Showing 1–50 of 93 results for author: Sahai, A

Measuring Psychological Depth in Language Models

Authors: Fabrice Harel-Canada, Hanyu Zhou, Sreya Mupalla, Zeynep Yildiz, Amit Sahai, Nanyun Peng

Abstract: Evaluations of creative stories generated by large language models (LLMs) often focus on objective properties of the text, such as its style, coherence, and toxicity. While these metrics are indispensable, they do not speak to a story's subjective, psychological impact from a reader's perspective. We introduce the Psychological Depth Scale (PDS), a novel framework rooted in literary theory that me… ▽ More Evaluations of creative stories generated by large language models (LLMs) often focus on objective properties of the text, such as its style, coherence, and toxicity. While these metrics are indispensable, they do not speak to a story's subjective, psychological impact from a reader's perspective. We introduce the Psychological Depth Scale (PDS), a novel framework rooted in literary theory that measures an LLM's ability to produce authentic and narratively complex stories that provoke emotion, empathy, and engagement. We empirically validate our framework by showing that humans can consistently evaluate stories based on PDS (0.72 Krippendorff's alpha). We also explore techniques for automating the PDS to easily scale future analyses. GPT-4o, combined with a novel Mixture-of-Personas (MoP) prompting strategy, achieves an average Spearman correlation of $0.51$ with human judgment while Llama-3-70B scores as high as 0.68 for empathy. Finally, we compared the depth of stories authored by both humans and LLMs. Surprisingly, GPT-4 stories either surpassed or were statistically indistinguishable from highly-rated human-written stories sourced from Reddit. By shifting the focus from text to reader, the Psychological Depth Scale is a validated, automated, and systematic means of measuring the capacity of LLMs to connect with humans through the stories they tell. △ Less

Submitted 18 June, 2024; originally announced June 2024.

Comments: Preprint. Under Review

Extreme plasmons

Authors: Aakash A. Sahai

Abstract: Nanosciences largely rely on plasmons which are quasiparticles constituted by collective oscillations of quantum electron gas composed of conduction band electrons that occupy discrete quantum states. Our work has introduced non-perturbative plasmons with oscillation amplitudes that approach the extreme limit set by breakdown in characteristic coherence. In contrast, conventional plasmons are smal… ▽ More Nanosciences largely rely on plasmons which are quasiparticles constituted by collective oscillations of quantum electron gas composed of conduction band electrons that occupy discrete quantum states. Our work has introduced non-perturbative plasmons with oscillation amplitudes that approach the extreme limit set by breakdown in characteristic coherence. In contrast, conventional plasmons are small-amplitude oscillations. Controlled excitation of extreme plasmons modeled in our work unleashes unprecedented Petavolts per meter fields. In this work, an analytical model of this new class of plasmons is developed based on quantum kinetic framework. A controllable extreme plasmon, the surface "crunch-in" plasmon, is modeled here using a modified independent electron approximation which takes into account the quantum oscillation frequency. Key characteristics of such realizable extreme plasmons that unlock unparalleled possibilities, are obtained. △ Less

Submitted 2 April, 2024; originally announced April 2024.

Precise Asymptotic Generalization for Multiclass Classification with Overparameterized Linear Models

Authors: David X. Wu, Anant Sahai

Abstract: We study the asymptotic generalization of an overparameterized linear model for multiclass classification under the Gaussian covariates bi-level model introduced in Subramanian et al.~'22, where the number of data points, features, and classes all grow together. We fully resolve the conjecture posed in Subramanian et al.~'22, matching the predicted regimes for generalization. Furthermore, our new… ▽ More We study the asymptotic generalization of an overparameterized linear model for multiclass classification under the Gaussian covariates bi-level model introduced in Subramanian et al.~'22, where the number of data points, features, and classes all grow together. We fully resolve the conjecture posed in Subramanian et al.~'22, matching the predicted regimes for generalization. Furthermore, our new lower bounds are akin to an information-theoretic strong converse: they establish that the misclassification rate goes to 0 or 1 asymptotically. One surprising consequence of our tight results is that the min-norm interpolating classifier can be asymptotically suboptimal relative to noninterpolating classifiers in the regime where the min-norm interpolating regressor is known to be optimal. The key to our tight analysis is a new variant of the Hanson-Wright inequality which is broadly useful for multiclass problems with sparse labels. As an application, we show that the same type of analysis can be used to analyze the related multilabel classification problem under the same bi-level ensemble. △ Less

Submitted 5 December, 2023; v1 submitted 22 June, 2023; originally announced June 2023.

Comments: NeurIPS 2023, 56 pages

Intraseasonal Oscillation of Land Surface Moisture and its role in the maintenance of land ITCZ during the active phases of the Indian Summer Monsoon

Authors: Pratibha Gautam, Rajib Chattopadhyay, Gill Martin, Susmitha Joseph, A. K. Sahai

Abstract: What is the role of soil moisture in maintaining the land ITCZ during the active phase of the monsoon? This question has been addressed in this study by using ERA5 reanalysis datasets, and then we evaluate the question in the CFS model-free run. Like rainfall, soil moisture also show intraseasonal oscillation. Furthermore, the sub-seasonal and seasonal features of soil moisture are different from… ▽ More What is the role of soil moisture in maintaining the land ITCZ during the active phase of the monsoon? This question has been addressed in this study by using ERA5 reanalysis datasets, and then we evaluate the question in the CFS model-free run. Like rainfall, soil moisture also show intraseasonal oscillation. Furthermore, the sub-seasonal and seasonal features of soil moisture are different from each other. During the summer monsoon season, the maximum soil moisture is found over western coastal regions, central parts of India, and the northeastern Indian subcontinent. However, during active phases of the monsoon, the maximum positive soil moisture anomaly was found in North West parts of India. soil moisture also play a pre-conditioning role during active phases of the monsoon over the monsoon core zone of India. When it is further divided into two boxes, the north monsoon core zone, and the south monsoon core zone, it is found that the preconditioning depends on that region's soil type and climate classification. Also, we calculate the moist static energy (MSE) budget during the monsoon phases to show how soil moisture feedback affects the boundary layer MSE and rainfall. A similar analysis is applied to the model run, but it cannot show the realistic preconditioning role of soil moisture and its feedback on the rainfall as in observations. We conclude that to get proper feedback between soil moisture and precipitation during the active phase of the monsoon in the model, the pre-conditioning of soil moisture should be realistic. △ Less

Submitted 30 March, 2023; originally announced March 2023.

Report of the Topical Group on Physics Beyond the Standard Model at Energy Frontier for Snowmass 2021

Authors: Tulika Bose, Antonio Boveia, Caterina Doglioni, Simone Pagan Griso, James Hirschauer, Elliot Lipeles, Zhen Liu, Nausheen R. Shah, Lian-Tao Wang, Kaustubh Agashe, Juliette Alimena, Sebastian Baum, Mohamed Berkat, Kevin Black, Gwen Gardner, Tony Gherghetta, Josh Greaves, Maxx Haehn, Phil C. Harris, Robert Harris, Julie Hogan, Suneth Jayawardana, Abraham Kahn, Jan Kalinowski, Simon Knapen , et al. (297 additional authors not shown)

Abstract: This is the Snowmass2021 Energy Frontier (EF) Beyond the Standard Model (BSM) report. It combines the EF topical group reports of EF08 (Model-specific explorations), EF09 (More general explorations), and EF10 (Dark Matter at Colliders). The report includes a general introduction to BSM motivations and the comparative prospects for proposed future experiments for a broad range of potential BSM mode… ▽ More This is the Snowmass2021 Energy Frontier (EF) Beyond the Standard Model (BSM) report. It combines the EF topical group reports of EF08 (Model-specific explorations), EF09 (More general explorations), and EF10 (Dark Matter at Colliders). The report includes a general introduction to BSM motivations and the comparative prospects for proposed future experiments for a broad range of potential BSM models and signatures, including compositeness, SUSY, leptoquarks, more general new bosons and fermions, long-lived particles, dark matter, charged-lepton flavor violation, and anomaly detection. △ Less

Submitted 18 October, 2022; v1 submitted 26 September, 2022; originally announced September 2022.

Comments: 108 pages + 38 pages references and appendix, 37 figures, Report of the Topical Group on Beyond the Standard Model Physics at Energy Frontier for Snowmass 2021. The first nine authors are the Conveners, with Contributions from the other authors

Approaching Petavolts per meter plasmonics using structured semiconductors

Authors: Aakash A. Sahai, M. Golkowski, T. Katsouleas, G. Andonian, G. White, C. Joshi, P. Taborek, V. Harid, J. Stohr

Abstract: A new class of strongly excited plasmonic modes that open access to unprecedented Petavolts per meter electromagnetic fields promise wide-ranging, transformative impact. These modes are constituted by large amplitude oscillations of the ultradense, delocalized free electron Fermi gas which is inherent in conductive media. Here structured semiconductors with appropriate concentration of n-type dopa… ▽ More A new class of strongly excited plasmonic modes that open access to unprecedented Petavolts per meter electromagnetic fields promise wide-ranging, transformative impact. These modes are constituted by large amplitude oscillations of the ultradense, delocalized free electron Fermi gas which is inherent in conductive media. Here structured semiconductors with appropriate concentration of n-type dopant are introduced to tune the properties of the Fermi gas for matched excitation of an electrostatic, surface "crunch-in" plasmon using readily available electron beams of ten micron overall dimensions and hundreds of picoCoulomb charge launched inside a tube. Strong excitation made possible by matching results in relativistic oscillations of the Fermi electron gas and uncovers unique phenomena. Relativistically induced ballistic electron transport comes about due to relativistic multifold increase in the mean free path. Acquired ballistic transport also leads to unconventional heat deposition beyond the Ohm's law. This explains the absence of observed damage or solid-plasma formation in experiments on interaction of conductive samples with electron bunches shorter than $\rm 10^{-13} seconds$. Furthermore, relativistic momentum leads to copious tunneling of electron gas allowing it to traverse the surface and crunch inside the tube. Relativistic effects along with large, localized variation of Fermi gas density underlying these modes necessitate the kinetic approach coupled with particle-in-cell simulations. Experimental verification of acceleration and focusing of electron beams modeled here using tens of Gigavolts per meter fields excited in semiconductors with $\rm 10^{18}cm^{-3}$ free electron density will pave the way for Petavolts per meter plasmonics. △ Less

Submitted 1 August, 2022; originally announced August 2022.

Comments: 16 pages, 10 figures

Generalization for multiclass classification with overparameterized linear models

Authors: Vignesh Subramanian, Rahul Arya, Anant Sahai

Abstract: Via an overparameterized linear model with Gaussian features, we provide conditions for good generalization for multiclass classification of minimum-norm interpolating solutions in an asymptotic setting where both the number of underlying features and the number of classes scale with the number of training points. The survival/contamination analysis framework for understanding the behavior of over… ▽ More Via an overparameterized linear model with Gaussian features, we provide conditions for good generalization for multiclass classification of minimum-norm interpolating solutions in an asymptotic setting where both the number of underlying features and the number of classes scale with the number of training points. The survival/contamination analysis framework for understanding the behavior of overparameterized learning problems is adapted to this setting, revealing that multiclass classification qualitatively behaves like binary classification in that, as long as there are not too many classes (made precise in the paper), it is possible to generalize well even in some settings where the corresponding regression tasks would not generalize. Besides various technical challenges, it turns out that the key difference from the binary classification setting is that there are relatively fewer positive training examples of each class in the multiclass setting as the number of classes increases, making the multiclass problem "harder" than the binary one. △ Less

Submitted 3 June, 2022; originally announced June 2022.

Comments: 44 pages, 4 figures

PetaVolts per meter Plasmonics: Snowmass21 White Paper

Authors: Aakash A. Sahai, Mark Golkowski, Stephen Gedney, Thomas Katsouleas, Gerard Andonian, Glen White, Joachim Stohr, Patric Muggli, Daniele Filipetto, Frank Zimmermann, Toshiki Tajima, Gerard Mourou, Javier Resta-Lopez

Abstract: Plasmonic modes offer the potential to achieve PetaVolts per meter fields, that would transform the current paradigm in collider development in addition to non-collider searches in fundamental physics. PetaVolts per meter plasmonics relies on collective oscillations of the free electron Fermi gas inherent in the conduction band of materials that have a suitable combination of constituent atoms and… ▽ More Plasmonic modes offer the potential to achieve PetaVolts per meter fields, that would transform the current paradigm in collider development in addition to non-collider searches in fundamental physics. PetaVolts per meter plasmonics relies on collective oscillations of the free electron Fermi gas inherent in the conduction band of materials that have a suitable combination of constituent atoms and ionic lattice structure. As the conduction band free electron density, at equilibrium, can be as high as $\rm 10^{24}cm^{-3}$, electromagnetic fields of the order of $\rm 0.1 \sqrt{\rm n_0(10^{24}cm^{-3})} ~ PVm^{-1}$ can be sustained by plasmonic modes. Engineered materials not only allow highly tunable material properties but quite critically make it possible to overcome disruptive instabilities that dominate the interactions in bulk media. Due to rapid shielding by the free electron Fermi gas, dielectric effects are strongly suppressed. Because the ionic lattice, the corresponding electronic energy bands and the free electron gas are governed by quantum mechanical effects, comparisons with plasmas are merely notional. Based on this framework, it is critical to address various challenges that underlie PetaVolts per meter plasmonics including stable excitation of plasmonic modes while accounting for their effects on the ionic lattice and the electronic energy band structure over femtosecond timescales. We summarize the ongoing theoretical and experimental efforts as well as map out strategies for the future. Extreme plasmonic fields can shape the future by not only bringing tens of TeV to multi-PeV center-of-mass-energies within reach but also by opening novel pathways in non-collider HEP. In view of this promise, we invite the scientific community to help realize the immense potential of PV/m plasmonics and call for significant expansion of the US and international R\&D program. △ Less

Submitted 22 March, 2022; originally announced March 2022.

Comments: contribution to Snowmass 2021

Snowmass21 Accelerator Modeling Community White Paper

Authors: S. Biedron, L. Brouwer, D. L. Bruhwiler, N. M. Cook, A. L. Edelen, D. Filippetto, C. -K. Huang, A. Huebl, T. Katsouleas, N. Kuklev, R. Lehe, S. Lund, C. Messe, W. Mori, C. -K. Ng, D. Perez, P. Piot, J. Qiang, R. Roussel, D. Sagan, A. Sahai, A. Scheinker, M. Thévenet, F. Tsung, J. -L. Vay , et al. (2 additional authors not shown)

Abstract: After a summary of relevant comments and recommendations from various reports over the last ten years, this paper examines the modeling needs in accelerator physics, from the modeling of single beams and individual accelerator elements, to the realization of virtual twins that replicate all the complexity to model a particle accelerator complex as accurately as possible. We then discuss cutting-ed… ▽ More After a summary of relevant comments and recommendations from various reports over the last ten years, this paper examines the modeling needs in accelerator physics, from the modeling of single beams and individual accelerator elements, to the realization of virtual twins that replicate all the complexity to model a particle accelerator complex as accurately as possible. We then discuss cutting-edge and emerging computing opportunities, such as advanced algorithms, AI/ML and quantum computing, computational needs in hardware, software performance, portability and scalability, and needs for scalable I/O and in-situ analysis. Considerations of reliability, long-term sustainability, user support and training are considered next, before discussing the benefits of ecosystems with integrated workflows based on standardized input and output, and with integrated frameworks and data repositories developed as a community. Last, we highlight how the community can work more collaboratively and efficiently through the development of consortia and centers, and via collaboration with industry. △ Less

Submitted 22 September, 2022; v1 submitted 15 March, 2022; originally announced March 2022.

Comments: contribution to Snowmass 2021

Classification and Adversarial examples in an Overparameterized Linear Model: A Signal Processing Perspective

Authors: Adhyyan Narang, Vidya Muthukumar, Anant Sahai

Abstract: State-of-the-art deep learning classifiers are heavily overparameterized with respect to the amount of training examples and observed to generalize well on "clean" data, but be highly susceptible to infinitesmal adversarial perturbations. In this paper, we identify an overparameterized linear ensemble, that uses the "lifted" Fourier feature map, that demonstrates both of these behaviors. The input… ▽ More State-of-the-art deep learning classifiers are heavily overparameterized with respect to the amount of training examples and observed to generalize well on "clean" data, but be highly susceptible to infinitesmal adversarial perturbations. In this paper, we identify an overparameterized linear ensemble, that uses the "lifted" Fourier feature map, that demonstrates both of these behaviors. The input is one-dimensional, and the adversary is only allowed to perturb these inputs and not the non-linear features directly. We find that the learned model is susceptible to adversaries in an intermediate regime where classification generalizes but regression does not. Notably, the susceptibility arises despite the absence of model mis-specification or label noise, which are commonly cited reasons for adversarial-susceptibility. These results are extended theoretically to a random-Fourier-sum setup that exhibits double-descent behavior. In both feature-setups, the adversarial vulnerability arises because of a phenomenon we term spatial localization: the predictions of the learned model are markedly more sensitive in the vicinity of training points than elsewhere. This sensitivity is a consequence of feature lifting and is reminiscent of Gibb's and Runge's phenomena from signal processing and functional analysis. Despite the adversarial susceptibility, we find that classification with these features can be easier than the more commonly studied "independent feature" models. △ Less

Submitted 27 September, 2021; originally announced September 2021.

Comments: 32 pages, 10 figures

On the role of Initial Error Growth in the Skill of Extended Range Prediction of Madden-Julian Oscillation (MJO)

Authors: Lekshmi S, Rajib Chattopadhyay, Manpreet Kaur, Susmitha Joseph, R. Phani, A Dey, R. Mandal, AK. Sahai

Abstract: The seamless forecast approach of subseasonal to seasonal scale variability has been succeeding in the forecast of multiple meteorological scales in a uniform framework. In this paradigm, it is hypothesized that reduction in initial error in dynamical forecast would help to reduce forecast error in extended lead-time up to 2-3 weeks. This is tested in a version of operational extended range foreca… ▽ More The seamless forecast approach of subseasonal to seasonal scale variability has been succeeding in the forecast of multiple meteorological scales in a uniform framework. In this paradigm, it is hypothesized that reduction in initial error in dynamical forecast would help to reduce forecast error in extended lead-time up to 2-3 weeks. This is tested in a version of operational extended range forecasts based on Climate Forecast System version 2 (CFSv2) developed at Indian Institute of Tropical Meteorology (IITM), Pune. Forecast skills are assessed to understand the role of initial errors on the prediction skill for MJO. A set of lowest and highest initial day error (LIDE & HIDE) cases are defined and the error-growth for these categories are analysed for the strong MJO events during May to September (MJJAS). The MJO forecast initial errors are categorized and defined using the well-known multivariate MJO index introduced by Wheeler &Hendon (2004). The probability distribution of bivariate RMSE and error growth evolution (first order difference of index error for each successive lead days) with respect to extended range lead-time are used as metrics in this analysis. The result showed that initial error is not showing any influence in the skill of model after a lead time of 7-10 days and the error growth remains the same for both set of errors. A rapid error growth evolution of same order is seen for both the classified cases. Further the physical attribution of these errors is studied and found that the errors originate from the events with initial phase in Western Pacific and Indian Ocean. The spatial distribution of OLR and the zonal winds also confirms the same. The study emphasises the importance of better representation of MJO phases especially over Indian ocean in the model to improve the MJO prediction rather than focusing primarily on the initial condition △ Less

Submitted 11 May, 2021; originally announced May 2021.

Comments: 27 pages, 9 figures

Solid-state Tube Wakefield Accelerator using Surface Waves in Crystals

Authors: Aakash A. Sahai, Toshiki Tajima, Peter Taborek, Vladimir D. Shiltsev

Abstract: Solid-state or crystal acceleration has for long been regarded as an attractive frontier in advanced particle acceleration. However, experimental investigations of solid-state acceleration mechanisms which offer $\rm TVm^{-1}$ acceleration gradients have been hampered by several technological constraints. The primary constraint has been the unavailability of attosecond particle or photon sources s… ▽ More Solid-state or crystal acceleration has for long been regarded as an attractive frontier in advanced particle acceleration. However, experimental investigations of solid-state acceleration mechanisms which offer $\rm TVm^{-1}$ acceleration gradients have been hampered by several technological constraints. The primary constraint has been the unavailability of attosecond particle or photon sources suitable for excitation of collective modes in bulk crystals. Secondly, there are significant difficulties with direct high-intensity irradiation of bulk solids, such as beam instabilities due to crystal imperfections and collisions etc. In this work, we model an experimentally practicable solid-state acceleration mechanism using collective electron oscillations in crystals that sustain propagating surface waves. These surface waves are driven in the wake of a submicron long particle beam in tube shaped nanostructured crystals with tube wall densities, $n_{\rm tube}\sim10^{22-24}\rm cm^{-3}$. Particle-In-Cell (PIC) simulations carried out under experimental constraints demonstrate the possibility of accessing average acceleration gradients of several $\rm TVm^{-1}$ using the solid-state tube wakefield acceleration regime. Furthermore, our modeling demonstrates the possibility that as the surface oscillations and resultantly the surface wave transitions into a nonlinear or "crunch-in" regime under $n_{\rm beam}/n_{\rm tube} \gtrsim 0.05$, not only does the average gradient increase but strong transverse focusing fields extend down to the tube axis. This work thus demonstrates the near-term experimental realizability of Solid-State Tube Wakefield Accelerator (SOTWA). (truncated to comply with submission requirements) △ Less

Submitted 27 January, 2021; originally announced January 2021.

Comments: based upon long-standing work on the surface crunch-in mode of A. A. Sahai published in https://doi.org/10.1103/PhysRevAccelBeams.20.081004, arXiv:1610.03289 and http://doi.org/10.18429/JACoW-IPAC2015-WEPJE001

Journal ref: International Journal of Modern Physics A, Vol. 34, No. 34, 1943009 (2019)

On the Impossibility of Convergence of Mixed Strategies with No Regret Learning

Authors: Vidya Muthukumar, Soham Phade, Anant Sahai

Abstract: We study the limiting behavior of the mixed strategies that result from optimal no-regret learning strategies in a repeated game setting where the stage game is any 2 by 2 competitive game. We consider optimal no-regret algorithms that are mean-based and monotonic in their argument. We show that for any such algorithm, the limiting mixed strategies of the players cannot converge almost surely to a… ▽ More We study the limiting behavior of the mixed strategies that result from optimal no-regret learning strategies in a repeated game setting where the stage game is any 2 by 2 competitive game. We consider optimal no-regret algorithms that are mean-based and monotonic in their argument. We show that for any such algorithm, the limiting mixed strategies of the players cannot converge almost surely to any Nash equilibrium. This negative result is also shown to hold under a broad relaxation of these assumptions, including popular variants of Online-Mirror-Descent with optimism and/or adaptive step-sizes. Finally, we conjecture that the monotonicity assumption can be removed, and provide partial evidence for this conjecture. Our results identify the inherent stochasticity in players' realizations as a critical factor underlying this divergence in outcomes between using the opponent's mixtures and realizations to make updates. △ Less

Submitted 2 March, 2022; v1 submitted 3 December, 2020; originally announced December 2020.

Comments: 47 pages, 12 figures

Indistinguishability Obfuscation from Well-Founded Assumptions

Authors: Aayush Jain, Huijia Lin, Amit Sahai

Abstract: In this work, we show how to construct indistinguishability obfuscation from subexponential hardness of four well-founded assumptions. We prove: Let $τ\in (0,\infty), δ\in (0,1), ε\in (0,1)$ be arbitrary constants. Assume sub-exponential security of the following assumptions, where $λ$ is a security parameter, and the parameters $\ell,k,n$ below are large enough polynomials in $λ$: - The SXDH… ▽ More In this work, we show how to construct indistinguishability obfuscation from subexponential hardness of four well-founded assumptions. We prove: Let $τ\in (0,\infty), δ\in (0,1), ε\in (0,1)$ be arbitrary constants. Assume sub-exponential security of the following assumptions, where $λ$ is a security parameter, and the parameters $\ell,k,n$ below are large enough polynomials in $λ$: - The SXDH assumption on asymmetric bilinear groups of a prime order $p = O(2^λ)$, - The LWE assumption over $\mathbb{Z}_{p}$ with subexponential modulus-to-noise ratio $2^{k^ε}$, where $k$ is the dimension of the LWE secret, - The LPN assumption over $\mathbb{Z}_p$ with polynomially many LPN samples and error rate $1/\ell^δ$, where $\ell$ is the dimension of the LPN secret, - The existence of a Boolean PRG in $\mathsf{NC}^0$ with stretch $n^{1+τ}$, Then, (subexponentially secure) indistinguishability obfuscation for all polynomial-size circuits exists. △ Less

Submitted 21 August, 2020; originally announced August 2020.

Nanostructure Accelerators: Novel concept and path to its realization

Authors: A. Sahai, M. Golkowski, F. Zimmermann, J. Resta-Lopez, T. Tajima, V. Shiltsev

Abstract: TeV/m acceleration gradients using crystals as originally envisioned by R. Hofstadter, an early pioneer of HEP, have remained unrealizable. Fundamental obstacles that have hampered efforts on particle acceleration using bulk-crystals arise from collisional energy loss and emittance degradation in addition to severe beam disruption despite the favorable effect of particle channeling along interatom… ▽ More TeV/m acceleration gradients using crystals as originally envisioned by R. Hofstadter, an early pioneer of HEP, have remained unrealizable. Fundamental obstacles that have hampered efforts on particle acceleration using bulk-crystals arise from collisional energy loss and emittance degradation in addition to severe beam disruption despite the favorable effect of particle channeling along interatomic planes in bulk. We aspire for the union of nanoscience with accelerator science to not only overcome these problems using nanostructured tubes to avoid direct impact of the beam on bulk ion-lattice but also to utilize the highly tunable characteristics of nanomaterials. We pioneer a novel surface wave mechanism in nanostructured materials with a strong electrostatic component which not only attains tens of TeV/m gradients but also has focusing fields. Under our initiative, the proof-of-principle demonstration of tens of TeV/m gradients and beam nanomodulation is underway. Realizable nanostructure accelerators naturally promise new horizons in HEP as well as in a wide range of areas of research that utilize beams of high-energy particles or photons. △ Less

Submitted 17 June, 2020; originally announced June 2020.

Comments: submission to Snowmass'21 Accelerator Frontier

Classification vs regression in overparameterized regimes: Does the loss function matter?

Authors: Vidya Muthukumar, Adhyyan Narang, Vignesh Subramanian, Mikhail Belkin, Daniel Hsu, Anant Sahai

Abstract: We compare classification and regression tasks in an overparameterized linear model with Gaussian features. On the one hand, we show that with sufficient overparameterization all training points are support vectors: solutions obtained by least-squares minimum-norm interpolation, typically used for regression, are identical to those produced by the hard-margin support vector machine (SVM) that mini… ▽ More We compare classification and regression tasks in an overparameterized linear model with Gaussian features. On the one hand, we show that with sufficient overparameterization all training points are support vectors: solutions obtained by least-squares minimum-norm interpolation, typically used for regression, are identical to those produced by the hard-margin support vector machine (SVM) that minimizes the hinge loss, typically used for training classifiers. On the other hand, we show that there exist regimes where these interpolating solutions generalize well when evaluated by the 0-1 test loss function, but do not generalize if evaluated by the square loss function, i.e. they approach the null risk. Our results demonstrate the very different roles and properties of loss functions used at the training phase (optimization) and the testing phase (generalization). △ Less

Submitted 14 October, 2021; v1 submitted 16 May, 2020; originally announced May 2020.

Journal ref: Journal of Machine Learning Research, 22(222):1-69, 2021

Nanostructured Tube Wakefield Accelerator

Authors: Aakash A. Sahai, Toshiki Tajima, Vladimir D. Shiltsev

Abstract: Unprecedented $\rm TeVm^{-1}$ acceleration gradients are modeled to be realizable using a nonlinear surface crunch-in mode in nanostructured tubes. This mode is realizable using advances in nanofabrication and solid energy density attosecond bunch compression. Three dimensional computational and analytical modeling demonstrates GeV energy gain in sub-millimeter long tubes with effective wall densi… ▽ More Unprecedented $\rm TeVm^{-1}$ acceleration gradients are modeled to be realizable using a nonlinear surface crunch-in mode in nanostructured tubes. This mode is realizable using advances in nanofabrication and solid energy density attosecond bunch compression. Three dimensional computational and analytical modeling demonstrates GeV energy gain in sub-millimeter long tubes with effective wall densities $n_{\rm t}\sim10^{22-24}\rm cm^{-3}$ and hundreds of nanometer core radius when driven by submicron near solid electron beams, $n_{\rm b}\sim0.05n_{\rm t}$. Besides the many $\rm TVm^{-1}$ average gradients, strong self-focusing and nanomodulation of the beam which increases its peak density and the wakefield strength also opens up controlled high-energy photon production. △ Less

Submitted 20 April, 2020; originally announced April 2020.

Comments: 6 pages, 4 figures. Submitted for peer-review in March 2020 (patent filed). The mechanism in this manuscript is founded on the crunch-in nonlinear surface mode in fiber-like tube structures in continuity with the work published by the lead author in Proc. IPAC 2015 (WEPJE001) (ref.13), PRAB 20, 081004, 2017 (ref.12) and IJMPA 34, 1943009, 2019 (ref.3)

Blind interactive learning of modulation schemes: Multi-agent cooperation without co-design

Authors: Anant Sahai, Joshua Sanz, Vignesh Subramanian, Caryn Tran, Kailas Vodrahalli

Abstract: We examine the problem of learning to cooperate in the context of wireless communication. In our setting, two agents must learn modulation schemes that enable them to communicate across a power-constrained additive white Gaussian noise channel. We investigate whether learning is possible under different levels of information sharing between distributed agents which are not necessarily co-designed.… ▽ More We examine the problem of learning to cooperate in the context of wireless communication. In our setting, two agents must learn modulation schemes that enable them to communicate across a power-constrained additive white Gaussian noise channel. We investigate whether learning is possible under different levels of information sharing between distributed agents which are not necessarily co-designed. We employ the "Echo" protocol, a "blind" interactive learning protocol where an agent hears, understands, and repeats (echoes) back the message received from another agent, simultaneously training itself to communicate. To capture the idea of cooperation between "not necessarily co-designed" agents we use two different populations of function approximators - neural networks and polynomials. We also include interactions between learning agents and non-learning agents with fixed modulation protocols such as QPSK and 16QAM. We verify the universality of the Echo learning approach, showing it succeeds independent of the inner workings of the agents. In addition to matching the communication expectations of others, we show that two learning agents can collaboratively invent a successful communication approach from independent random initializations. We complement our simulations with an implementation of the Echo protocol in software-defined radios. To explore the continuum of co-design, we study how learning is impacted by different levels of information sharing between agents, including sharing training symbols, losses, and full gradients. We find that co-design (increased information sharing) accelerates learning. Learning higher order modulation schemes is a more difficult task, and the beneficial effect of co-design becomes more pronounced as the task becomes harder. △ Less

Submitted 1 April, 2020; v1 submitted 21 October, 2019; originally announced October 2019.

Comments: 33 pages, 25 figures, code can be found at https://github.com/ml4wireless/echo, accepted for publication in IEEE Access

Robust Commitments and Partial Reputation

Authors: Vidya Muthukumar, Anant Sahai

Abstract: Agents rarely act in isolation -- their behavioral history, in particular, is public to others. We seek a non-asymptotic understanding of how a leader agent should shape this history to its maximal advantage, knowing that follower agent(s) will be learning and responding to it. We study Stackelberg leader-follower games with finite observations of the leader commitment, which commonly models secur… ▽ More Agents rarely act in isolation -- their behavioral history, in particular, is public to others. We seek a non-asymptotic understanding of how a leader agent should shape this history to its maximal advantage, knowing that follower agent(s) will be learning and responding to it. We study Stackelberg leader-follower games with finite observations of the leader commitment, which commonly models security games and network routing in engineering, and persuasion mechanisms in economics. First, we formally show that when the game is not zero-sum and the vanilla Stackelberg commitment is mixed, it is not robust to observational uncertainty. We propose observation-robust, polynomial-time-computable commitment constructions for leader strategies that approximate the Stackelberg payoff, and also show that these commitment rules approximate the maximum obtainable payoff (which could in general be greater than the Stackelberg payoff). △ Less

Submitted 27 May, 2019; originally announced May 2019.

Comments: 29 pages, extended abstract at ACM Economics and Computation 2019

Learning Physical-Layer Communication with Quantized Feedback

Authors: **xiang Song, Bile Peng, Christian Häger, Henk Wymeersch, Anant Sahai

Abstract: Data-driven optimization of transmitters and receivers can reveal new modulation and detection schemes and enable physical-layer communication over unknown channels. Previous work has shown that practical implementations of this approach require a feedback signal from the receiver to the transmitter. In this paper, we study the impact of quantized feedback in data-driven learning of physical-layer… ▽ More Data-driven optimization of transmitters and receivers can reveal new modulation and detection schemes and enable physical-layer communication over unknown channels. Previous work has shown that practical implementations of this approach require a feedback signal from the receiver to the transmitter. In this paper, we study the impact of quantized feedback in data-driven learning of physical-layer communication. A novel quantization method is proposed, which exploits the specific properties of the feedback signal and is suitable for non-stationary signal distributions. The method is evaluated for linear and nonlinear channels. Simulation results show that feedback quantization does not appreciably affect the learning process and can lead to excellent performance, even with $1$-bit quantization. In addition, it is shown that learning is surprisingly robust to noisy feedback where random bit flips are applied to the quantization bits. △ Less

Submitted 4 November, 2019; v1 submitted 19 April, 2019; originally announced April 2019.

Harmless interpolation of noisy data in regression

Authors: Vidya Muthukumar, Kailas Vodrahalli, Vignesh Subramanian, Anant Sahai

Abstract: A continuing mystery in understanding the empirical success of deep neural networks is their ability to achieve zero training error and generalize well, even when the training data is noisy and there are more parameters than data points. We investigate this overparameterized regime in linear regression, where all solutions that minimize training error interpolate the data, including noise. We char… ▽ More A continuing mystery in understanding the empirical success of deep neural networks is their ability to achieve zero training error and generalize well, even when the training data is noisy and there are more parameters than data points. We investigate this overparameterized regime in linear regression, where all solutions that minimize training error interpolate the data, including noise. We characterize the fundamental generalization (mean-squared) error of any interpolating solution in the presence of noise, and show that this error decays to zero with the number of features. Thus, overparameterization can be explicitly beneficial in ensuring harmless interpolation of noise. We discuss two root causes for poor generalization that are complementary in nature -- signal "bleeding" into a large number of alias features, and overfitting of noise by parsimonious feature selectors. For the sparse linear model with noise, we provide a hybrid interpolating scheme that mitigates both these issues and achieves order-optimal MSE over all possible interpolating solutions. △ Less

Submitted 9 September, 2019; v1 submitted 21 March, 2019; originally announced March 2019.

Comments: 52 pages, expanded version of the paper presented at ITA in San Diego in Feb 2019, ISIT in Paris in July 2019, at Simons in July, and as a plenary at ITW in Visby in August 2019

Spectrogram Feature Losses for Music Source Separation

Authors: Abhimanyu Sahai, Romann Weber, Brian McWilliams

Abstract: In this paper we study deep learning-based music source separation, and explore using an alternative loss to the standard spectrogram pixel-level L2 loss for model training. Our main contribution is in demonstrating that adding a high-level feature loss term, extracted from the spectrograms using a VGG net, can improve separation quality vis-a-vis a pure pixel-level loss. We show this improvement… ▽ More In this paper we study deep learning-based music source separation, and explore using an alternative loss to the standard spectrogram pixel-level L2 loss for model training. Our main contribution is in demonstrating that adding a high-level feature loss term, extracted from the spectrograms using a VGG net, can improve separation quality vis-a-vis a pure pixel-level loss. We show this improvement in the context of the MMDenseNet, a State-of-the-Art deep learning model for this task, for the extraction of drums and vocal sounds from songs in the musdb18 database, covering a broad range of western music genres. We believe that this finding can be generalized and applied to broader machine learning-based systems in the audio domain. △ Less

Submitted 26 June, 2019; v1 submitted 15 January, 2019; originally announced January 2019.

Comments: Accepted for presentation at the 27th European Signal Processing Conference (EUSIPCO 2019)

MSC Class: 62; 68 ACM Class: I.2.6; H.5.5

Expander Graphs are Non-Malleable Codes

Authors: Peter M. R. Rasmussen, Amit Sahai

Abstract: Any $d$-regular graph on $n$ vertices with spectral expansion $λ$ satisfying $n = Ω(d^3\log(d)/λ)$ yields a $O\left(\frac{λ^{3/2}}{d}\right)$-non-malleable code for single-bit messages in the split-state model. Any $d$-regular graph on $n$ vertices with spectral expansion $λ$ satisfying $n = Ω(d^3\log(d)/λ)$ yields a $O\left(\frac{λ^{3/2}}{d}\right)$-non-malleable code for single-bit messages in the split-state model. △ Less

Submitted 20 March, 2019; v1 submitted 28 September, 2018; originally announced October 2018.

Comments: 10 pages Resubmitted with revised introduction and acknowledgement

Wireless Channel Dynamics and Robustness for Ultra-Reliable Low-Latency Communications

Authors: Vasuki Narasimha Swamy, Paul Rigge, Gireeja Ranade, Borivoje Nikolic, Anant Sahai

Abstract: Interactive, immersive and critical applications demand ultra-reliable low-latency communication (URLLC). To build wireless communication systems that can support these applications, understanding the characteristics of the wireless medium is paramount. Although wireless channel characteristics and dynamics have been extensively studied, it is important to revisit these concepts in the context of… ▽ More Interactive, immersive and critical applications demand ultra-reliable low-latency communication (URLLC). To build wireless communication systems that can support these applications, understanding the characteristics of the wireless medium is paramount. Although wireless channel characteristics and dynamics have been extensively studied, it is important to revisit these concepts in the context of the strict demands of low latency and ultra-reliability. In this paper, we bring a modeling approach from robust control to wireless communication -- the wireless channel characteristics are given a nominal model around which we allow for some quantified uncertainty. We propose certain key "directions" along which to bound model uncertainty that are relevant to URLLC. For the nominal model, we take an in-depth look at wireless channel characteristics such as spatial and temporal correlations based on Jakes' model. Contrary to what has been claimed in the literature, we find that standard Rayleigh fading processes are not bandlimited. This has significant implications on the predictability of channels. We also find that under reasonable conditions the spatial correlation of channels provide a fading distribution that is not too far off from an independent spatial fading model. Additionally, we look at the impact of these channel models on cooperative communication based systems. We find that while spatial-diversity-based techniques are necessary to combat the adverse effects of fading, time-diversity-based techniques are necessary to be robust against unmodeled errors. Robust URLLC systems need to operate with both an SNR margin and a time/repetition margin. △ Less

Submitted 22 June, 2018; originally announced June 2018.

Comments: Submitted to IEEE JSAC Special Issue on Ultra-Reliable Low-Latency Communications in Wireless Networks

Best of many worlds: Robust model selection for online supervised learning

Authors: Vidya Muthukumar, Mitas Ray, Anant Sahai, Peter L. Bartlett

Abstract: We introduce algorithms for online, full-information prediction that are competitive with contextual tree experts of unknown complexity, in both probabilistic and adversarial settings. We show that by incorporating a probabilistic framework of structural risk minimization into existing adaptive algorithms, we can robustly learn not only the presence of stochastic structure when it exists (leading… ▽ More We introduce algorithms for online, full-information prediction that are competitive with contextual tree experts of unknown complexity, in both probabilistic and adversarial settings. We show that by incorporating a probabilistic framework of structural risk minimization into existing adaptive algorithms, we can robustly learn not only the presence of stochastic structure when it exists (leading to constant as opposed to $\mathcal{O}(\sqrt{T})$ regret), but also the correct model order. We thus obtain regret bounds that are competitive with the regret of an optimal algorithm that possesses strong side information about both the complexity of the optimal contextual tree expert and whether the process generating the data is stochastic or adversarial. These are the first constructive guarantees on simultaneous adaptivity to the model and the presence of stochasticity. △ Less

Submitted 22 May, 2018; originally announced May 2018.

Comments: 33 pages, 5 figures

Network Coding for Real-time Wireless Communication for Automation

Authors: Vasuki Narasimha Swamy, Paul Rigge, Gireeja Ranade, Anant Sahai, Borivoje Nikolic

Abstract: Real-time applications require latencies on the order of a millisecond with very high reliabilities, paralleling the requirements for high-performance industrial control. Current wireless technologies like WiFi, Bluetooth, LTE, etc. are unable to meet these stringent latency and reliability requirements, forcing the use of wired systems. This paper introduces a wireless communication protocol base… ▽ More Real-time applications require latencies on the order of a millisecond with very high reliabilities, paralleling the requirements for high-performance industrial control. Current wireless technologies like WiFi, Bluetooth, LTE, etc. are unable to meet these stringent latency and reliability requirements, forcing the use of wired systems. This paper introduces a wireless communication protocol based on network coding that in conjunction with cooperative communication techniques builds the necessary diversity to achieve the target reliability. The proposed protocol is analyzed using a communication theoretic delay-limited-capacity framework and compared to proposed protocols without network coding. The results show that for larger network sizes or payloads employing network coding lowers the minimum SNR required to achieve the target reliability. For a scenario inspired by an industrial printing application with $30$ nodes in the control loop, aggregate throughput of $4.8$ Mb/s, $20$MHz of bandwidth and cycle time under $2$ ms, the protocol can robustly achieve a system probability of error better than $10^{-9}$ with a nominal SNR less than $2$ dB under ideal channel conditions. △ Less

Submitted 14 March, 2018; originally announced March 2018.

Comments: A preliminary version of this work appeared at IEEE WCNC 2016

Quasi-monoenergetic Laser-Plasma Positron Accelerator using Particle-Shower Plasma-Wave interactions

Authors: Aakash A. Sahai

Abstract: An all-optical centimeter-scale laser-plasma positron accelerator is modeled to produce quasi-monoenergetic beams with tunable ultra-relativistic energies. A new principle elucidated here describes the trap** of divergent positrons that are part of a laser-driven electromagnetic shower with a large energy spread and their acceleration into a quasi-monoenergetic positron beam in a laser-driven pl… ▽ More An all-optical centimeter-scale laser-plasma positron accelerator is modeled to produce quasi-monoenergetic beams with tunable ultra-relativistic energies. A new principle elucidated here describes the trap** of divergent positrons that are part of a laser-driven electromagnetic shower with a large energy spread and their acceleration into a quasi-monoenergetic positron beam in a laser-driven plasma wave. Proof of this principle using analysis and Particle-In-Cell simulations demonstrates that, under limits defined here, existing lasers can accelerate hundreds of MeV pC quasi-monoenergetic positron bunches. By providing an affordable alternative to kilometer-scale radio-frequency accelerators, this compact positron accelerator opens up new avenues of research. △ Less

Submitted 18 April, 2018; v1 submitted 19 January, 2018; originally announced January 2018.

Comments: submitted to Physical Review Letters, January 2018

Journal ref: Phys. Rev. Accel. Beams 21, 081301 (2018)

Cooperative Multi-Agent Reinforcement Learning for Low-Level Wireless Communication

Authors: Colin de Vrieze, Shane Barratt, Daniel Tsai, Anant Sahai

Abstract: Traditional radio systems are strictly co-designed on the lower levels of the OSI stack for compatibility and efficiency. Although this has enabled the success of radio communications, it has also introduced lengthy standardization processes and imposed static allocation of the radio spectrum. Various initiatives have been undertaken by the research community to tackle the problem of artificial sp… ▽ More Traditional radio systems are strictly co-designed on the lower levels of the OSI stack for compatibility and efficiency. Although this has enabled the success of radio communications, it has also introduced lengthy standardization processes and imposed static allocation of the radio spectrum. Various initiatives have been undertaken by the research community to tackle the problem of artificial spectrum scarcity by both making frequency allocation more dynamic and building flexible radios to replace the static ones. There is reason to believe that just as computer vision and control have been overhauled by the introduction of machine learning, wireless communication can also be improved by utilizing similar techniques to increase the flexibility of wireless networks. In this work, we pose the problem of discovering low-level wireless communication schemes ex-nihilo between two agents in a fully decentralized fashion as a reinforcement learning problem. Our proposed approach uses policy gradients to learn an optimal bi-directional communication scheme and shows surprisingly sophisticated and intelligent learning behavior. We present the results of extensive experiments and an analysis of the fidelity of our approach. △ Less

Submitted 14 January, 2018; originally announced January 2018.

Strongly-Mismatched Regime of Nonlinear Laser-Plasma Acceleration: Optimization of Laser to Energetic Particle Efficiency

Authors: Aakash A. Sahai

Abstract: A strongly mismatched regime of self-guided nonlinear laser-plasma acceleration in the bubble regime is modeled for optimization of Laser to Particle energy efficiency with application to recently proposed laser positron accelerator. The strong mismatch, in contrast with the matched condition, arises from the incident laser spot-size being much larger than that needed for equilibration of the lase… ▽ More A strongly mismatched regime of self-guided nonlinear laser-plasma acceleration in the bubble regime is modeled for optimization of Laser to Particle energy efficiency with application to recently proposed laser positron accelerator. The strong mismatch, in contrast with the matched condition, arises from the incident laser spot-size being much larger than that needed for equilibration of the laser ponderomotive and electron-ion charge-separation force in the plasma bubble. This is shown to be favorable for optimization of large self-injected electron charge and ultra-low transverse emittance. The prominent signatures of the mismatched regime, strong optical-shock excitation and bubble elongation, are validated using multi-dimensional Particle-In-Cell simulations. This work thus uncovers a generalized regime that is shown to have been favored by many laser-plasma acceleration experiments and opens a novel pathway for a wide-range of future applications. △ Less

Submitted 16 December, 2018; v1 submitted 1 November, 2017; originally announced November 2017.

Comments: 12 pages, 8 figures

Laser-driven plasma acceleration in a regime of strong-mismatch between the incident laser envelope and the nonlinear plasma response

Authors: A. A. Sahai, K. Poder, J. C. Wood, J. M. Cole, N. C. Lopes, S. P. D. Mangles, Z. Najmudin

Abstract: We explore a regime of laser-driven plasma acceleration of electrons where the radial envelope of the laser-pulse incident at the plasma entrance is strongly mismatched to the nonlinear plasma electron response excited by it. This regime has been experimentally studied with the gemini laser using f/40 focusing optics in August 2015 and f/20 in 2008. The physical mechanisms and the scaling laws of… ▽ More We explore a regime of laser-driven plasma acceleration of electrons where the radial envelope of the laser-pulse incident at the plasma entrance is strongly mismatched to the nonlinear plasma electron response excited by it. This regime has been experimentally studied with the gemini laser using f/40 focusing optics in August 2015 and f/20 in 2008. The physical mechanisms and the scaling laws of electron acceleration achievable in a laser-plasma accelerator have been studied in the radially matched laser regime and thus are not accurate in the strongly mismatched regime explored here. In this work, we show that a novel adjusted-a0 model applicable over a specific range of densities where the laser enters the state of a strong optical shock, describes the mismatched regime. Beside several novel aspects of laser-plasma interaction dynamics relating to an elongating bubble shape and the corresponding self-injection mechanism, importantly we find that in this strongly mismatched regime when the laser pulse transforms into an optical shock it is possible to achieve beam-energies that significantly exceed the incident intensity matched regime scaling laws. △ Less

Submitted 10 April, 2017; originally announced April 2017.

Comments: Theory paper explaining the strongly mis-matched regime of LWFA and proving that there are several advantages to using such a regime over the perfectly matched regime of LWFA which has been so far considered to be the optimum for LWFA. Rutherford Appleton Lab (RAL) - Central Laser Facility (CLF) - 2017 Annual report contribution

Layered black-box, behavioral interconnection perspective and applications to the problem of communication with fidelity criteria, Part II: stationary sources satisfying ψ-mixing criterion

Authors: Mukul Agarwal, Sanjoy Mitter, Anant Sahai

Abstract: Theorems from Part 1 of this paper are generalized to ψ-mixing sources in this paper. Application to Markoff chains and order m Markoff chains is presented. The main result is the generalization of Theorem 1 in Part 1. Theorems from Part 1 of this paper are generalized to ψ-mixing sources in this paper. Application to Markoff chains and order m Markoff chains is presented. The main result is the generalization of Theorem 1 in Part 1. △ Less

Submitted 23 March, 2018; v1 submitted 15 March, 2017; originally announced March 2017.

Layered black-box, behavioral interconnection perspective and applications to the problem of communication with fidelity criteria, Part I: i.i.d. sources

Authors: Mukul Agarwal, Sanjoy Mitter, Anant Sahai

Abstract: In this paper, the problem of communication over an essentially unknown channel, which is known to be able to communicate a source to a destination to within a certain distortion level, is considered from a behavioral, interconnection view-point. Rates of reliable communication are derived and source-channel separation for communication with fidelity criteria is proved. The results are then genera… ▽ More In this paper, the problem of communication over an essentially unknown channel, which is known to be able to communicate a source to a destination to within a certain distortion level, is considered from a behavioral, interconnection view-point. Rates of reliable communication are derived and source-channel separation for communication with fidelity criteria is proved. The results are then generalized to the multi-user setting under certain assumptions. Other applications of this problem problem which follow from this perspective are discussed. △ Less

Submitted 26 March, 2018; v1 submitted 15 March, 2017; originally announced March 2017.

Control Capacity

Authors: Gireeja Ranade, Anant Sahai

Abstract: Feedback control actively dissipates uncertainty from a dynamical system by means of actuation. We develop a notion of "control capacity" that gives a fundamental limit (in bits) on the rate at which a controller can dissipate the uncertainty from a system, i.e. stabilize to a known fixed point. We give a computable single-letter characterization of control capacity for memoryless stationary scala… ▽ More Feedback control actively dissipates uncertainty from a dynamical system by means of actuation. We develop a notion of "control capacity" that gives a fundamental limit (in bits) on the rate at which a controller can dissipate the uncertainty from a system, i.e. stabilize to a known fixed point. We give a computable single-letter characterization of control capacity for memoryless stationary scalar multiplicative actuation channels. Control capacity allows us to answer questions of stabilizability for scalar linear systems: a system with actuation uncertainty is stabilizable if and only if the control capacity is larger than the log of the unstable open-loop eigenvalue. For second-moment senses of stability, we recover the classic uncertainty threshold principle result. However, our definition of control capacity can quantify the stabilizability limits for any moment of stability. Our formulation parallels the notion of Shannon's communication capacity, and thus yields both a strong converse and a way to compute the value of side-information in control. The results in our paper are motivated by bit-level models for control that build on the deterministic models that are widely used to understand information flows in wireless network information theory. △ Less

Submitted 16 January, 2017; originally announced January 2017.

Comments: 52 pages

Non-linear Ion-Wake Excitation by the Time-Asymmetric Electron Wakefields of Intense Energy Sources with applications to the Crunch-in regime

Authors: Aakash A. Sahai

Abstract: A model for the excitation of a non-linear ion-wake mode by a train of plasma electron oscillations in the non-linear time-asymmetric regime is developed using analytical theory and particle-in-cell based computational solutions. The ion-wake is shown to be a driven non-linear ion-acoustic wave in the form of a cylindrical ion-soliton. The near-void and radially-outwards propagating ion-wake chann… ▽ More A model for the excitation of a non-linear ion-wake mode by a train of plasma electron oscillations in the non-linear time-asymmetric regime is developed using analytical theory and particle-in-cell based computational solutions. The ion-wake is shown to be a driven non-linear ion-acoustic wave in the form of a cylindrical ion-soliton. The near-void and radially-outwards propagating ion-wake channel of a few plasma skin-depth radius, is explored for application to "Crunch-in" regime of positron acceleration. The coupling from the electron wakefield mode to the ion-mode dictates the long-term evolution of the plasma and the time for its relaxation back to an equilibrium, limiting the repetition-rate of a plasma accelerator. Using an analytical model it is shown that it is the time asymmetric phases of the oscillating radial electric fields of the nearly-stationary electron bubble that excite time-averaged inertial ion motion radially. The electron compression in the back of the bubble sucks-in the ions whereas the space-charge within the bubble cavity expels them, driving a cylindrical ion-soliton structure with on-axis and bubble-edge density-spikes. Once formed, the channel-edge density-spike is sustained over the length of the plasma and driven radially outwards by the thermal pressure of the wake energy in electrons. Its channel-like structure is independent of the energy-source, electromagnetic wave or particle beam, driving the bubble electron wake. Particle-In-Cell simulations are used to study the ion-wake soliton structure, its driven propagation and its use for positron acceleration in the "Crunch-in" regime. △ Less

Submitted 11 December, 2016; originally announced December 2016.

Comments: Crunch-in regime strongly contradicts the established conclusions of ZERO focusing fields in a hollow-channel (claimed and presented in several PRL, PoP, PRE and Nature papers). Since this entirely opposes and goes against established conclusions of over 20 years old work, severely damaging to the reputations of the senior physicists, it is not being allowed through the peer-review process

Journal ref: Phys. Rev. Accel. Beams 20, 081004 (2017)

Compact ring-based X-ray source with on-orbit and on-energy laser-plasma injection

Authors: Marlene Turner, Jeremy Cheatam, Auralee Edelen, James Gerity, Andrew Lajoie, Gerard Lawler, Osip Lishilin, Kook** Moon, Aakash Ajit Sahai, Andrei Seryi, Kai Shih, Brandon Zerbe

Abstract: We report here the results of a one week long investigation into the conceptual design of an X-ray source based on a compact ring with on-orbit and on-energy laser-plasma accelerator. We performed these studies during the June 2016 USPAS class "Physics of Accelerators, Lasers, and Plasma..." applying the art of inventiveness TRIZ. We describe three versions of the light source with the constraints… ▽ More We report here the results of a one week long investigation into the conceptual design of an X-ray source based on a compact ring with on-orbit and on-energy laser-plasma accelerator. We performed these studies during the June 2016 USPAS class "Physics of Accelerators, Lasers, and Plasma..." applying the art of inventiveness TRIZ. We describe three versions of the light source with the constraints of the electron beam with energy $1\,\rm{GeV}$ or $3\,\rm{GeV}$ and a magnetic lattice design being normal conducting (only for the $1\,\rm{GeV}$ beam) or superconducting (for either beam). The electron beam recirculates in the ring, to increase the effective photon flux. We describe the design choices, present relevant parameters, and describe insights into such machines. △ Less

Submitted 17 October, 2016; originally announced October 2016.

Comments: 4 pages, 1 figure, Conference Proceedings of NAPAC 2016

Crunch-in regime - Non-linearly driven hollow-channel plasma

Authors: Aakash A. Sahai

Abstract: Plasma wakefields driven inside a hollow-channel plasma are significantly different from those driven in a homogeneous plasma. This work investigates the scaling laws of the accelerating and focusing fields in the "crunch-in" regime. This regime is excited due to the collapse of the electron-rings from the channel walls onto the propagation axis of the energy-source, in its wake. This regime is th… ▽ More Plasma wakefields driven inside a hollow-channel plasma are significantly different from those driven in a homogeneous plasma. This work investigates the scaling laws of the accelerating and focusing fields in the "crunch-in" regime. This regime is excited due to the collapse of the electron-rings from the channel walls onto the propagation axis of the energy-source, in its wake. This regime is thus the non-linearly driven hollow channel, since the electron-ring displacement is of the order of the channel radius. We present the properties of the coherent structures in the "crunch-in" regime where the channel radius is matched to the beam properties such that channel-edge to on-axis collapse time has a direct correspondence to the energy source intensity. We also investigate the physical mechanisms that underlie the "crunch-in" wakefields by tuning the channel radius. Using a theoretical framework and results from PIC simulations the possible applications of the "crunch-in" regime for acceleration of positron beams with collider-scale parameters is presented. △ Less

Submitted 11 October, 2016; originally announced October 2016.

Comments: presented as a "oral contribution" at Advanced Accelerator Conference, Aug 2016, MD, USA & submitted to the proceedings of North American Particle Accelerator Conference, Oct 2016, IL, USA

Real-time Cooperative Communication for Automation over Wireless

Authors: Vasuki Narasimha Swamy, Sahaana Suri, Paul Rigge, Matthew Weiner, Gireeja Ranade, Anant Sahai, Borivoje Nikolic

Abstract: High-performance industrial automation systems rely on tens of simultaneously active sensors and actuators and have stringent communication latency and reliability requirements. Current wireless technologies like WiFi, Bluetooth, and LTE are unable to meet these requirements, forcing the use of wired communication in industrial control systems. This paper introduces a wireless communication protoc… ▽ More High-performance industrial automation systems rely on tens of simultaneously active sensors and actuators and have stringent communication latency and reliability requirements. Current wireless technologies like WiFi, Bluetooth, and LTE are unable to meet these requirements, forcing the use of wired communication in industrial control systems. This paper introduces a wireless communication protocol that capitalizes on multiuser diversity and cooperative communication to achieve the ultra-reliability with a low-latency constraint. Our protocol is analyzed using the communication-theoretic delay-limited-capacity framework and compared to baseline schemes that primarily exploit frequency diversity. For a scenario inspired by an industrial printing application with thirty nodes in the control loop, 20B messages transmitted between pairs of nodes and a cycle time of $2$ ms, an idealized protocol can achieve a cycle failure probability (probability that any packet in a cycle is not successfully delivered) lower than $10^{-9}$ with nominal SNR below 5 dB in a 20MHz wide channel. △ Less

Submitted 23 January, 2017; v1 submitted 9 September, 2016; originally announced September 2016.

Comments: A preliminary version of this work appeared at IEEE International Conference on Communications 2015

Optimal positron-beam excited plasma wakefields in Hollow and Ion-Wake channels

Authors: Aakash A. Sahai, T. C. Katsouleas

Abstract: A positron-beam interacting with the plasma electrons drives radial suck-in, in contrast to an electron-beam driven blow-out in the over-dense regime, $n_b>n_0$. In a homogeneous plasma, the electrons are radially sucked-in from all the different radii. The electrons collapsing from different radii do not simultaneously compress on-axis driving weak fields. A hollow-channel allows electrons from i… ▽ More A positron-beam interacting with the plasma electrons drives radial suck-in, in contrast to an electron-beam driven blow-out in the over-dense regime, $n_b>n_0$. In a homogeneous plasma, the electrons are radially sucked-in from all the different radii. The electrons collapsing from different radii do not simultaneously compress on-axis driving weak fields. A hollow-channel allows electrons from its channel-radius to collapse simultaneously exciting coherent fields. We analyze the optimal channel radius. Additionally, the low ion density in the hollow allows a larger region with focusing phase which we show is linearly focusing. We have shown the formation of an ion-wake channel behind a blow-out electron bubble-wake. Here we explore positron acceleration in the over-dense regime comparing an optimal hollow-plasma channel to the ion-wake channel. The condition for optimal hollow-channel radius is also compared. We also address the effects of a non-ideal ion-wake channel on positron-beam excited fields. △ Less

Submitted 25 December, 2015; originally announced December 2015.

Comments: Proceedings of IPAC2015, Richmond, VA, USA 3: Alternative Particle Sources and Acceleration Techniques A22 - Plasma Wake eld Acceleration http://accelconf.web.cern.ch/AccelConf/IPAC2015/papers/wepje001.pdf, 2015 (ISBN 978-3-95450-168-7) pp 2674-2677

Report number: WEPJE001

Non-linear Ion-wake Excitation by Ultra-relativistic Electron Wakefields

Authors: Aakash A. Sahai, Thomas C. Katsouleas

Abstract: The excitation of a non-linear ion-wake by a train of ultra-relativistic electron bubble wake [1]-[8] is modeled. The ion-wake is shown to be a driven non-linear ion-acoustic wave in the form of a cylindrical ion-soliton [11]-[14]. The phases of the oscillating radial electric fields of the slowly-propagating [21] electron bubble is asymmetric in time and excites time-averaged inertial ion motion… ▽ More The excitation of a non-linear ion-wake by a train of ultra-relativistic electron bubble wake [1]-[8] is modeled. The ion-wake is shown to be a driven non-linear ion-acoustic wave in the form of a cylindrical ion-soliton [11]-[14]. The phases of the oscillating radial electric fields of the slowly-propagating [21] electron bubble is asymmetric in time and excites time-averaged inertial ion motion radially. The electron compression in the back of the bubble sucks-in the ions and the space-charge within the bubble cavity expels them, driving a cylindrical ion-soliton structure with on-axis and bubble-edge density-spikes [5][6]. Once formed, the channel-edge density-spike is driven radially outwards by the thermal pressure of the wake energy [12]. Its channel-like structure due to the flat-residue left behind by the propagating ion-soliton, is independent of the energy-source driving the bubble [3][4] electron wake. We explore the use of the partially- filled channel formed by the cylindrical ion-soliton for a novel regime of positron acceleration [18]. OSIRIS PIC [25] simulations are used to study the ion-wake soliton structure, its driven propagation and its use for positron acceleration. △ Less

Submitted 27 June, 2015; v1 submitted 14 April, 2015; originally announced April 2015.

Longitudinal instabilities affecting the moving critical layer laser-plasma ion accelerators

Authors: Aakash Ajit Sahai, Thomas C. Katsouleas

Abstract: In this work we analyze the longitudinal instabilities of propagating acceleration structures that are driven by a relativistically intense laser at the moving plasma critical layer [1]. These instabilities affect the energy-spectra of the accelerated ion-beams in propagating critical layer acceleration schemes [2][3]. Specifically, using analytical theory and PIC simulations we look into three fu… ▽ More In this work we analyze the longitudinal instabilities of propagating acceleration structures that are driven by a relativistically intense laser at the moving plasma critical layer [1]. These instabilities affect the energy-spectra of the accelerated ion-beams in propagating critical layer acceleration schemes [2][3]. Specifically, using analytical theory and PIC simulations we look into three fundamental physical processes and their interplay that are crucial to the understanding of energy spectral control by making the laser-plasma ion accelerators stable. The interacting processes are (i) Doppler-shifted ponderomotive bunching [1][4] (ii) potential quenching by beam-loading [2] and (iii) two-stream instabilities. These phenomenon have been observed in simulations analyzing these acceleration processes [5][6][7]. From the preliminary models and results we present in this work, we can infer measures by which these instabilities can be controlled [8] for improving the energy-spread of the beams. △ Less

Submitted 10 November, 2014; originally announced November 2014.

Comments: submitted to the proceeding of the Advanced Accelerator Concepts workshop July 2014, San Jose, CA, USA

Journal ref: American Institute of Physics, 2014

Self-injection by trap** of plasma electrons oscillating in rising density gradient at the vacuum-plasma interface

Authors: Aakash A. Sahai, Thomas C. Katsouleas, Patric Muggli

Abstract: We model the trap** of plasma $e^-$ within the density structures excited by a propagating energy source ($β_{S}\simeq1$) in a rising plasma density gradient. Rising density gradient leads to spatially contiguous coupled up-chirped plasmons ($d{ω^2_{pe}(x)}/{dx}>0$). Therefore phase mixing between plasmons can lead to trap** until the plasmon field is high enough such that $e^-$ trajectories r… ▽ More We model the trap** of plasma $e^-$ within the density structures excited by a propagating energy source ($β_{S}\simeq1$) in a rising plasma density gradient. Rising density gradient leads to spatially contiguous coupled up-chirped plasmons ($d{ω^2_{pe}(x)}/{dx}>0$). Therefore phase mixing between plasmons can lead to trap** until the plasmon field is high enough such that $e^-$ trajectories returning towards a longer wavelength see a trap** potential. Rising plasma density gradients are ubiquitous for confining the plasma within sources at the vacuum-plasma interfaces. Therefore trap** of plasma-$e^-$ in a rising ramp is important for acceleration diagnostics and to understand the energy dissipation from the excited plasmon train \cite{LTE-2013}. Down-ramp in density \cite{density-transition-2001} has been used for plasma-$e^-$ trap** within the first bucket behind the driver. Here, in rising density gradient the trap** does not occur in the first plasmon bucket but in subsequent plasmon buckets behind the driver. Trap** reduces the Hamiltonian of each bucket where $e^-$ are trapped, so it is a wakefield-decay probe. Preliminary computational results for beam and laser-driven wakefield are shown. △ Less

Submitted 12 July, 2014; originally announced July 2014.

Comments: Proceedings of International Particle Accelerator Conference, IPAC 2014, Dresden, Germany, June 2014, http://accelconf.web.cern.ch/AccelConf/IPAC2014/papers/tupme051.pdf

Report number: ISBN 978-3-95450-132-8, 03 - Particle Sources and Alternative Acceleration Techniques, A22 - Plasma Wakefield Acceleration

Evaluation of Machine Learning Techniques for Green Energy Prediction

Authors: Ankur Sahai

Abstract: We evaluate the following Machine Learning techniques for Green Energy (Wind, Solar) Prediction: Bayesian Inference, Neural Networks, Support Vector Machines, Clustering techniques (PCA). Our objective is to predict green energy using weather forecasts, predict deviations from forecast green energy, find correlation amongst different weather parameters and green energy availability, recover lost o… ▽ More We evaluate the following Machine Learning techniques for Green Energy (Wind, Solar) Prediction: Bayesian Inference, Neural Networks, Support Vector Machines, Clustering techniques (PCA). Our objective is to predict green energy using weather forecasts, predict deviations from forecast green energy, find correlation amongst different weather parameters and green energy availability, recover lost or missing energy (/ weather) data. We use historical weather data and weather forecasts for the same. △ Less

Submitted 14 June, 2014; originally announced June 2014.

Refraction of $e^-$ beams due to plasma lensing at a plasma-vacuum interface -- applied to beam deflection in a Copper cell with electrical RF-breakdown plasma

Authors: Aakash A. Sahai, T. C. Katsouleas

Abstract: We formulate a possible description of the deflection of a relativistic $e^-$ beam in an inhomogeneous copper plasma, encountered by the beam when propagating through a accelerating cell that has undergone a high electric-field RF-breakdown. It is well known that an inhomogeneous plasma forms and may last for up to a few micro-seconds, until recombination in an accelerating structure where a field… ▽ More We formulate a possible description of the deflection of a relativistic $e^-$ beam in an inhomogeneous copper plasma, encountered by the beam when propagating through a accelerating cell that has undergone a high electric-field RF-breakdown. It is well known that an inhomogeneous plasma forms and may last for up to a few micro-seconds, until recombination in an accelerating structure where a field-emission triggers melting and ionization of RF-cell wall deformity. We present a preliminary model for the beam deflection due to collective plasma response based upon the beam density, plasma density and interaction length. △ Less

Submitted 16 May, 2014; originally announced May 2014.

Comments: presented to Compact Linear Collider (CLIC) team at CERN, Geneva, Switzerland on 29 August 2013, http://indico.cern.ch/event/269506/material/slides/2?contribId=2

Proton acceleration by a relativistic laser frequency-chirp driven plasma snowplow

Authors: Aakash A. Sahai, T. C. Katsouleas, R. A. Bingham, F. S. Tsung, A. R. Tableman, M. Tzoufras, W. B. Mori

Abstract: We analyze the use of a relativistic laser pulse with a controlled frequency chirp incident on a rising plasma density gradient to drive an acceleration structure for proton and light-ion acceleration. The Chirp Induced Transparency Acceleration (ChITA) scheme is described with an analytical model of the velocity of the snowplow at critical density on a pre-formed rising plasma density gradient th… ▽ More We analyze the use of a relativistic laser pulse with a controlled frequency chirp incident on a rising plasma density gradient to drive an acceleration structure for proton and light-ion acceleration. The Chirp Induced Transparency Acceleration (ChITA) scheme is described with an analytical model of the velocity of the snowplow at critical density on a pre-formed rising plasma density gradient that is driven by a positive-chirp in the frequency of a relativistic laser pulse. The velocity of the ChITA-snowplow is shown to depend upon rate of rise of the frequency of the relativistic laser pulse represented by $\frac{ε_0}θ$ where, $ε_0 = \frac{Δω_0}{ω_0}$ and chir** spatial scale-length, $θ$, the normalized magnetic vector potential of the laser pulse $a_0$ and the plasma density gradient scale-length, $α$. We observe using 1-D OSIRIS simulations the formation and forward propagation of ChITA-snowplow, being continuously pushed by the chir** laser at a velocity in accordance with the analytical results. The trace protons reflect off of this propagating snowplow structure and accelerate mono-energetically. The control over ChITA-snowplow velocity allows the tuning of accelerated proton energies. △ Less

Submitted 16 May, 2014; originally announced May 2014.

Comments: Proceedings of IPAC2012, New Orleans, Louisiana, USA, pp.2654-2656, 07 Accelerator Technology and Main Systems, T25 Lasers, ISBN 978-3-95450-115-1. http://accelconf.web.cern.ch/accelconf/ipac2012/papers/weppd059.pdf

Journal ref: Proceedings of IPAC2012, pp.2654-2656, 07 Accelerator Technology and Main Systems, T25 Lasers, ISBN 978-3-95450-115-1

Long Term Evolution of Plasma Wakefields

Authors: Aakash A. Sahai, T. C. Katsouleas, F. S. Tsung, W. B. Mori

Abstract: We study the long-term evolution (LTE) of plasma wakefields over multiple plasma-electron periods and few plasma-ion periods, much less than a recombination time. The evolution and relaxation of such a wakefield-perturbed plasma over these timescales has important implications for the upper limits of repetition-rates in plasma colliders. Intense fields in relativistic lasers (or intense beams) cre… ▽ More We study the long-term evolution (LTE) of plasma wakefields over multiple plasma-electron periods and few plasma-ion periods, much less than a recombination time. The evolution and relaxation of such a wakefield-perturbed plasma over these timescales has important implications for the upper limits of repetition-rates in plasma colliders. Intense fields in relativistic lasers (or intense beams) create plasma wakefields (modes around ωpe) by transferring energy to the plasma electrons. Charged-particle beams in the right phase may be accelerated with acceleration/focusing gradients of tens of GeV/m. However, wakefields leave behind a plasma not in equilibrium, with a relaxation time of multiple plasma-electron periods. Ion motion over ion timescales, caused by energy transfer from the driven plasma-electrons to the plasma-ions can create interesting plasma states. Eventually during LTE, the dynamics of plasma de-coheres (multiple modes through instability driven mixing), thermalizing into random motion (second law of thermodynamics), dissipating energy away from the wakefields. Wakefield-drivers interacting with such a relativistically hot-plasma lead to plasma wakefields that differ from the wakefields in a cold-plasma. △ Less

Submitted 16 May, 2014; originally announced May 2014.

Comments: North American Particle Accelerator Conference, Sep 2013, Pasadena, CA, USA (MOPAC10, ISBN 978-3-95450-138-0) http://accelconf.web.cern.ch/accelconf/pac2013/papers/mopac10.pdf 03- Alternative Acceleration Schemes, A23 - Laser-driven Plasma Acceleration, pp.90-92

Motion of the Plasma Critical Layer During Relativistic-electron Laser Interaction with Immobile and Comoving Ion Plasma for Ion Acceleration

Authors: Aakash A. Sahai

Abstract: We analyze the motion of the plasma critical layer by two different processes in the relativistic-electron laser-plasma interaction regime ($a_0>1$). The differences are highlighted when the critical layer ions are stationary in contrast to when they move with it. Controlling the speed of the plasma critical layer in this regime is essential for creating low-$β$ traveling acceleration structures o… ▽ More We analyze the motion of the plasma critical layer by two different processes in the relativistic-electron laser-plasma interaction regime ($a_0>1$). The differences are highlighted when the critical layer ions are stationary in contrast to when they move with it. Controlling the speed of the plasma critical layer in this regime is essential for creating low-$β$ traveling acceleration structures of sufficient laser-excited potential for laser ion accelerators (LIA). In Relativistically Induced Transparency Acceleration (RITA) scheme the heavy plasma-ions are fixed and only trace-density light-ions are accelerated. The relativistic critical layer and the acceleration structure move longitudinally forward by laser inducing transparency through apparent relativistic increase in electron mass. In the Radiation Pressure Acceleration (RPA) scheme the whole plasma is longitudinally pushed forward under the action of the laser radiation pressure, possible only when plasma ions co-propagate with the laser front. In RPA the acceleration structure velocity critically depends upon plasma-ion mass in addition to the laser intensity and plasma density. In RITA, mass of the heavy immobile plasma-ions does not affect the speed of the critical layer. Inertia of the bared immobile ions in RITA excites the charge separation potential whereas RPA is not possible when ions are stationary. △ Less

Submitted 31 March, 2014; originally announced March 2014.

Comments: Invited paper (submitted), Division of Plasma Physics, American Physical Society, Nov 2013, Denver, CO

Report number: Paper GI2 6, Bull. Am. Phys. Soc. 58, 104 (2013)

Journal ref: Physics of Plasma, Vol.21, iss. 5, 056707 (2014)

Renewable Energy Prediction using Weather Forecasts for Optimal Scheduling in HPC Systems

Authors: Ankur Sahai

Abstract: The objective of the GreenPAD project is to use green energy (wind, solar and biomass) for powering data-centers that are used to run HPC jobs. As a part of this it is important to predict the Renewable (Wind) energy for efficient scheduling (executing jobs that require higher energy when there is more green energy available and vice-versa). For predicting the wind energy we first analyze the hist… ▽ More The objective of the GreenPAD project is to use green energy (wind, solar and biomass) for powering data-centers that are used to run HPC jobs. As a part of this it is important to predict the Renewable (Wind) energy for efficient scheduling (executing jobs that require higher energy when there is more green energy available and vice-versa). For predicting the wind energy we first analyze the historical data to find a statistical model that gives relation between wind energy and weather attributes. Then we use this model based on the weather forecast data to predict the green energy availability in the future. Using the green energy prediction obtained from the statistical model we are able to precompute job schedules for maximizing the green energy utilization in the future. We propose a model which uses live weather data in addition to machine learning techniques (which can predict future deviations in weather conditions based on current deviations from the forecast) to make on-the-fly changes to the precomputed schedule (based on green energy prediction). For this we first analyze the data using histograms and simple statistical tools such as correlation. In addition we build (correlation) regression model for finding the relation between wind energy availability and weather attributes (temperature, cloud cover, air pressure, wind speed / direction, precipitation and sunshine). We also analyze different algorithms and machine learning techniques for optimizing the job schedules for maximizing the green energy utilization. △ Less

Submitted 26 February, 2014; originally announced February 2014.

VM Power Prediction in Distributed Systems for Maximizing Renewable Energy Usage

Authors: Ankur Sahai

Abstract: In the context of GreenPAD project it is important to predict the energy consumption of individual (and mixture of) VMs / workload for optimal scheduling (running those VMs which require higher energy when there is more green energy available and vice-versa) in order to maximize green energy utilization. For this we execute the following experiments on an Openstack cloud testbed consisting of Fu… ▽ More In the context of GreenPAD project it is important to predict the energy consumption of individual (and mixture of) VMs / workload for optimal scheduling (running those VMs which require higher energy when there is more green energy available and vice-versa) in order to maximize green energy utilization. For this we execute the following experiments on an Openstack cloud testbed consisting of Fujitsu servers: VM energy measurement for different configurations (flavor + workload) and VM energy prediction for a new configuration. The automation framework for running these experiments uses bash scripts which call tools like 'stress' (simulating workloads), 'collected' (resource usage) and 'IPMI' (power measurement). We propose a linear model for predicting the power usage of the VMs based on regression. We first collect the resource usage (using collected) and the associated power usage (using IPMI) for different VM configurations and use this to build a (multi-) regression model (between resource usage and VM energy consumption). Then we use the information about the resource usage patterns of the new workload to predict the power usage. For predicting power for mix of workloads we execute (build a regression model based on) experiments with random workloads. We observe the highest energy usage for CPU-intensive workloads followed by memory-intensive workloads. △ Less

Submitted 23 February, 2014; originally announced February 2014.

Relativistically Induced Transparency Acceleration (RITA) of Protons and Light-ions with Ultrashort Laser Interaction with Heavy-ion Plasma Density Gradient

Authors: Aakash A. Sahai, F. S. Tsung, A. R. Tableman, W. B. Mori, T. C. Katsouleas

Abstract: The relativistically induced transparency acceleration (RITA) scheme of proton and ion acceleration using laser-plasma interactions is introduced, modeled, and compared to the existing schemes. Protons are accelerated with femtosecond relativistic pulses to produce quasimonoenergetic bunches with controllable peak energy. The RITA scheme works by a relativistic laser inducing transparency to densi… ▽ More The relativistically induced transparency acceleration (RITA) scheme of proton and ion acceleration using laser-plasma interactions is introduced, modeled, and compared to the existing schemes. Protons are accelerated with femtosecond relativistic pulses to produce quasimonoenergetic bunches with controllable peak energy. The RITA scheme works by a relativistic laser inducing transparency to densities higher than the cold-electron critical density, while the background heavy ions are stationary. The rising laser pulse creates a traveling acceleration structure at the relativistic critical density by ponderomotively driving a local electron density inflation, creating an electron snowplow and a co-propagating electrostatic potential. The snowplow advances with a velocity determined by the rate of the rise of the laser's intensity envelope and the heavy-ion-plasma density gradient scale length. The rising laser is incrementally rendered transparent to higher densities such that the relativistic-electron plasma frequency is resonant with the laser frequency. In the snowplow frame, trace density protons reflect off the electrostatic potential and get snowplowed, while the heavier background ions are relatively unperturbed. Quasimonoenergetic bunches of velocity equal to twice the snowplow velocity can be obtained and tuned by controlling the snowplow velocity using laser-plasma parameters. An analytical model for the proton energy as a function of laser intensity, rise time, and plasma density gradient is developed and compared to 1D and 2D PIC OSIRIS simulations. We model the acceleration of protons to GeV energies with tens-of-femtoseconds laser pulses of a few petawatts. The scaling of proton energy with laser power compares favorably to other mechanisms for ultrashort pulses. △ Less

Submitted 21 February, 2014; originally announced February 2014.

Journal ref: Physical Review E 88, 043105 Published 28 October 2013

Adaptive Protocols for Interactive Communication

Authors: Shweta Agrawal, Ran Gelles, Amit Sahai

Abstract: How much adversarial noise can protocols for interactive communication tolerate? This question was examined by Braverman and Rao (IEEE Trans. Inf. Theory, 2014) for the case of "robust" protocols, where each party sends messages only in fixed and predetermined rounds. We consider a new class of non-robust protocols for Interactive Communication, which we call adaptive protocols. Such protocols ada… ▽ More How much adversarial noise can protocols for interactive communication tolerate? This question was examined by Braverman and Rao (IEEE Trans. Inf. Theory, 2014) for the case of "robust" protocols, where each party sends messages only in fixed and predetermined rounds. We consider a new class of non-robust protocols for Interactive Communication, which we call adaptive protocols. Such protocols adapt structurally to the noise induced by the channel in the sense that both the order of speaking, and the length of the protocol may vary depending on observed noise. We define models that capture adaptive protocols and study upper and lower bounds on the permissible noise rate in these models. When the length of the protocol may adaptively change according to the noise, we demonstrate a protocol that tolerates noise rates up to $1/3$. When the order of speaking may adaptively change as well, we demonstrate a protocol that tolerates noise rates up to $2/3$. Hence, adaptivity circumvents an impossibility result of $1/4$ on the fraction of tolerable noise (Braverman and Rao, 2014). △ Less

Submitted 7 August, 2015; v1 submitted 15 December, 2013; originally announced December 2013.

Comments: Content is similar to previous version yet with an improved presentation