Search

We consider the problem of stragglers in distributed computing systems. Stragglers, which are compute nodes that unpredictably slow down, often increase the completion times of tasks. One common approach to mitigating stragglers is work replication, where only the first completion among replicated tasks is accepted, discarding the others. However, discarding work leads to resource wastage. In this paper, we propose a method for exploiting the work completed by stragglers rather than discarding it. The idea is to increase the granularity of the assigned work, and to increase the frequency of worker updates. We show that the proposed method reduces the completion time of tasks via experiments performed on a simulated cluster as well as on Amazon EC2 with Apache Hadoop., Comment: This paper has been accepted for publication in IEEE Transactions on Services Computing. The initial results presented in this paper appeared in the proceedings of the Allerton Conference on Communication, Control, and Computing in 2023

For applications in concatenated coding for optical communications systems, we examine soft-demapping of short spherical codes constructed as constant-energy shells of the Cartesian power of pulse amplitude modulation constellations. These are unions of permutation codes having the same average power. We construct a list decoder for permutation codes by adapting Murty's algorithm, which is then used to determine mutual information curves for these permutation codes. In the process, we discover a straightforward expression for determining the likelihood of large subcodes of permutation codes called orbits. We introduce a simple process, called orbit demapping, that allows us to extract soft information from noisy permutation codewords. In a sample communication system with probabilistic amplitude shaping protected by a standard low-density parity-check code that employs short permutation codes, we demonstrate that orbit demapping provides a gain of about 0.3 dB in signal-to-noise ratio compared to the traditional symbol-by-symbol demapping. By using spherical codes composed of unions of permutation codes, we can increase the input entropy compared to using permutation codes alone. In one scheme, we consider a union of a small number of permutation codes. In this case, orbit demapping provides about 0.2 dB gain compared to the traditional method. In another scheme, we use all possible permutations to form a spherical code that exhibits a computationally feasible trellis representation. The soft information obtained using the BCJR algorithm outperforms the traditional symbol-by-symbol method by 0.1 dB. Using the spherical codes containing all possible permutation codes of the same average power and the BCJR algorithm, a gain of 0.5 dB is observed. Comparison of the achievable information rates of bit-metric decoding verifies the observed gains.

We consider the controllability of large-scale linear networked dynamical systems when complete knowledge of network structure is unavailable and knowledge is limited to coarse summaries. We provide conditions under which average controllability of the fine-scale system can be well approximated by average controllability of the (synthesized, reduced-order) coarse-scale system. To this end, we require knowledge of some inherent parametric structure of the fine-scale network that makes this type of approximation possible. Therefore, we assume that the underlying fine-scale network is generated by the stochastic block model (SBM) -- often studied in community detection. We then provide an algorithm that directly estimates the average controllability of the fine-scale system using a coarse summary of SBM. Our analysis indicates the necessity of underlying structure (e.g., in-built communities) to be able to quantify accurately the controllability from coarsely characterized networked dynamics. We also compare our method to that of the reduced-order method and highlight the regimes where both can outperform each other. Finally, we provide simulations to confirm our theoretical results for different scalings of network size and density, and the parameter that captures how much community-structure is retained in the coarse summary.

Elasticity is offered by cloud service providers to exploit under-utilized computing resources. The low-cost elastic nodes can leave and join any time during the computation cycle. The possibility of elastic events occurring together with the problem of slow nodes, referred to as stragglers, increases the uncertainty of the system, leading to computation delay. Recent results have shown that coded computing can be used to reduce the negative effect of elasticity and stragglers. In this paper, we propose two hierarchical coded elastic computing schemes that can further speed up the system by exploiting stragglers and effectively allocating tasks among available nodes. In our simulations, our scheme realizes 45% improvement in average finishing time compared to the state-of-the-art coded elastic computing scheme.

Resistive memories store information in a crossbar arrangement of two-terminal devices that can be programmed to patterns of high or low resistance. While extremely compact, this technology suffers from the "sneak-path" problem: certain information patterns cannot be recovered, as multiple low resistances in parallel make a high resistance indistinguishable from a low resistance. In this paper, a multi-layer device is considered, and the number of bits it can store is derived exactly and asymptotic bounds are developed. The information density of a series of isolated arrays with extreme aspect ratios is derived in the single- and multi-layer cases with and without peripheral selection circuitry. This density is shown to be non-zero in the limit, unlike that of the arrays with moderate aspect ratios previously considered. A simple encoding scheme that achieves capacity asymptotically is presented.

Coded distributed computing was recently introduced to mitigate the effect of stragglers on distributed computing. This paper combines ideas of approximate computing with coded computing to further accelerate computation. We propose successive approximation coding (SAC) techniques that realize a tradeoff between accuracy and speed, allowing the distributed computing system to produce approximations that increase in accuracy over time. If a sufficient number of compute nodes finish their tasks, SAC exactly recovers the desired computation. We theoretically provide design guidelines for our SAC techniques, and numerically show that SAC achieves a better accuracy-speed tradeoff in comparison with previous methods.

An increasing bottleneck in decentralized optimization is communication. Bigger models and growing datasets mean that decentralization of computation is important and that the amount of information exchanged is quickly growing. While compression techniques have been introduced to cope with the latter, none has considered leveraging the temporal correlations that exist in consecutive vector updates. An important example is distributed momentum-SGD where temporal correlation is enhanced by the low-pass-filtering effect of applying momentum. In this paper we design and analyze compression methods that exploit temporal correlation in systems both with and without error-feedback. Experiments with the ImageNet dataset demonstrate that our proposed methods offer significant reduction in the rate of communication at only a negligible increase in computation complexity. We further analyze the convergence of SGD when compression is applied with error-feedback. In the literature, convergence guarantees are developed only for compressors that provide error-bounds point-wise, i.e., for each input to the compressor. In contrast, many important codes (e.g. rate-distortion codes) provide error-bounds only in expectation and thus provide a more general guarantee. In this paper we prove the convergence of SGD under an expected error assumption by establishing a bound for the minimum gradient norm., Comment: This paper was presented in part at the 11th International Symposium on Topics in Coding (ISTC), Montreal, QC, Canada, August 2021, and the paper has been accepted for publication in the IEEE Journal on Selected Areas in Information Theory (JSAIT) Volume 2, Issue 3 (2021), https://ieeexplore.ieee.org/document/9511618

We leverage state-of-the-art machine learning methods and a decade's worth of archival data from CFHT to predict observatory image quality (IQ) from environmental conditions and observatory operating parameters. Specifically, we develop accurate and interpretable models of the complex dependence between data features and observed IQ for CFHT's wide-field camera, MegaCam. Our contributions are several-fold. First, we collect, collate and reprocess several disparate data sets gathered by CFHT scientists. Second, we predict probability distribution functions (PDFs) of IQ and achieve a mean absolute error of $\sim0.07''$ for the predicted medians. Third, we explore the data-driven actuation of the 12 dome "vents" installed in 2013-14 to accelerate the flushing of hot air from the dome. We leverage epistemic and aleatoric uncertainties in conjunction with probabilistic generative modeling to identify candidate vent adjustments that are in-distribution (ID); for the optimal configuration for each ID sample, we predict the reduction in required observing time to achieve a fixed SNR. On average, the reduction is $\sim12\%$. Finally, we rank input features by their Shapley values to identify the most predictive variables for each observation. Our long-term goal is to construct reliable and real-time models that can forecast optimal observatory operating parameters to optimize IQ. We can then feed such forecasts into scheduling protocols and predictive maintenance routines. We anticipate that such approaches will become standard in automating observatory operations and maintenance by the time CFHT's successor, the Maunakea Spectroscopic Explorer, is installed in the next decade., Comment: 25 pages and 12 figures (main) + 7 pages and 8 figures (2 appendices). Accepted for publication in MNRAS

We study the problem of community recovery from coarse measurements of a graph. In contrast to the problem of community recovery of a fully observed graph, one often encounters situations when measurements of a graph are made at low-resolution, each measurement integrating across multiple graph nodes. Such low-resolution measurements effectively induce a coarse graph with its own communities. Our objective is to develop conditions on the graph structure, the quantity, and properties of measurements, under which we can recover the community organization in this coarse graph. In this paper, we build on the stochastic block model by mathematically formalizing the coarsening process, and characterizing its impact on the community members and connections. Through this novel setup and modeling, we characterize an error bound for community recovery. The error bound yields simple and closed-form asymptotic conditions to achieve the perfect recovery of the coarse graph communities.

Distributed optimization is widely deployed in practice to solve a broad range of problems. In a typical asynchronous scheme, workers calculate gradients with respect to out-of-date optimization parameters while the master uses stale (i.e., delayed) gradients to update the parameters. While using stale gradients can slow the convergence, asynchronous methods speed up the overall optimization with respect to wall clock time by allowing more frequent updates and reducing idling times. In this paper, we present a variable per-epoch minibatch scheme called Anytime Minibatch with Delayed Gradients (AMB-DG). In AMB-DG, workers compute gradients in epochs of a fixed time while the master uses stale gradients to update the optimization parameters. We analyze AMB-DG in terms of its regret bound and convergence rate. We prove that for convex smooth objective functions, AMB-DG achieves the optimal regret bound and convergence rate. We compare the performance of AMB-DG with that of Anytime Minibatch (AMB) which is similar to AMB-DG but does not use stale gradients. In AMB, workers stay idle after each gradient transmission to the master until they receive the updated parameters from the master while in AMB-DG workers never idle. We also extend AMB-DG to the fully distributed setting. We compare AMB-DG with AMB when the communication delay is long and observe that AMB-DG converges faster than AMB in wall clock time. We also compare the performance of AMB-DG with the state-of-the-art fixed minibatch approach that uses delayed gradients. We run our experiments on a real distributed system and observe that AMB-DG converges more than two times., Comment: Accepted for publication in the IEEE Transaction on Signal and Information Processing over Networks

Intelligent scheduling of the sequence of scientific exposures taken at ground-based astronomical observatories is massively challenging. Observing time is over-subscribed and atmospheric conditions are constantly changing. We propose to guide observatory scheduling using machine learning. Leveraging a 15-year archive of exposures, environmental, and operating conditions logged by the Canada-France-Hawaii Telescope, we construct a probabilistic data-driven model that accurately predicts image quality. We demonstrate that, by optimizing the opening and closing of twelve vents placed on the dome of the telescope, we can reduce dome-induced turbulence and improve telescope image quality by (0.05-0.2 arc-seconds). This translates to a reduction in exposure time (and hence cost) of $\sim 10-15\%$. Our study is the first step toward data-based optimization of the multi-million dollar operations of current and next-generation telescopes., Comment: 4 pages, 3 figures. Accepted to Machine Learning and the Physical Sciences Workshop at the 34th Conference on Neural Information Processing Systems (NeurIPS)

Data-driven graph learning models a network by determining the strength of connections between its nodes. The data refers to a graph signal which associates a value with each graph node. Existing graph learning methods either use simplified models for the graph signal, or they are prohibitively expensive in terms of computational and memory requirements. This is particularly true when the number of nodes is high or there are temporal changes in the network. In order to consider richer models with a reasonable computational tractability, we introduce a graph learning method based on representation learning on graphs. Representation learning generates an embedding for each graph node, taking the information from neighbouring nodes into account. Our graph learning method further modifies the embeddings to compute the graph similarity matrix. In this work, graph learning is used to examine brain networks for brain state identification. We infer time-varying brain graphs from an extensive dataset of intracranial electroencephalographic (iEEG) signals from ten patients. We then apply the graphs as input to a classifier to distinguish seizure vs. non-seizure brain states. Using the binary classification metric of area under the receiver operating characteristic curve (AUC), this approach yields an average of 9.13 percent improvement when compared to two widely used brain network modeling methods.

Motivated by the emerging area of graph signal processing (GSP), we introduce a novel method to draw inference from spatiotemporal signals. Data acquisition in different locations over time is common in sensor networks, for diverse applications ranging from object tracking in wireless networks to medical uses such as electroencephalography (EEG) signal processing. In this paper we leverage novel techniques of GSP to develop a hierarchical feature extraction approach by mapping the data onto a series of spatiotemporal graphs. Such a model maps signals onto vertices of a graph and the time-space dependencies among signals are modeled by the edge weights. Signal components acquired from different locations and time often have complicated functional dependencies. Accordingly, their corresponding graph weights are learned from data and used in two ways. First, they are used as a part of the embedding related to the topology of graph, such as density. Second, they provide the connectivities of the base graph for extracting higher level GSP-based features. The latter include the energies of the signal's graph Fourier transform in different frequency bands. We test our approach on the intracranial EEG (iEEG) data set of the Kaggle epileptic seizure detection contest. In comparison to the winning code, the results show a slight net improvement and up to 6 percent improvement in per subject analysis, while the number of features are decreased by 75 percent on average.

In this paper we propose a new framework for distributed source coding of structured sources, such as sparse signals. Our framework capitalizes on recent advances in the theory of linear inverse problems and signal representations using incoherent projections. Our approach acquires and quantizes incoherent linear measurements of the signal, which are represented as separate bitplanes. Each bitplane is coded using a distributed source code of the appropriate rate, and transmitted. The decoder, starts from the least significant biplane and, using a prediction of the signal as side information, iteratively recovers each bitplane based on the source prediction and the signal, assuming all the previous bitplanes of lower significance have already been recovered. We provide theoretical results guiding the rate selection, relying only on the least squares prediction error of the source. This is in contrast to existing approaches which rely on difficult-to-estimate information-theoretic metrics to set the rate. We validate our approach using simulations on remote-sensing multispectral images, comparing them with existing approaches of similar complexity., Comment: 16 pages, 8 figures

Distributed optimization is vital in solving large-scale machine learning problems. A widely-shared feature of distributed optimization techniques is the requirement that all nodes complete their assigned tasks in each computational epoch before the system can proceed to the next epoch. In such settings, slow nodes, called stragglers, can greatly slow progress. To mitigate the impact of stragglers, we propose an online distributed optimization method called Anytime Minibatch. In this approach, all nodes are given a fixed time to compute the gradients of as many data samples as possible. The result is a variable per-node minibatch size. Workers then get a fixed communication time to average their minibatch gradients via several rounds of consensus, which are then used to update primal variables via dual averaging. Anytime Minibatch prevents stragglers from holding up the system without wasting the work that stragglers can complete. We present a convergence analysis and analyze the wall time performance. Our numerical results show that our approach is up to 1.5 times faster in Amazon EC2 and it is up to five times faster when there is greater variability in compute node performance., Comment: International Conference on Learning Representations (ICLR), May 2019, New Orleans, LA, USA

In distributed computing systems slow working nodes, known as stragglers, can greatly extend finishing times. Coded computing is a technique that enables straggler-resistant computation. Most coded computing techniques presented to date provide robustness by ensuring that the time to finish depends only on a set of the fastest nodes. However, while stragglers do compute less work than non-stragglers, in real-world commercial cloud computing systems (e.g., Amazon's Elastic Compute Cloud (EC2)) the distinction is often a soft one. In this paper, we develop hierarchical coded computing that exploits the work completed by all nodes, both fast and slow, automatically integrating the potential contribution of each. We first present a conceptual framework to represent the division of work amongst nodes in coded matrix multiplication as a cuboid partitioning problem. This framework allows us to unify existing methods and motivates new techniques. We then develop three methods of hierarchical coded computing that we term bit-interleaved coded computation (BICC), multilevel coded computation (MLCC), and hybrid hierarchical coded computation (HHCC). In this paradigm, each worker is tasked with completing a sequence (a hierarchy) of ordered subtasks. The sequence of subtasks, and the complexity of each, is designed so that partial work completed by stragglers can be used, rather than ignored. We note that our methods can be used in conjunction with any coded computing method. We illustrate this by showing how we can use our methods to accelerate all previously developed coded computing techniques by enabling them to exploit stragglers. Under a widely studied statistical model of completion time, our approach realizes a $66\%$ improvement in the expected finishing time. On Amazon EC2, the gain was $27\%$ when stragglers are simulated.

Coded matrix multiplication is a technique to enable straggler-resistant multiplication of large matrices in distributed computing systems. In this paper, we first present a conceptual framework to represent the division of work amongst processors in coded matrix multiplication as a cuboid partitioning problem. This framework allows us to unify existing methods and motivates new techniques. Building on this framework, we apply the idea of hierarchical coding (Ferdinand & Draper, 2018) to coded matrix multiplication. The hierarchical scheme we develop is able to exploit the work completed by all processors (fast and slow), rather than ignoring the slow ones, even if the amount of work completed by stragglers is much less than that completed by the fastest workers. On Amazon EC2, we achieve a 37% improvement in average finishing time compared to non-hierarchical schemes.

We introduce a semiparametric approach to neighbor-based classification. We build off the recently proposed Boundary Trees algorithm by Mathy et al.(2015) which enables fast neighbor-based classification, regression and retrieval in large datasets. While boundary trees use an Euclidean measure of similarity, the Differentiable Boundary Tree algorithm by Zoran et al.(2017) was introduced to learn low-dimensional representations of complex input data, on which semantic similarity can be calculated to train boundary trees. As is pointed out by its authors, the differentiable boundary tree approach contains a few limitations that prevents it from scaling to large datasets. In this paper, we introduce Differentiable Boundary Sets, an algorithm that overcomes the computational issues of the differentiable boundary tree scheme and also improves its classification accuracy and data representability. Our algorithm is efficiently implementable with existing tools and offers a significant reduction in training time. We test and compare the algorithms on the well known MNIST handwritten digits dataset and the newer Fashion-MNIST dataset by Xiao et al.(2017)., Comment: 9 pages, 2 figures

In cloud computing systems slow processing nodes, often referred to as "stragglers", can significantly extend the computation time. Recent results have shown that error correction coding can be used to reduce the effect of stragglers. In this work we introduce a scheme that, in addition to using error correction to distribute mixed jobs across nodes, is also able to exploit the work completed by all nodes, including stragglers. We first consider vector-matrix multiplication and apply maximum distance separable (MDS) codes to small blocks of sub-matrices. The worker nodes process blocks sequentially, working block-by-block, transmitting partial per-block results to the master as they are completed. Sub-blocking allows a more continuous completion process, which thereby allows us to exploit the work of a much broader spectrum of processors and reduces computation time. We then apply this technique to matrix-matrix multiplication using product code. In this case, we show that the order of computing sub-tasks is a new degree of design freedom that can be exploited to reduce computation time further. We propose a novel approach to analyze the finishing time, which is different from typical order statistics. Simulation results show that the expected computation time decreases by a factor of at least two in compared to previous methods.

We present a hardware-based implementation of Linear Program (LP) decoding for binary linear codes. LP decoding frames error-correction as an optimization problem. In contrast, variants of Belief Propagation (BP) decoding frame error-correction as a problem of graphical inference. LP decoding has several advantages over BP-based methods, including convergence guarantees and better error-rate performance in high-reliability channels. The latter makes LP decoding attractive for optical transport and storage applications. However, LP decoding, when implemented with general solvers, does not scale to large blocklengths and is not suitable for a parallelized implementation in hardware. It has been recently shown that the Alternating Direction Method of Multipliers (ADMM) can be applied to decompose the LP decoding problem. The result is a message-passing algorithm with a structure very similar to BP. We present new intuition for this decoding algorithm as well as for its major computational primitive: projection onto the parity polytope. Furthermore, we present results for a fixed-point Verilog implementation of ADMM-LP decoding. This implementation targets a Field-Programmable Gate Array (FPGA) platform to evaluate error-rate performance and estimate resource usage. We show that Frame Error Rate (FER) performance well within 0.5dB of double-precision implementations is possible with 10-bit messages. Finally, we outline a number of research opportunities that should be explored en-route to the realization of an Application Specific Integrated Circuit (ASIC) implementation capable of gigabit per second throughput.

In this paper we present an FPGA-based implementation of linear programming (LP) decoding. LP decoding frames error correction as an optimization problem. This is in contrast to variants of belief propagation (BP) decoding that view error correction as a problem of graphical inference. There are many advantages to taking the optimization perspective: convergence guarantees, improved performance in certain regimes, and a methodology for incorporating the latest developments in optimization techniques. However, LP decoding, when implemented with standard LP solvers, does not easily scale to the blocklengths of modern error-correction codes. In earlier work, we showed that by drawing on decomposition methods from optimization theory, specifically the alternating direction method of multipliers (ADMM), we could build an LP decoding solver that was competitive with BP, both in terms of performance and speed. We also observed empirically that LP decoders have much better high-SNR performance in the "error floor" regime, a trait of particular relevance to optical transport and storage applications. While our previous implementation was in floating point, in this paper we report initial results of a fixed-point, hardware-based realization of our ADMM-LP decoder., Comment: Submitted for publication to the 2016 IEEE International Workshop on Signal Processing Systems

This paper is concerned with the adaptation to hardware of methods for Euclidean norm projections onto the parity polytope and probability simplex. We first refine recent efforts to develop efficient methods of projection onto the parity polytope. Our resulting algorithm can be configured to have either average computational complexity $\mathcal{O}\left(d\right)$ or worst case complexity $\mathcal{O}\left(d\log{d}\right)$ on a serial processor where $d$ is the dimension of projection space. We show how to adapt our projection routine to hardware. Our projection method uses a sub-routine that involves another Euclidean projection; onto the probability simplex. We therefore explain how to adapt to hardware a well know simplex projection algorithm. The hardware implementations of both projection algorithms achieve area scalings of $\mathcal{O}(d\left(\log{d}\right)^2)$ at a delay of $\mathcal{O}(\left(\log{d}\right)^2)$. Finally, we present numerical results in which we evaluate the fixed-point accuracy and resource scaling of these algorithms when targeting a modern FPGA.

Future servers will incorporate many active lowpower modes for different system components, such as cores and memory. Though these modes provide flexibility for power management via Dynamic Voltage and Frequency Scaling (DVFS), they must be operated in a coordinated manner. Such coordinated control creates a combinatorial space of possible power mode configurations. Given the rapid growth of the number of cores, it is becoming increasingly challenging to quickly select the configuration that maximizes the performance under a given power budget. Prior power capping techniques do not scale well to large numbers of cores, and none of those works has considered memory DVFS. In this paper, we present FastCap, our optimization approach for system-wide power capping, using both CPU and memory DVFS. Based on a queuing model, FastCap formulates power capping as a non-linear optimization problem where we seek to maximize the system performance under a power budget, while promoting fairness across applications. Our FastCap algorithm solves the optimization online and efficiently (low complexity on the number of cores), using a small set of performance counters as input. To evaluate FastCap, we simulate it for a many-core server running different types of workloads. Our results show that FastCap caps power draw accurately, while producing better application performance and fairness than many existing CPU power capping methods (even after they are extended to use of memory DVFS as well)., Comment: Accepted by ISPASS 2016

In this paper, we introduce a new way of constructing and decoding multipermutation codes. Multipermutations are permutations of a multiset that generally consist of duplicate entries. We first introduce a class of binary matrices called multipermutation matrices, each of which corresponds to a unique and distinct multipermutation. By enforcing a set of linear constraints on these matrices, we define a new class of codes that we term LP-decodable multipermutation codes. In order to decode these codes using a linear program (LP), thereby enabling soft decoding, we characterize the convex hull of multipermutation matrices. This characterization allows us to relax the coding constraints to a polytope and to derive two LP decoding problems. These two problems are respectively formulated by relaxing the maximum likelihood decoding problem and the minimum Chebyshev distance decoding problem. Because these codes are non-linear, we also study efficient encoding and decoding algorithms. We first describe an algorithm that maps consecutive integers, one by one, to an ordered list of multipermutations. Based on this algorithm, we develop an encoding algorithm for a code proposed by Shieh and Tsai, a code that falls into our class of LP-decodable multipermutation codes. Regarding decoding algorithms, we propose an efficient distributed decoding algorithm based on the alternating direction method of multipliers (ADMM). Finally, we observe from simulation results that the soft decoding techniques we introduce can significantly outperform hard decoding techniques that are based on quantized channel outputs., Comment: This work was supported by the National Science Foundation (NSF) under Grants CCF-1217058 and by a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Research Grant. This paper was submitted to IEEE Trans. Inf. Theory

In this paper, we introduce a new way of constructing and decoding multipermutation codes. Multipermutations are permutations of a multiset that may consist of duplicate entries. We first introduce a new class of matrices called multipermutation matrices. We characterize the convex hull of multipermutation matrices. Based on this characterization, we propose a new class of codes that we term LP-decodable multipermutation codes. Then, we derive two LP decoding algorithms. We first formulate an LP decoding problem for memoryless channels. We then derive an LP algorithm that minimizes the Chebyshev distance. Finally, we show a numerical example of our algorithm., Comment: This work was supported by NSF and NSERC. To appear at the 2014 Allerton Conference

In this paper, we develop efficient decoders for non-binary low-density parity-check (LDPC) codes using the alternating direction method of multipliers (ADMM). We apply ADMM to two decoding problems. The first problem is linear programming (LP) decoding. In order to develop an efficient algorithm, we focus on non-binary codes in fields of characteristic two. This allows us to transform each constraint in $\mathbb{F}_{2^m}$ to a set of constraints in $\mathbb{F}_{2}$ that has a factor graph representation. Applying ADMM to the LP decoding problem results in two types of non-trivial sub-routines. The first type requires us to solve an unconstrained quadratic program. We solve this problem efficiently by leveraging new results obtained from studying the above factor graphs. The second type requires Euclidean projection onto polytopes that are studied in the literature, a projection that can be solved efficiently using off-the-shelf techniques, which scale linearly in the dimension of the vector to project. ADMM LP decoding scales linearly with block length, linearly with check degree, and quadratically with field size. The second problem we consider is a penalized LP decoding problem. This problem is obtained by incorporating a penalty term into the LP decoding objective. The purpose of the penalty term is to make non-integer solutions (pseudocodewords) more expensive and hence to improve decoding performance. The ADMM algorithm for the penalized LP problem requires Euclidean projection onto a polytope formed by embedding the constraints specified by the non-binary single parity-check code, which can be solved by applying the ADMM technique to the resulting quadratic program. Empirically, this decoder achieves a much reduced error rate than LP decoding at low signal-to-noise ratios., Comment: This work was supported by the National Science Foundation (NSF) under Grants CCF-1217058 and by a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Research Grant. This paper was submitted to IEEE Trans. Inf. Theory

Linear programming (LP) decoding for low-density parity-check (LDPC) codes proposed by Feldman et al. is shown to have theoretical guarantees in several regimes and empirically is not observed to suffer from an error floor. However at low signal-to-noise ratios (SNRs), LP decoding is observed to have worse error performance than belief propagation (BP) decoding. In this paper, we seek to improve LP decoding at low SNRs while still achieving good high SNR performance. We first present a new decoding framework obtained by trying to solve a non-convex optimization problem using the alternating direction method of multipliers (ADMM). This non-convex problem is constructed by adding a penalty term to the LP decoding objective. The goal of the penalty term is to make "pseudocodewords", which are the non-integer vertices of the LP relaxation to which the LP decoder fails, more costly. We name this decoder class the "ADMM penalized decoder". In our simulation results, the ADMM penalized decoder with $\ell_1$ and $\ell_2$ penalties outperforms both BP and LP decoding at all SNRs. For high SNR regimes where it is infeasible to simulate, we use an instanton analysis and show that the ADMM penalized decoder has better high SNR performance than BP decoding. We also develop a reweighted LP decoder using linear approximations to the objective with an $\ell_1$ penalty. We show that this decoder has an improved theoretical recovery threshold compared to LP decoding. In addition, we show that the empirical gain of the reweighted LP decoder is significant at low SNRs., Comment: This work was supported by the National Science Foundation (NSF) under Grants CCF-1217058 and by a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Research Grant. This paper was submitted to IEEE Trans. Inf. Theory

We study a form of unequal error protection that we term "unequal message protection" (UMP). The message set of a UMP code is a union of $m$ disjoint message classes. Each class has its own error protection requirement, with some classes needing better error protection than others. We analyze the tradeoff between rates of message classes and the levels of error protection of these codes. We demonstrate that there is a clear performance loss compared to homogeneous (classical) codes with equivalent parameters. This is in sharp contrast to previous literature that considers UMP codes. To obtain our results we generalize finite block length achievability and converse bounds due to Polyanskiy-Poor-Verd\'{u}. We evaluate our bounds for the binary symmetric and binary erasure channels, and analyze the asymptotic characteristic of the bounds in the fixed error and moderate deviations regimes. In addition, we consider two questions related to the practical construction of UMP codes. First, we study a "header" construction that prefixes the message class into a header followed by data protection using a standard homogeneous code. We show that, in general, this construction is not optimal at finite block lengths. We further demonstrate that our main UMP achievability bound can be obtained using coset codes, which suggests a path to implementation of tractable UMP codes.

Power consumption in data centers has been growing significantly in recent years. To reduce power, servers are being equipped with increasingly sophisticated power management mechanisms. Different mechanisms offer dramatically different trade-offs between power savings and performance penalties. Considering the complexity, variety, and temporally varying nature of the applications hosted in a typical data center, intelligently determining which power management policy to use and when is a complicated task. In this paper we analyze a system model featuring both performance scaling and low-power states. We reveal the interplay between performance scaling and low-power states via intensive simulation and analytic verification. Based on the observations, we present SleepScale, a runtime power management tool designed to efficiently exploit existing power control mechanisms. At run time, SleepScale characterizes power consumption and quality-of-service (QoS) for each low-power state and frequency setting, and selects the best policy for a given QoS constraint. We evaluate SleepScale using workload traces from data centers and achieve significant power savings relative to conventional power management strategies., Comment: Accepted by ACM/IEEE International Symposium on Computer Architecture (ISCA) 2014

BIOMETRICS are an important and widely used class of methods for identity verification and access control. Biometrics are attractive because they are inherent properties of an individual. They need not be remembered like passwords, and are not easily lost or forged like identifying documents. At the same time, bio- metrics are fundamentally noisy and irreplaceable. There are always slight variations among the measurements of a given biometric, and, unlike passwords or identification numbers, biometrics are derived from physical characteristics that cannot easily be changed. The proliferation of biometric usage raises critical privacy and security concerns that, due to the noisy nature of biometrics, cannot be addressed using standard cryptographic methods. In this article we present an overview of "secure biometrics", also referred to as "biometric template protection", an emerging class of methods that address these concerns., Comment: 16 pages, 11 figures, 1 table

In this paper we study the power-performance relationship of power-efficient computing from a queuing theoretic perspective. We investigate the interplay of several system operations including processing speed, system on/off decisions, and server farm size. We identify that there are oftentimes "sweet spots" in power-efficient operations: there exist optimal combinations of processing speed and system settings that maximize power efficiency. For the single server case, a widely deployed threshold mechanism is studied. We show that there exist optimal processing speed and threshold value pairs that minimize the power consumption. This holds for the threshold mechanism with job batching. For the multi-server case, it is shown that there exist best processing speed and server farm size combinations., Comment: Paper published in CISS 2013

This paper investigates generation of a secret key from a reciprocal wireless channel. In particular we consider wireless channels that exhibit sparse structure in the wideband regime and the impact of the sparsity on the secret key capacity. We explore this problem in two steps. First, we study key generation from a state-dependent discrete memoryless multiple source. The state of source captures the effect of channel sparsity. Secondly, we consider a wireless channel model that captures channel sparsity and correlation between the legitimate users' channel and the eavesdropper's channel. Such dependency can significantly reduce the secret key capacity. According to system delay requirements, two performance measures are considered: (i) ergodic secret key capacity and (ii) outage probability. We show that in the wideband regime when a white sounding sequence is adopted, a sparser channel can achieve a higher ergodic secret key rate than a richer channel can. For outage performance, we show that if the users generate secret keys at a fraction of the ergodic capacity, the outage probability will decay exponentially in signal bandwidth. Moreover, a larger exponent is achieved by a richer channel.

When binary linear error-correcting codes are used over symmetric channels, a relaxed version of the maximum likelihood decoding problem can be stated as a linear program (LP). This LP decoder can be used to decode error-correcting codes at bit-error-rates comparable to state-of-the-art belief propagation (BP) decoders, but with significantly stronger theoretical guarantees. However, LP decoding when implemented with standard LP solvers does not easily scale to the block lengths of modern error correcting codes. In this paper we draw on decomposition methods from optimization theory, specifically the Alternating Directions Method of Multipliers (ADMM), to develop efficient distributed algorithms for LP decoding. The key enabling technical result is a "two-slice" characterization of the geometry of the parity polytope, which is the convex hull of all codewords of a single parity check code. This new characterization simplifies the representation of points in the polytope. Using this simplification, we develop an efficient algorithm for Euclidean norm projection onto the parity polytope. This projection is required by ADMM and allows us to use LP decoding, with all its theoretical guarantees, to decode large-scale error correcting codes efficiently. We present numerical results for LDPC codes of lengths more than 1000. The waterfall region of LP decoding is seen to initiate at a slightly higher signal-to-noise ratio than for sum-product BP, however an error floor is not observed for LP decoding, which is not the case for BP. Our implementation of LP decoding using ADMM executes as fast as our baseline sum-product BP decoder, is fully parallelizable, and can be seen to implement a type of message-passing with a particularly simple schedule., Comment: 35 pages, 11 figures. An early version of this work appeared at the 49th Annual Allerton Conference, September 2011. This version to appear in IEEE Transactions on Information Theory

We present a theoretical framework for the analysis of privacy and security tradeoffs in secure biometric authentication systems. We use this framework to conduct a comparative information-theoretic analysis of two biometric systems that are based on linear error correction codes, namely fuzzy commitment and secure sketches. We derive upper bounds for the probability of false rejection ($P_{FR}$) and false acceptance ($P_{FA}$) for these systems. We use mutual information to quantify the information leaked about a user's biometric identity, in the scenario where one or multiple biometric enrollments of the user are fully or partially compromised. We also quantify the probability of successful attack ($P_{SA}$) based on the compromised information. Our analysis reveals that fuzzy commitment and secure sketch systems have identical $P_{FR}, P_{FA}, P_{SA}$ and information leakage, but secure sketch systems have lower storage requirements. We analyze both single-factor (keyless) and two-factor (key-based) variants of secure biometrics, and consider the most general scenarios in which a single user may provide noisy biometric enrollments at several access control devices, some of which may be subsequently compromised by an attacker. Our analysis highlights the revocability and reusability properties of key-based systems and exposes a subtle design tradeoff between reducing information leakage from compromised systems and preventing successful attacks on systems whose data have not been compromised., Comment: 15 pages

We present a general approach to designing capacity-approaching high-girth low-density parity-check (LDPC) codes that are friendly to hardware implementation. Our methodology starts by defining a new class of "hierarchical" quasi-cyclic (HQC) LDPC codes that generalizes the structure of quasi-cyclic (QC) LDPC codes. Whereas the parity check matrices of QC LDPC codes are composed of circulant sub-matrices, those of HQC LDPC codes are composed of a hierarchy of circulant sub-matrices that are in turn constructed from circulant sub-matrices, and so on, through some number of levels. We show how to map any class of codes defined using a protograph into a family of HQC LDPC codes. Next, we present a girth-maximizing algorithm that optimizes the degrees of freedom within the family of codes to yield a high-girth HQC LDPC code. Finally, we discuss how certain characteristics of a code protograph will lead to inevitable short cycles, and show that these short cycles can be eliminated using a "squashing" procedure that results in a high-girth QC LDPC code, although not a hierarchical one. We illustrate our approach with designed examples of girth-10 QC LDPC codes obtained from protographs of one-sided spatially-coupled codes., Comment: Submitted to IEEE Transactions on Information THeory

We study the fundamental limits to secret key generation from an excited distributed source (EDS). In an EDS a pair of terminals observe dependent sources of randomness excited by a pre-arranged signal. We first determine the secret key capacity for such systems with one-way public messaging. We then characterize a tradeoff between the secret key rate and exponential bounds on the probability of key agreement failure and on the secrecy of the key generated. We find that there is a fundamental tradeoff between reliability and secrecy. We then explore this framework within the context of reciprocal wireless channels. In this setting, the users transmit pre-arranged excitation signals to each other. When the fading is Rayleigh, the observations of the users are jointly Gaussian sources. We show that an on-off excitation signal with an SNR-dependent duty cycle achieves the secret key capacity of this system. Furthermore, we characterize a fundamental metric -- minimum energy per key bit for reliable key generation -- and show that in contrast to conventional AWGN channels, there is a non-zero threshold SNR that achieves the minimum energy per key bit. The capacity achieving on-off excitation signal achieves the minimum energy per key bit at any SNR below the threshold. Finally, we build off our error exponent results to investigate the energy required to generate a key using a finite block length. Again we find that on-off excitation signals yield an improvement when compared to constant excitation signals. In addition to Rayleigh fading, we analyze the performance of a system based on binary channel phase quantization., Comment: accepted for publication, IEEE Transactions on Information Theory

In this paper we develop scheduling policies that maximize the stability region of a wireless network under the assumption that mutual information accumulation is implemented at the physical layer. When the link quality between nodes is not sufficiently high that a packet can be decoded within a single slot, the system can accumulate information across multiple slots, eventually decoding the packet. The result is an expanded stability region. The accumulation process over weak links is temporally coupled and therefore does not satisfy the independent and identically distributed (i.i.d) assumption that underlies many previous analysis in this area. Therefore the problem setting also poses new analytic challenges. We propose two dynamic scheduling algorithms to cope with the non-i.i.d nature of the decoding. The first performs scheduling every $T$ slots, and approaches the boundary of the stability region as $T$ gets large, but at the cost of increased average delay. The second introduces virtual queues for each link and constructs a virtual system wherein two virtual nodes are introduced for each link. The constructed virtual system is shown to have the same stability region as the original system. Through controlling the virtual queues in the constructed system, we avoid the non-i.i.d analysis difficulty and attain the full stability region. We derive performance bounds for both algorithms and compare them through simulation results., Comment: submitted to IEEE Trans. on Information Theory, September 2011; revised, May 2012

We consider the resource allocation problem in cooperative wireless networks wherein nodes perform mutual information accumulation. We consider a unicast setting and arbitrary arrival processes at the source node. Source arrivals can be broken down into numerous packets to better exploit the spatial and temporal diversity of the routes available in the network. We devise a linear-program-based algorithm which allocates network resource to meet a certain transmission objective. Given a network, a source with multiple arriving packets and a destination, our algorithm generates a policy that regulates which nodes should participate in transmitting which packets, when and with what resource. By routing different packets through different nodes the policy exploits spatial route diversity, and by sequencing packet transmissions along the same route it exploits temporal route diversity.

We consider the secret key generation problem when sources are randomly excited by the sender and there is a noiseless public discussion channel. Our setting is thus similar to recent works on channels with action-dependent states where the channel state may be influenced by some of the parties involved. We derive single-letter expressions for the secret key capacity through a type of source emulation analysis. We also derive lower bounds on the achievable reliability and secrecy exponents, i.e., the exponential rates of decay of the probability of decoding error and of the information leakage. These exponents allow us to determine a set of strongly-achievable secret key rates. For degraded eavesdroppers the maximum strongly-achievable rate equals the secret key capacity; our exponents can also be specialized to previously known results. In deriving our strong achievability results we introduce a coding scheme that combines wiretap coding (to excite the channel) and key extraction (to distill keys from residual randomness). The secret key capacity is naturally seen to be a combination of both source- and channel-type randomness. Through examples we illustrate a fundamental interplay between the portion of the secret key rate due to each type of randomness. We also illustrate inherent tradeoffs between the achievable reliability and secrecy exponents. Our new scheme also naturally accommodates rate limits on the public discussion. We show that under rate constraints we are able to achieve larger rates than those that can be attained through a pure source emulation strategy., Comment: 18 pages, 8 figures; Submitted to the IEEE Transactions on Information Theory; Revised in Oct 2013

We design and analyze a method to extract secret keys from the randomness inherent to wireless channels. We study a channel model for multipath wireless channel and exploit the channel diversity in generating secret key bits. We compare the key extraction methods based both on entire channel state information (CSI) and on single channel parameter such as the received signal strength indicators (RSSI). Due to the reduction in the degree-of-freedom when going from CSI to RSSI, the rate of key extraction based on CSI is far higher than that based on RSSI. This suggests that exploiting channel diversity and making CSI information available to higher layers would greatly benefit the secret key generation. We propose a key generation system based on low-density parity-check (LDPC) codes and describe the design and performance of two systems: one based on binary LDPC codes and the other (useful at higher signal-to-noise ratios) based on four-ary LDPC codes.

This paper establishes information-theoretic limits in estimating a finite field low-rank matrix given random linear measurements of it. These linear measurements are obtained by taking inner products of the low-rank matrix with random sensing matrices. Necessary and sufficient conditions on the number of measurements required are provided. It is shown that these conditions are sharp and the minimum-rank decoder is asymptotically optimal. The reliability function of this decoder is also derived by appealing to de Caen's lower bound on the probability of a union. The sufficient condition also holds when the sensing matrices are sparse - a scenario that may be amenable to efficient decoding. More precisely, it is shown that if the n\times n-sensing matrices contain, on average, \Omega(nlog n) entries, the number of measurements required is the same as that when the sensing matrices are dense and contain entries drawn uniformly at random from the field. Analogies are drawn between the above results and rank-metric codes in the coding theory literature. In fact, we are also strongly motivated by understanding when minimum rank distance decoding of random rank-metric codes succeeds. To this end, we derive distance properties of equiprobable and sparse rank-metric codes. These distance properties provide a precise geometric interpretation of the fact that the sparse ensemble requires as few measurements as the dense one. Finally, we provide a non-exhaustive procedure to search for the unknown low-rank matrix., Comment: Accepted to the IEEE Transactions on Information Theory; Presented at IEEE International Symposium on Information Theory (ISIT) 2011

Consider the following unequal error protection scenario. One special message, dubbed the "red alert" message, is required to have an extremely small probability of missed detection. The remainder of the messages must keep their average probability of error and probability of false alarm below a certain threshold. The goal then is to design a codebook that maximizes the error exponent of the red alert message while ensuring that the average probability of error and probability of false alarm go to zero as the blocklength goes to infinity. This red alert exponent has previously been characterized for discrete memoryless channels. This paper completely characterizes the optimal red alert exponent for additive white Gaussian noise channels with block power constraints., Comment: 13 pages, 10 figures, To appear in IEEE Transactions on Information Theory

The "Divide and Concur'' (DC) algorithm, recently introduced by Gravel and Elser, can be considered a competitor to the belief propagation (BP) algorithm, in that both algorithms can be applied to a wide variety of constraint satisfaction, optimization, and probabilistic inference problems. We show that DC can be interpreted as a message-passing algorithm on a constraint graph, which helps make the comparison with BP more clear. The "difference-map'' dynamics of the DC algorithm enables it to avoid "traps'' which may be related to the "trapping sets'' or "pseudo-codewords'' that plague BP decoders of low-density parity check (LDPC) codes in the error-floor regime. We investigate two decoders for low-density parity-check (LDPC) codes based on these ideas. The first decoder is based directly on DC, while the second decoder borrows the important "difference-map'' concept from the DC algorithm and translates it into a BP-like decoder. We show that this "difference-map belief propagation'' (DMBP) decoder has dramatically improved error-floor performance compared to standard BP decoders, while maintaining a similar computational complexity. We present simulation results for LDPC codes on the additive white Gaussian noise and binary symmetric channels, comparing DC and DMBP decoders with other decoders based on BP, linear programming, and mixed-integer linear programming.

Cooperation between the nodes of wireless multihop networks can increase communication reliability, reduce energy consumption, and decrease latency. The possible improvements are even greater when nodes perform mutual information accumulation using rateless codes. In this paper, we investigate routing problems in such networks. Given a network, a source, and a destination, our objective is to minimize end-to-end transmission delay under energy and bandwidth constraints. We provide an algorithm that determines which nodes should participate in forwarding the message and what resources (time, energy, bandwidth) should be allocated to each. Our approach factors into two sub-problems, each of which can be solved efficiently. For any transmission order we show that solving for the optimum resource allocation can be formulated as a linear programming problem. We then show that the transmission order can be improved systematically by swapping nodes based on the solution of the linear program. Solving a sequence of linear programs leads to a locally optimal solution in a very efficient manner. In comparison to the proposed cooperative routing solution, it is observed that conventional shortest path multihop routing typically incurs additional delays and energy expenditures on the order of 70%. Our first algorithm is centralized, assuming that routing computations can be done at a central processor with full access to channel state information for the entire system. We also design two distributed routing algorithms that require only local channel state information. We provide simulations showing that for the same networks the distributed algorithms find routes that are only about two to five percent less efficient than the centralized algorithm.