms of weight k on the upper half plane, with respect to the unit group of a quaternion division algebra over Q. More precisely we show that when the L-2 norm of an eigenfunction f is one, parallel to f parallel to(infinity) <<(epsilon) k(1/2-1/33+epsilon) for any epsilon > 0 and for all k sufficiently large.

Novel architecture for anomalous strengthening of a particulate filled polymer matrix composite

We propose an architecture for dramatically enhancing the stress bearing and energy absorption capacities of a polymer based composite. Different weight fractions of iron oxide nano-particles (NPs) are mixed in a poly(dimethylesiloxane) (PDMS) matrix either uniformly or into several vertically aligned cylindrical pillars. These composites are compressed up to a strain of 60% at a strain rate of 0.01 s(-1) following which they are fully unloaded at the same rate. Load bearing and energy absorption capacities of the composite with uniform distribution of NPs increase by similar to 50% upon addition of 5 wt% of NPs; however, these properties monotonically decrease with further addition of NPs so much so that the load bearing capacity of the composite becomes 1/6th of PDMS upon addition of 20 wt% of NPs. On the contrary, stress at a strain of 60% and energy absorption capacity of the composites with pillar configuration monotonically increase with the weight fraction of NPs in the pillars wherein the load bearing capacity becomes 1.5 times of PDMS when the pillars consisted of 20 wt% of NPs. In situ mechanical testing of composites with pillars reveals outward bending of the pillars wherein the pillars and the PDMS in between two pillars, located along a radius, are significantly compressed. Reasoning based on effects of compressive hydrostatic stress and shape of fillers is developed to explain the observed anomalous strengthening of the composite with pillar architecture.

Minimization Problems Based on Relative alpha-Entropy I: Forward Projection

Minimization problems with respect to a one-parameter family of generalized relative entropies are studied. These relative entropies, which we term relative alpha-entropies (denoted I-alpha), arise as redundancies under mismatched compression when cumulants of compressed lengths are considered instead of expected compressed lengths. These parametric relative entropies are a generalization of the usual relative entropy (Kullback-Leibler divergence). Just like relative entropy, these relative alpha-entropies behave like squared Euclidean distance and satisfy the Pythagorean property. Minimizers of these relative alpha-entropies on closed and convex sets are shown to exist. Such minimizations generalize the maximum Renyi or Tsallis entropy principle. The minimizing probability distribution (termed forward I-alpha-projection) for a linear family is shown to obey a power-law. Other results in connection with statistical inference, namely subspace transitivity and iterated projections, are also established. In a companion paper, a related minimization problem of interest in robust statistics that leads to a reverse I-alpha-projection is studied.

Minimization Problems Based on Relative alpha-Entropy II: Reverse Projection

In part I of this two-part work, certain minimization problems based on a parametric family of relative entropies (denoted I-alpha) were studied. Such minimizers were called forward I-alpha-projections. Here, a complementary class of minimization problems leading to the so-called reverse I-alpha-projections are studied. Reverse I-alpha-projections, particularly on log-convex or power-law families, are of interest in robust estimation problems (alpha > 1) and in constrained compression settings (alpha < 1). Orthogonality of the power-law family with an associated linear family is first established and is then exploited to turn a reverse I-alpha-projection into a forward I-alpha-projection. The transformed problem is a simpler quasi-convex minimization subject to linear constraints.

On Locally Gabriel Geometric Graphs

Let be a set of points in the plane. A geometric graph on is said to be locally Gabriel if for every edge in , the Euclidean disk with the segment joining and as diameter does not contain any points of that are neighbors of or in . A locally Gabriel graph(LGG) is a generalization of Gabriel graph and is motivated by applications in wireless networks. Unlike a Gabriel graph, there is no unique LGG on a given point set since no edge in a LGG is necessarily included or excluded. Thus the edge set of the graph can be customized to optimize certain network parameters depending on the application. The unit distance graph(UDG), introduced by Erdos, is also a LGG. In this paper, we show the following combinatorial bounds on edge complexity and independent sets of LGG: (i) For any , there exists LGG with edges. This improves upon the previous best bound of . (ii) For various subclasses of convex point sets, we show tight linear bounds on the maximum edge complexity of LGG. (iii) For any LGG on any point set, there exists an independent set of size .