text stringlengths 17 3.36M | source stringlengths 3 333 | __index_level_0__ int64 0 518k |
|---|---|---|
Financing high-tech projects always entails a great deal of risk. The lack of a systematic method to pinpoint the risk of such projects has been recognized as one of the most salient barriers for evaluating them. So, in order to develop a mechanism for evaluating high-tech projects, an Artificial Neural Network (ANN) has been developed through this study. The structure of this paper encompasses four parts. The first part deals with introducing paper's whole body. The second part gives a literature review. The collection process of risk related variables and the process of developing a Risk Assessment Index system (RAIS) through Principal Component Analysis (PCA) are those issues that are discussed in the third part. The fourth part particularly deals with pharmaceutical industry. Finally, the fifth part has focused on developing an ANN for pattern recognition of failure or success of high-tech projects. Analysis of model's results and a final conclusion are also presented in this part. | A Neural Network Model for Determining the Success or Failure of
High-tech Projects Development: A Case of Pharmaceutical industry | 6,800 |
Due to the nonlinearity of artificial neural networks, designing topologies for deep convolutional neural networks (CNN) is a challenging task and often only heuristic approach, such as trial and error, can be applied. An evolutionary algorithm can solve optimization problems where the fitness landscape is unknown. However, evolutionary algorithms are computing resource intensive, which makes it difficult for problems when deep CNNs are involved. In this paper, we propose an evolutionary strategy to find better topologies for deep CNNs. Incorporating the concept of knowledge inheritance and knowledge learning, our evolutionary algorithm can be executed with limited computing resources. We applied the proposed algorithm in finding effective topologies of deep CNNs for the image classification task using CIFAR-10 dataset. After the evolution, we analyzed the topologies that performed well for this task. Our studies verify the techniques that have been commonly used in human designed deep CNNs. We also discovered that some of the graph properties greatly affect the system performance. We applied the guidelines learned from the evolution and designed new network topologies that outperform Residual Net with less layers on CIFAR-10, CIFAR-100, and SVHN dataset. | Finding Better Topologies for Deep Convolutional Neural Networks by
Evolution | 6,801 |
Tuning parameters is an important step for the application of metaheuristics to problem classes of interest. In this work we present a tuning framework based on the sequential optimization of perturbed regression models. Besides providing algorithm configurations with good expected performance, the proposed methodology can also provide insights on the relevance of each parameter and their interactions, as well as models of expected algorithm performance for a given problem class, conditional on the parameter values. A test case is presented for the tuning of six parameters of a decomposition-based multiobjective optimization algorithm, in which an instantiation of the proposed framework is compared against the results obtained by the most recent version the Iterated Racing (Irace) procedure. The results suggest that the proposed approach returns solutions that are as good as those of Irace in terms of mean performance, with the advantage of providing more information on the relevance and effect of each parameter on the expected performance of the algorithm. | Tuning metaheuristics by sequential optimization of regression models | 6,802 |
This paper presents and demonstrates a stochastic logic time delay reservoir design in FPGA hardware. The reservoir network approach is analyzed using a number of metrics, such as kernel quality, generalization rank, performance on simple benchmarks, and is also compared to a deterministic design. A novel re-seeding method is introduced to reduce the adverse effects of stochastic noise, which may also be implemented in other stochastic logic reservoir computing designs, such as echo state networks. Benchmark results indicate that the proposed design performs well on noise-tolerant classification problems, but more work needs to be done to improve the stochastic logic time delay reservoirs robustness for regression problems. In addition, we show that the stochastic design can significantly reduce area cost if the conversion between binary and stochastic representations implemented efficiently. | An FPGA Implementation of a Time Delay Reservoir Using Stochastic Logic | 6,803 |
Spiking neural networks (SNNs) that enables energy efficient implementation on emerging neuromorphic hardware are gaining more attention. Yet now, SNNs have not shown competitive performance compared with artificial neural networks (ANNs), due to the lack of effective learning algorithms and efficient programming frameworks. We address this issue from two aspects: (1) We propose a neuron normalization technique to adjust the neural selectivity and develop a direct learning algorithm for deep SNNs. (2) Via narrowing the rate coding window and converting the leaky integrate-and-fire (LIF) model into an explicitly iterative version, we present a Pytorch-based implementation method towards the training of large-scale SNNs. In this way, we are able to train deep SNNs with tens of times speedup. As a result, we achieve significantly better accuracy than the reported works on neuromorphic datasets (N-MNIST and DVS-CIFAR10), and comparable accuracy as existing ANNs and pre-trained SNNs on non-spiking datasets (CIFAR10). {To our best knowledge, this is the first work that demonstrates direct training of deep SNNs with high performance on CIFAR10, and the efficient implementation provides a new way to explore the potential of SNNs. | Direct Training for Spiking Neural Networks: Faster, Larger, Better | 6,804 |
Many Pareto-based multi-objective evolutionary algorithms require to rank the solutions of the population in each iteration according to the dominance principle, what can become a costly operation particularly in the case of dealing with many-objective optimization problems. In this paper, we present a new efficient algorithm for computing the non-dominated sorting procedure, called Merge Non-Dominated Sorting (MNDS), which has a best computational complexity of $\Theta(NlogN)$ and a worst computational complexity of $\Theta(MN^2)$. Our approach is based on the computation of the dominance set of each solution by taking advantage of the characteristics of the merge sort algorithm. We compare the MNDS against four well-known techniques that can be considered as the state-of-the-art. The results indicate that the MNDS algorithm outperforms the other techniques in terms of number of comparisons as well as the total running time. | Merge Non-Dominated Sorting Algorithm for Many-Objective Optimization | 6,805 |
Genetic Programming (GP) is a computationally intensive technique which is naturally parallel in nature. Consequently, many attempts have been made to improve its run-time from exploiting highly parallel hardware such as GPUs. However, a second methodology of improving the speed of GP is through efficiency techniques such as subtree caching. However achieving parallel performance and efficiency is a difficult task. This paper will demonstrate an efficiency saving for GP compatible with the harnessing of parallel CPU hardware by exploiting tournament selection. Significant efficiency savings are demonstrated whilst retaining the capability of a high performance parallel implementation of GP. Indeed, a 74% improvement in the speed of GP is achieved with a peak rate of 96 billion GPop/s for classification type problems. | Exploiting Tournament Selection for Efficient Parallel Genetic
Programming | 6,806 |
Even though dense networks have lost importance today, they are still used as final logic elements. It could be shown that these dense networks can be simplified by the sparse graph interpretation. This in turn shows that the information flow between input and output is not optimal with an initialization common today. The lightning initialization sets the weights so that complete information paths exist between input and output from the start. It turned out that pure dense networks and also more complex networks with additional layers benefit from this initialization. The networks accuracy increases faster. The lightning initialization has two parameters which behaved robustly in the tests carried out. However, especially with more complex networks, an improvement effect only occurs at lower learning rates, which shows that the initialization retains its positive effect over the epochs with learning rate reduction. | Dense neural networks as sparse graphs and the lightning initialization | 6,807 |
Designing search algorithms for finding global optima is one of the most active research fields, recently. These algorithms consist of two main categories, i.e., classic mathematical and metaheuristic algorithms. This article proposes a meta-algorithm, Tree-Based Optimization (TBO), which uses other heuristic optimizers as its sub-algorithms in order to improve the performance of search. The proposed algorithm is based on mathematical tree subject and improves performance and speed of search by iteratively removing parts of the search space having low fitness, in order to minimize and purify the search space. The experimental results on several well-known benchmarks show the outperforming performance of TBO algorithm in finding the global solution. Experiments on high dimensional search spaces show significantly better performance when using the TBO algorithm. The proposed algorithm improves the search algorithms in both accuracy and speed aspects, especially for high dimensional searching such as in VLSI CAD tools for Integrated Circuit (IC) design. | Tree-Based Optimization: A Meta-Algorithm for Metaheuristic Optimization | 6,808 |
Successive linear transforms followed by nonlinear "activation" functions can approximate nonlinear functions to arbitrary precision given sufficient layers. The number of necessary layers is dependent on, in part, by the nature of the activation function. The hyperbolic tangent (tanh) has been a favorable choice as an activation until the networks grew deeper and the vanishing gradients posed a hindrance during training. For this reason the Rectified Linear Unit (ReLU) defined by max(0, x) has become the prevailing activation function in deep neural networks. Unlike the tanh function which is smooth, the ReLU yields networks that are piecewise linear functions with a limited number of facets. This paper presents a new activation function, the Piecewise Linear Unit (PLU) that is a hybrid of tanh and ReLU and shown to outperform the ReLU on a variety of tasks while avoiding the vanishing gradients issue of the tanh. | PLU: The Piecewise Linear Unit Activation Function | 6,809 |
Creating catchy slogans is a demanding and clearly creative job for ad agencies. The process of slogan creation by humans involves finding key concepts of the company and its products, and developing a memorable short phrase to describe the key concept. We attempt to follow the same sequence, but with an evolutionary algorithm. A user inputs a paragraph describing describing the company or product to be promoted. The system randomly samples initial slogans from a corpus of existing slogans. The initial slogans are then iteratively mutated and improved using an evolutionary algorithm. Mutation randomly replaces words in an individual with words from the input paragraphs. Internal evaluation measures a combination of grammatical correctness, and semantic similarity to the input paragraphs. Subjective analysis of output slogans leads to the conclusion that the algorithm certainly outputs valuable slogans. External evaluation found that the slogans were somewhat successful in conveying a message, because humans were generally able to select the correct promoted item given a slogan. | Slogatron: Advanced Wealthiness Generator | 6,810 |
This paper develops Penguin search Optimisation Algorithm (PeSOA), a new metaheuristic algorithm which is inspired by the foraging behaviours of penguins. A population of penguins located in the solution space of the given search and optimisation problem is divided into groups and tasked with finding optimal solutions. The penguins of a group perform simultaneous dives and work as a team to collaboratively feed on fish the energy content of which corresponds to the fitness of candidate solutions. Fish stocks have higher fitness and concentration near areas of solution optima and thus drive the search. Penguins can migrate to other places if their original habitat lacks food. We identify two forms of penguin communication both intra-group and inter-group which are useful in designing intensification and diversification strategies. An efficient intensification strategy allows fast convergence to a local optimum, whereas an effective diversification strategy avoids cyclic behaviour around local optima and explores more effectively the space of potential solutions. The proposed PeSOA algorithm has been validated on a well-known set of benchmark functions. Comparative performances with six other nature-inspired metaheuristics show that the PeSOA performs favourably in these tests. A run-time analysis shows that the performance obtained by the PeSOA is very stable at any time of the evolution horizon, making the PeSOA a viable approach for real world applications. | PeSOA: Penguins Search Optimisation Algorithm for Global Optimisation
Problems | 6,811 |
The Artificial Neural Networks (ANNs) have been originally designed to function like a biological neural network, but does an ANN really work in the same way as a biological neural network? As we know, the human brain holds information in its memory cells, so if the ANNs use the same model as our brains, they should store datasets in a similar manner. The most popular type of ANN architecture is based on a layered structure of neurons, whereas a human brain has trillions of complex interconnections of neurons continuously establishing new connections, updating existing ones, and removing the irrelevant connections across different parts of the brain. In this paper, we propose a novel approach to building ANNs which are truly inspired by the biological network containing a mesh of subnets controlled by a central mechanism. A subnet is a network of neurons that hold the dataset values. We attempt to address the following fundamental questions: (1) What is the architecture of the ANN model? Whether the layered architecture is the most appropriate choice? (2) Whether a neuron is a process or a memory cell? (3) What is the best way of interconnecting neurons and what weight-assignment mechanism should be used? (4) How to incorporate prior knowledge, bias, and generalizations for features extraction and prediction? Our proposed ANN architecture leverages the accuracy on textual data and our experimental findings confirm the effectiveness of our model. We also collaborate with the construction of the ANN model for storing and processing the images. | Rethinking the Artificial Neural Networks: A Mesh of Subnets with a
Central Mechanism for Storing and Predicting the Data | 6,812 |
While the history of machine learning so far largely encompasses a series of problems posed by researchers and algorithms that learn their solutions, an important question is whether the problems themselves can be generated by the algorithm at the same time as they are being solved. Such a process would in effect build its own diverse and expanding curricula, and the solutions to problems at various stages would become stepping stones towards solving even more challenging problems later in the process. The Paired Open-Ended Trailblazer (POET) algorithm introduced in this paper does just that: it pairs the generation of environmental challenges and the optimization of agents to solve those challenges. It simultaneously explores many different paths through the space of possible problems and solutions and, critically, allows these stepping-stone solutions to transfer between problems if better, catalyzing innovation. The term open-ended signifies the intriguing potential for algorithms like POET to continue to create novel and increasingly complex capabilities without bound. Our results show that POET produces a diverse range of sophisticated behaviors that solve a wide range of environmental challenges, many of which cannot be solved by direct optimization alone, or even through a direct-path curriculum-building control algorithm introduced to highlight the critical role of open-endedness in solving ambitious challenges. The ability to transfer solutions from one environment to another proves essential to unlocking the full potential of the system as a whole, demonstrating the unpredictable nature of fortuitous stepping stones. We hope that POET will inspire a new push towards open-ended discovery across many domains, where algorithms like POET can blaze a trail through their interesting possible manifestations and solutions. | Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly
Complex and Diverse Learning Environments and Their Solutions | 6,813 |
A key property underlying the success of evolutionary algorithms (EAs) is their global search behavior, which allows the algorithms to `jump' from a current state to other parts of the search space, thereby avoiding to get stuck in local optima. This property is obtained through a random choice of the radius at which offspring are sampled from previously evaluated solutions. It is well known that, thanks to this global search behavior, the probability that an EA using standard bit mutation finds a global optimum of an arbitrary function $f:\{0,1\}^n \to \mathbb{R}$ tends to one as the number of function evaluations grows. This advantage over heuristics using a fixed search radius, however, comes at the cost of using non-optimal step sizes also in those regimes in which the optimal rate is stable for a long time. This downside results in significant performance losses for many standard benchmark problems. We introduce in this work a simple way to interpolate between the random global search of EAs and their deterministic counterparts which sample from a fixed radius only. To this end, we introduce \emph{normalized standard bit mutation}, in which the binomial choice of the search radius is replaced by a normal distribution. Normalized standard bit mutation allows a straightforward way to control its variance, and hence the degree of randomness involved. We experiment with a self-adjusting choice of this variance, and demonstrate its effectiveness for the two classic benchmark problems LeadingOnes and OneMax. Our work thereby also touches a largely ignored question in discrete evolutionary computation: multi-dimensional parameter control. | Interpolating Local and Global Search by Controlling the Variance of
Standard Bit Mutation | 6,814 |
Very expensive problems are very common in practical system that one fitness evaluation costs several hours or even days. Surrogate assisted evolutionary algorithms (SAEAs) have been widely used to solve this crucial problem in the past decades. However, most studied SAEAs focus on solving problems with a budget of at least ten times of the dimension of problems which is unacceptable in many very expensive real-world problems. In this paper, we employ Voronoi diagram to boost the performance of SAEAs and propose a novel framework named Voronoi-based efficient surrogate assisted evolutionary algorithm (VESAEA) for very expensive problems, in which the optimization budget, in terms of fitness evaluations, is only 5 times of the problem's dimension. In the proposed framework, the Voronoi diagram divides the whole search space into several subspace and then the local search is operated in some potentially better subspace. Additionally, in order to trade off the exploration and exploitation, the framework involves a global search stage developed by combining leave-one-out cross-validation and radial basis function surrogate model. A performance selector is designed to switch the search dynamically and automatically between the global and local search stages. The empirical results on a variety of benchmark problems demonstrate that the proposed framework significantly outperforms several state-of-art algorithms with extremely limited fitness evaluations. Besides, the efficacy of Voronoi-diagram is furtherly analyzed, and the results show its potential to optimize very expensive problems. | Voronoi-based Efficient Surrogate-assisted Evolutionary Algorithm for
Very Expensive Problems | 6,815 |
This paper presents an evolutionary metaheuristic called Multiple Search Neuroevolution (MSN) to optimize deep neural networks. The algorithm attempts to search multiple promising regions in the search space simultaneously, maintaining sufficient distance between them. It is tested by training neural networks for two tasks, and compared with other optimization algorithms. The first task is to solve Global Optimization functions with challenging topographies. We found to MSN to outperform classic optimization algorithms such as Evolution Strategies, reducing the number of optimization steps performed by at least 2X. The second task is to train a convolutional neural network (CNN) on the popular MNIST dataset. Using 3.33% of the training set, MSN reaches a validation accuracy of 90%. Stochastic Gradient Descent (SGD) was able to match the same accuracy figure, while taking 7X less optimization steps. Despite lagging, the fact that the MSN metaheurisitc trains a 4.7M-parameter CNN suggests promise for future development. This is by far the largest network ever evolved using a pool of only 50 samples. | Optimizing Deep Neural Networks with Multiple Search Neuroevolution | 6,816 |
This paper thoroughly investigates a range of popular DE configurations to identify components responsible for the emergence of structural bias - recently identified tendency of the algorithm to prefer some regions of the search space for reasons directly unrelated to the objective function values. Such tendency was already studied in GA and PSO where a connection was established between the strength of structural bias and population sizes and potential weaknesses of these algorithms was highlighted. For DE, this study goes further and extends the range of aspects that can contribute to presence of structural bias by including algorithmic component which is usually overlooked - constraint handling technique. A wide range of DE configurations were subjected to the protocol for testing for bias. Results suggest that triggering mechanism for the bias in DE differs to the one previously found for GA and PSO - no clear dependency on population size exists. Setting of DE parameters is based on a separate study which on its own leads to interesting directions of new research. Overall, DE turned out to be robust against structural bias - only DE/current-to-best/1/bin is clearly biased but this effect is mitigated by the use of penalty constraint handling technique. | Infeasibility and structural bias in Differential Evolution | 6,817 |
Liquid State Machine (LSM) is a brain-inspired architecture used for solving problems like speech recognition and time series prediction. LSM comprises of a randomly connected recurrent network of spiking neurons. This network propagates the non-linear neuronal and synaptic dynamics. Maass et al. have argued that the non-linear dynamics of LSMs is essential for its performance as a universal computer. Lyapunov exponent (mu), used to characterize the "non-linearity" of the network, correlates well with LSM performance. We propose a complementary approach of approximating the LSM dynamics with a linear state space representation. The spike rates from this model are well correlated to the spike rates from LSM. Such equivalence allows the extraction of a "memory" metric (tau_M) from the state transition matrix. tau_M displays high correlation with performance. Further, high tau_M system require lesser epochs to achieve a given accuracy. Being computationally cheap (1800x time efficient compared to LSM), the tau_M metric enables exploration of the vast parameter design space. We observe that the performance correlation of the tau_M surpasses the Lyapunov exponent (mu), (2-4x improvement) in the high-performance regime over multiple datasets. In fact, while mu increases monotonically with network activity, the performance reaches a maxima at a specific activity described in literature as the "edge of chaos". On the other hand, tau_M remains correlated with LSM performance even as mu increases monotonically. Hence, tau_M captures the useful memory of network activity that enables LSM performance. It also enables rapid design space exploration and fine-tuning of LSM parameters for high performance. | Predicting Performance using Approximate State Space Model for Liquid
State Machines | 6,818 |
Information coding by precise timing of spikes can be faster and more energy-efficient than traditional rate coding. However, spike-timing codes are often brittle, which has limited their use in theoretical neuroscience and computing applications. Here, we propose a novel type of attractor neural network in complex state space, and show how it can be leveraged to construct spiking neural networks with robust computational properties through a phase-to-timing mapping. Building on Hebbian neural associative memories, like Hopfield networks, we first propose threshold phasor associative memory (TPAM) networks. Complex phasor patterns whose components can assume continuous-valued phase angles and binary magnitudes can be stored and retrieved as stable fixed points in the network dynamics. TPAM achieves high memory capacity when storing sparse phasor patterns, and we derive the energy function that governs its fixed point attractor dynamics. Second, through simulation experiments we show how the complex algebraic computations in TPAM can be approximated by a biologically plausible network of integrate-and-fire neurons with synaptic delays and recurrently connected inhibitory interneurons. The fixed points of TPAM in the complex domain are commensurate with stable periodic states of precisely timed spiking activity that are robust to perturbation. The link established between rhythmic firing patterns and complex attractor dynamics has implications for the interpretation of spike patterns seen in neuroscience, and can serve as a framework for computation in emerging neuromorphic devices. | Robust computation with rhythmic spike patterns | 6,819 |
Theoretical analyses of evolution strategies are indispensable for gaining a deep understanding of their inner workings. For constrained problems, rather simple problems are of interest in the current research. This work presents a theoretical analysis of a multi-recombinative evolution strategy with cumulative step size adaptation applied to a conically constrained linear optimization problem. The state of the strategy is modeled by random variables and a stochastic iterative mapping is introduced. For the analytical treatment, fluctuations are neglected and the mean value iterative system is considered. Non-linear difference equations are derived based on one-generation progress rates. Based on that, expressions for the steady state of the mean value iterative system are derived. By comparison with real algorithm runs, it is shown that for the considered assumptions, the theoretical derivations are able to predict the dynamics and the steady state values of the real runs. | Analysis of the $(μ/μ_I,λ)$-CSA-ES with Repair by Projection
Applied to a Conically Constrained Problem | 6,820 |
Pre-training of models in pruning algorithms plays an important role in pruning decision-making. We find that excessive pre-training is not necessary for pruning algorithms. According to this idea, we propose a pruning algorithm---Incremental pruning based on less training (IPLT). Compared with the traditional pruning algorithm based on a large number of pre-training, IPLT has competitive compression effect than the traditional pruning algorithm under the same simple pruning strategy. On the premise of ensuring accuracy, IPLT can achieve 8x-9x compression for VGG-19 on CIFAR-10 and only needs to pre-train few epochs. For VGG-19 on CIFAR-10, we can not only achieve 10 times test acceleration, but also about 10 times training acceleration. At present, the research mainly focuses on the compression and acceleration in the application stage of the model, while the compression and acceleration in the training stage are few. We newly proposed a pruning algorithm that can compress and accelerate in the training stage. It is novel to consider the amount of pre-training required by pruning algorithm. Our results have implications: Too much pre-training may be not necessary for pruning algorithms. | Really should we pruning after model be totally trained? Pruning based
on a small amount of training | 6,821 |
Evolutionary computation (EC) algorithms, such as discrete and multi-objective versions of particle swarm optimization (PSO), have been applied to solve the Feature selection (FS) problem, tackling the combinatorial explosion of search spaces that are peppered with local minima. Furthermore, high-dimensional FS problems such as finding a small set of biomarkers to make a diagnostic call add an additional challenge as such methods ability to pick out the most important features must remain unchanged in decision spaces of increasing dimensions and presence of irrelevant features. We developed a combinatorial PSO algorithm, called COMB-PSO, that scales up to high-dimensional gene expression data while still selecting the smallest subsets of genes that allow reliable classification of samples. In particular, COMB-PSO enhances the encoding, speed of convergence, control of divergence and diversity of the conventional PSO algorithm, balancing exploration and exploitation of the search space. Applying our approach on real gene expression data of different cancers, COMB-PSO finds gene sets of smallest size that allow a reliable classification of the underlying disease classes. | A Stable Combinatorial Particle Swarm Optimization for Scalable Feature
Selection in Gene Expression Data | 6,822 |
The way how recurrently connected networks of spiking neurons in the brain acquire powerful information processing capabilities through learning has remained a mystery. This lack of understanding is linked to a lack of learning algorithms for recurrent networks of spiking neurons (RSNNs) that are both functionally powerful and can be implemented by known biological mechanisms. Since RSNNs are simultaneously a primary target for implementations of brain-inspired circuits in neuromorphic hardware, this lack of algorithmic insight also hinders technological progress in that area. The gold standard for learning in recurrent neural networks in machine learning is back-propagation through time (BPTT), which implements stochastic gradient descent with regard to a given loss function. But BPTT is unrealistic from a biological perspective, since it requires a transmission of error signals backwards in time and in space, i.e., from post- to presynaptic neurons. We show that an online merging of locally available information during a computation with suitable top-down learning signals in real-time provides highly capable approximations to BPTT. For tasks where information on errors arises only late during a network computation, we enrich locally available information through feedforward eligibility traces of synapses that can easily be computed in an online manner. The resulting new generation of learning algorithms for recurrent neural networks provides a new understanding of network learning in the brain that can be tested experimentally. In addition, these algorithms provide efficient methods for on-chip training of RSNNs in neuromorphic hardware. | Biologically inspired alternatives to backpropagation through time for
learning in recurrent neural nets | 6,823 |
Long Short-Term Memory (LSTM) has achieved state-of-the-art performances on a wide range of tasks. Its outstanding performance is guaranteed by the long-term memory ability which matches the sequential data perfectly and the gating structure controlling the information flow. However, LSTMs are prone to be memory-bandwidth limited in realistic applications and need an unbearable period of training and inference time as the model size is ever-increasing. To tackle this problem, various efficient model compression methods have been proposed. Most of them need a big and expensive pre-trained model which is a nightmare for resource-limited devices where the memory budget is strictly limited. To remedy this situation, in this paper, we incorporate the Sparse Evolutionary Training (SET) procedure into LSTM, proposing a novel model dubbed SET-LSTM. Rather than starting with a fully-connected architecture, SET-LSTM has a sparse topology and dramatically fewer parameters in both phases, training and inference. Considering the specific architecture of LSTMs, we replace the LSTM cells and embedding layers with sparse structures and further on, use an evolutionary strategy to adapt the sparse connectivity to the data. Additionally, we find that SET-LSTM can provide many different good combinations of sparse connectivity to substitute the overparameterized optimization problem of dense neural networks. Evaluated on four sentiment analysis classification datasets, the results demonstrate that our proposed model is able to achieve usually better performance than its fully connected counterpart while having less than 4\% of its parameters. | Intrinsically Sparse Long Short-Term Memory Networks | 6,824 |
This paper investigates a new hybridization of multi-objective particle swarm optimization (MOPSO) and cooperative agents (MOPSO-CA) to handle the problem of stagnation encounters in MOPSO, which leads solutions to trap in local optima. The proposed approach involves a new distribution strategy based on the idea of having a set of a sub-population, each of which is processed by one agent. The number of the sub-population and agents are adjusted dynamically through the Pareto ranking. This method allocates a dynamic number of sub-population as required to improve diversity in the search space. Additionally, agents are used for better management for the exploitation within a sub-population, and for exploration among sub-populations. Furthermore, we investigate the automated negotiation within agents in order to share the best knowledge. To validate our approach, several benchmarks are performed. The results show that the introduced variant ensures the trade-off between the exploitation and exploration with respect to the comparative algorithms | Multi Objective Particle Swarm Optimization based Cooperative Agents
with Automated Negotiation | 6,825 |
P300-based spellers are one of the main methods for EEG-based brain-computer interface, and the detection of the P300 target event with high accuracy is an important prerequisite. The rapid serial visual presentation (RSVP) protocol is of high interest because it can be used by patients who have lost control over their eyes. In this study we wish to explore the suitability of recurrent neural networks (RNNs) as a machine learning method for identifying the P300 signal in RSVP data. We systematically compare RNN with alternative methods such as linear discriminant analysis (LDA) and convolutional neural network (CNN). Our results indicate that LDA performs as well as the neural network models or better on single subject data, but a network combining CNN and RNN has advantages when transferring learning among subejcts, and is significantly more resilient to temporal noise than other methods. | Recurrent Neural Networks for P300-based BCI | 6,826 |
Many long short-term memory (LSTM) applications need fast yet compact models. Neural network compression approaches, such as the grow-and-prune paradigm, have proved to be promising for cutting down network complexity by skipping insignificant weights. However, current compression strategies are mostly hardware-agnostic and network complexity reduction does not always translate into execution efficiency. In this work, we propose a hardware-guided symbiotic training methodology for compact, accurate, yet execution-efficient inference models. It is based on our observation that hardware may introduce substantial non-monotonic behavior, which we call the latency hysteresis effect, when evaluating network size vs. inference latency. This observation raises question about the mainstream smaller-dimension-is-better compression strategy, which often leads to a sub-optimal model architecture. By leveraging the hardware-impacted hysteresis effect and sparsity, we are able to achieve the symbiosis of model compactness and accuracy with execution efficiency, thus reducing LSTM latency while increasing its accuracy. We have evaluated our algorithms on language modeling and speech recognition applications. Relative to the traditional stacked LSTM architecture obtained for the Penn Treebank dataset, we reduce the number of parameters by 18.0x (30.5x) and measured run-time latency by up to 2.4x (5.2x) on Nvidia GPUs (Intel Xeon CPUs) without any accuracy degradation. For the DeepSpeech2 architecture obtained for the AN4 dataset, we reduce the number of parameters by 7.0x (19.4x), word error rate from 12.9% to 9.9% (10.4%), and measured run-time latency by up to 1.7x (2.4x) on Nvidia GPUs (Intel Xeon CPUs). Thus, our method yields compact, accurate, yet execution-efficient inference models. | Hardware-Guided Symbiotic Training for Compact, Accurate, yet
Execution-Efficient LSTM | 6,827 |
We introduce the perceptron Turing machine and show how it can be used to create a system of neuroevolution. Advantages of this approach include automatic scaling of solutions to larger problem sizes, the ability to experiment with hand-coded solutions, and an enhanced potential for understanding evolved solutions. Hand-coded solutions may be implemented in the low-level language of Turing machines, which is the genotype used in neuroevolution, but a high-level language called Lopro is introduced to make the job easier. | Neuroevolution with Perceptron Turing Machines | 6,828 |
This paper computationally obtains optimal bounded-weight, binary, error-correcting codes for a variety of distance bounds and dimensions. We compare the sizes of our codes to the sizes of optimal constant-weight, binary, error-correcting codes, and evaluate the differences. | On Bounded-Weight Error-Correcting Codes | 6,829 |
Let A(q,n,d) denote the maximum size of a q-ary code of length n and distance d. We study the minimum asymptotic redundancy \rho(q,n,d)=n-log_q A(q,n,d) as n grows while q and d are fixed. For any d and q<=d-1, long algebraic codes are designed that improve on the BCH codes and have the lowest asymptotic redundancy \rho(q,n,d) <= ((d-3)+1/(d-2)) log_q n known to date. Prior to this work, codes of fixed distance that asymptotically surpass BCH codes and the Gilbert-Varshamov bound were designed only for distances 4,5 and 6. | Long Nonbinary Codes Exceeding the Gilbert - Varshamov Bound for any
Fixed Distance | 6,830 |
We give a new lower bound on the expansion coefficient of an edge-vertex graph of a $d$-regular graph. As a consequence, we obtain an improvement on the lower bound on relative minimum distance of the expander codes constructed by Sipser and Spielman. We also derive some improved results on the vertex expansion of graphs that help us in improving the parameters of the expander codes of Alon, Bruck, Naor, Naor, and Roth. | On Expanders Graphs: Parameters and Applications | 6,831 |
We address the problem of bounding below the probability of error under maximum likelihood decoding of a binary code with a known distance distribution used on a binary symmetric channel. An improved upper bound is given for the maximum attainable exponent of this probability (the reliability function of the channel). In particular, we prove that the ``random coding exponent'' is the true value of the channel reliability for code rate $R$ in some interval immediately below the critical rate of the channel. An analogous result is obtained for the Gaussian channel. | Distance distribution of binary codes and the error probability of
decoding | 6,832 |
We study codes on graphs combined with an iterative message passing algorithm for quantization. Specifically, we consider the binary erasure quantization (BEQ) problem which is the dual of the binary erasure channel (BEC) coding problem. We show that duals of capacity achieving codes for the BEC yield codes which approach the minimum possible rate for the BEQ. In contrast, low density parity check codes cannot achieve the minimum rate unless their density grows at least logarithmically with block length. Furthermore, we show that duals of efficient iterative decoding algorithms for the BEC yield efficient encoding algorithms for the BEQ. Hence our results suggest that graphical models may yield near optimal codes in source coding as well as in channel coding and that duality plays a key role in such constructions. | Iterative Quantization Using Codes On Graphs | 6,833 |
A variable-length code is a fix-free code if no codeword is a prefix or a suffix of any other codeword. In a fix-free code any finite sequence of codewords can be decoded in both directions, which can improve the robustness to channel noise and speed up the decoding process. In this paper we prove a new sufficient condition of the existence of fix-free codes and improve the upper bound on the redundancy of optimal fix-free codes. | Improved Upper Bound for the Redundancy of Fix-Free Codes | 6,834 |
Fundamental results concerning the dynamics of abelian group codes (behaviors) and their duals are developed. Duals of sequence spaces over locally compact abelian groups may be defined via Pontryagin duality; dual group codes are orthogonal subgroups of dual sequence spaces. The dual of a complete code or system is finite, and the dual of a Laurent code or system is (anti-)Laurent. If C and C^\perp are dual codes, then the state spaces of C act as the character groups of the state spaces of C^\perp. The controllability properties of C are the observability properties of C^\perp. In particular, C is (strongly) controllable if and only if C^\perp is (strongly) observable, and the controller memory of C is the observer memory of C^\perp. The controller granules of C act as the character groups of the observer granules of C^\perp. Examples of minimal observer-form encoder and syndrome-former constructions are given. Finally, every observer granule of C is an "end-around" controller granule of C. | The Dynamics of Group Codes: Dual Abelian Group Codes and Systems | 6,835 |
We consider lossy source coding when side information affecting the distortion measure may be available at the encoder, decoder, both, or neither. For example, such distortion side information can model reliabilities for noisy measurements, sensor calibration information, or perceptual effects like masking and sensitivity to context. When the distortion side information is statistically independent of the source, we show that in many cases (e.g, for additive or multiplicative distortion side information) there is no penalty for knowing the side information only at the encoder, and there is no advantage to knowing it at the decoder. Furthermore, for quadratic distortion measures scaled by the distortion side information, we evaluate the penalty for lack of encoder knowledge and show that it can be arbitrarily large. In this scenario, we also sketch transform based quantizers constructions which efficiently exploit encoder side information in the high-resolution limit. | Source Coding With Distortion Side Information At The Encoder | 6,836 |
We present two sequences of ensembles of non-systematic irregular repeat-accumulate codes which asymptotically (as their block length tends to infinity) achieve capacity on the binary erasure channel (BEC) with bounded complexity per information bit. This is in contrast to all previous constructions of capacity-achieving sequences of ensembles whose complexity grows at least like the log of the inverse of the gap (in rate) to capacity. The new bounded complexity result is achieved by puncturing bits, and allowing in this way a sufficient number of state nodes in the Tanner graph representing the codes. We also derive an information-theoretic lower bound on the decoding complexity of randomly punctured codes on graphs. The bound holds for every memoryless binary-input output-symmetric channel and is refined for the BEC. | Capacity-achieving ensembles for the binary erasure channel with bounded
complexity | 6,837 |
We present two sequences of ensembles of non-systematic irregular repeat-accumulate codes which asymptotically (as their block length tends to infinity) achieve capacity on the binary erasure channel (BEC) with bounded complexity per information bit. This is in contrast to all previous constructions of capacity-achieving sequences of ensembles whose complexity grows at least like the log of the inverse of the gap (in rate) to capacity. The new bounded complexity result is achieved by puncturing bits, and allowing in this way a sufficient number of state nodes in the Tanner graph representing the codes. We also derive an information-theoretic lower bound on the decoding complexity of randomly punctured codes on graphs. The bound holds for every memoryless binary-input output-symmetric channel, and is refined for the BEC. | Bounds on the decoding complexity of punctured codes on graphs | 6,838 |
We compare the elementary theories of Shannon information and Kolmogorov complexity, the extent to which they have a common purpose, and where they are fundamentally different. We discuss and relate the basic notions of both theories: Shannon entropy versus Kolmogorov complexity, the relation of both to universal coding, Shannon mutual information versus Kolmogorov (`algorithmic') mutual information, probabilistic sufficient statistic versus algorithmic sufficient statistic (related to lossy compression in the Shannon theory versus meaningful information in the Kolmogorov theory), and rate distortion theory versus Kolmogorov's structure function. Part of the material has appeared in print before, scattered through various publications, but this is the first comprehensive systematic comparison. The last mentioned relations are new. | Shannon Information and Kolmogorov Complexity | 6,839 |
Capacity formulas and random-coding exponents are derived for a generalized family of Gel'fand-Pinsker coding problems. These exponents yield asymptotic upper bounds on the achievable log probability of error. In our model, information is to be reliably transmitted through a noisy channel with finite input and output alphabets and random state sequence, and the channel is selected by a hypothetical adversary. Partial information about the state sequence is available to the encoder, adversary, and decoder. The design of the transmitter is subject to a cost constraint. Two families of channels are considered: 1) compound discrete memoryless channels (CDMC), and 2) channels with arbitrary memory, subject to an additive cost constraint, or more generally to a hard constraint on the conditional type of the channel output given the input. Both problems are closely connected. The random-coding exponent is achieved using a stacked binning scheme and a maximum penalized mutual information decoder, which may be thought of as an empirical generalized Maximum a Posteriori decoder. For channels with arbitrary memory, the random-coding exponents are larger than their CDMC counterparts. Applications of this study include watermarking, data hiding, communication in presence of partially known interferers, and problems such as broadcast channels, all of which involve the fundamental idea of binning. | Capacity and Random-Coding Exponents for Channel Coding with Side
Information | 6,840 |
We consider source coding with fixed lag side information at the decoder. We focus on the special case of perfect side information with unit lag corresponding to source coding with feedforward (the dual of channel coding with feedback) introduced by Pradhan. We use this duality to develop a linear complexity algorithm which achieves the rate-distortion bound for any memoryless finite alphabet source and distortion measure. | Source Coding with Fixed Lag Side Information | 6,841 |
We propose use of QR factorization with sort and Dijkstra's algorithm for decreasing the computational complexity of the sphere decoder that is used for ML detection of signals on the multi-antenna fading channel. QR factorization with sort decreases the complexity of searching part of the decoder with small increase in the complexity required for preprocessing part of the decoder. Dijkstra's algorithm decreases the complexity of searching part of the decoder with increase in the storage complexity. The computer simulation demonstrates that the complexity of the decoder is reduced by the proposed methods significantly. | Two Methods for Decreasing the Computational Complexity of the MIMO ML
Decoder | 6,842 |
In this paper, we analyze the performance of space-time block codes which enable symbolwise maximum likelihood decoding. We derive an upper bound of maximum mutual information (MMI) on space-time block codes that enable symbolwise maximum likelihood decoding for a frequency non-selective quasi-static fading channel. MMI is an upper bound on how much one can send information with vanishing error probability by using the target code. | Maximum Mutual Information of Space-Time Block Codes with Symbolwise
Decodability | 6,843 |
In this paper, we present two low complexity algorithms that achieve capacity for the noiseless (d,k) constrained channel when k=2d+1, or when k-d+1 is not prime. The first algorithm, called symbol sliding, is a generalized version of the bit flipping algorithm introduced by Aviran et al. [1]. In addition to achieving capacity for (d,2d+1) constraints, it comes close to capacity in other cases. The second algorithm is based on interleaving, and is a generalized version of the bit stuffing algorithm introduced by Bender and Wolf [2]. This method uses fewer than k-d biased bit streams to achieve capacity for (d,k) constraints with k-d+1 not prime. In particular, the encoder for (d,d+2^m-1) constraints, 1\le m<\infty, requires only m biased bit streams. | Capacity Achieving Code Constructions for Two Classes of (d,k)
Constraints | 6,844 |
Capacity analysis for channels with side information at the receiver has been an active area of interest. This problem is well investigated for the case of finite alphabet channels. However, the results are not easily generalizable to the case of continuous alphabet channels due to analytic difficulties inherent with continuous alphabets. In the first part of this two-part paper, we address an analytical framework for capacity analysis of continuous alphabet channels with side information at the receiver. For this purpose, we establish novel necessary and sufficient conditions for weak* continuity and strict concavity of the mutual information. These conditions are used in investigating the existence and uniqueness of the capacity-achieving measures. Furthermore, we derive necessary and sufficient conditions that characterize the capacity value and the capacity-achieving measure for continuous alphabet channels with side information at the receiver. | Capacity Analysis for Continuous Alphabet Channels with Side
Information, Part I: A General Framework | 6,845 |
In this part, we consider the capacity analysis for wireless mobile systems with multiple antenna architectures. We apply the results of the first part to a commonly known baseband, discrete-time multiple antenna system where both the transmitter and receiver know the channel's statistical law. We analyze the capacity for additive white Gaussian noise (AWGN) channels, fading channels with full channel state information (CSI) at the receiver, fading channels with no CSI, and fading channels with partial CSI at the receiver. For each type of channels, we study the capacity value as well as issues such as the existence, uniqueness, and characterization of the capacity-achieving measures for different types of moment constraints. The results are applicable to both Rayleigh and Rician fading channels in the presence of arbitrary line-of-sight and correlation profiles. | Capacity Analysis for Continuous Alphabet Channels with Side
Information, Part II: MIMO Channels | 6,846 |
We examine the structure of families of distortion balls from the perspective of Kolmogorov complexity. Special attention is paid to the canonical rate-distortion function of a source word which returns the minimal Kolmogorov complexity of all distortion balls containing that word subject to a bound on their cardinality. This canonical rate-distortion function is related to the more standard algorithmic rate-distortion function for the given distortion measure. Examples are given of list distortion, Hamming distortion, and Euclidean distortion. The algorithmic rate-distortion function can behave differently from Shannon's rate-distortion function. To this end, we show that the canonical rate-distortion function can and does assume a wide class of shapes (unlike Shannon's); we relate low algorithmic mutual information to low Kolmogorov complexity (and consequently suggest that certain aspects of the mutual information formulation of Shannon's rate-distortion function behave differently than would an analogous formulation using algorithmic mutual information); we explore the notion that low Kolmogorov complexity distortion balls containing a given word capture the interesting properties of that word (which is hard to formalize in Shannon's theory) and this suggests an approach to denoising; and, finally, we show that the different behavior of the rate-distortion curves of individual source words to some extent disappears after averaging over the source words. | Rate Distortion and Denoising of Individual Data Using Kolmogorov
complexity | 6,847 |
The feedback capacity of the stationary Gaussian additive noise channel has been open, except for the case where the noise is white. Here we find the feedback capacity of the stationary first-order moving average additive Gaussian noise channel in closed form. Specifically, the channel is given by $Y_i = X_i + Z_i,$ $i = 1, 2, ...,$ where the input $\{X_i\}$ satisfies a power constraint and the noise $\{Z_i\}$ is a first-order moving average Gaussian process defined by $Z_i = \alpha U_{i-1} + U_i,$ $|\alpha| \le 1,$ with white Gaussian innovations $U_i,$ $i = 0,1,....$ We show that the feedback capacity of this channel is $-\log x_0,$ where $x_0$ is the unique positive root of the equation $ \rho x^2 = (1-x^2) (1 - |\alpha|x)^2,$ and $\rho$ is the ratio of the average input power per transmission to the variance of the noise innovation $U_i$. The optimal coding scheme parallels the simple linear signalling scheme by Schalkwijk and Kailath for the additive white Gaussian noise channel -- the transmitter sends a real-valued information-bearing signal at the beginning of communication and subsequently refines the receiver's error by processing the feedback noise signal through a linear stationary first-order autoregressive filter. The resulting error probability of the maximum likelihood decoding decays doubly-exponentially in the duration of the communication. This feedback capacity of the first-order moving average Gaussian channel is very similar in form to the best known achievable rate for the first-order \emph{autoregressive} Gaussian noise channel studied by Butman, Wolfowitz, and Tiernan, although the optimality of the latter is yet to be established. | Feedback Capacity of the First-Order Moving Average Gaussian Channel | 6,848 |
Geographic routing with greedy relaying strategies have been widely studied as a routing scheme in sensor networks. These schemes assume that the nodes have perfect information about the location of the destination. When the distance between the source and destination is normalized to unity, the asymptotic routing delays in these schemes are $\Theta(\frac{1}{M(n)}),$ where M(n) is the maximum distance traveled in a single hop (transmission range of a radio). In this paper, we consider routing scenarios where nodes have location errors (imprecise GPS), or where only coarse geographic information about the destination is available, and only a fraction of the nodes have routing information. We show that even with such imprecise or limited destination-location information, the routing delays are $\Theta(\frac{1}{M(n)})$. We also consider the throughput-capacity of networks with progressive routing strategies that take packets closer to the destination in every step, but not necessarily along a straight-line. We show that the throughput-capacity with progressive routing is order-wise the same as the maximum achievable throughput-capacity. | Geographic Routing with Limited Information in Sensor Networks | 6,849 |
We obtain the first term in the high signal-to-noise ratio (SNR) expansion of the capacity of fading networks where the transmitters and receivers--while fully cognizant of the fading \emph{law}--have no access to the fading \emph{realization}. This term is an integer multiple of $\log \log \textnormal{SNR}$ with the coefficient having a simple combinatorial characterization. | On the High-SNR Capacity of Non-Coherent Networks | 6,850 |
The dependence of the Gaussian input information rate on the line-of-sight (LOS) matrix in multiple-input multiple-output coherent Rician fading channels is explored. It is proved that the outage probability and the mutual information induced by a multivariate circularly symmetric Gaussian input with any covariance matrix are monotonic in the LOS matrix D, or more precisely, monotonic in D'D in the sense of the Loewner partial order. Conversely, it is also demonstrated that this ordering on the LOS matrices is a necessary condition for the uniform monotonicity over all input covariance matrices. This result is subsequently applied to prove the monotonicity of the isotropic Gaussian input information rate and channel capacity in the singular values of the LOS matrix. Extensions to multiple-access channels are also discussed. | Monotonicity Results for Coherent MIMO Rician Channels | 6,851 |
Recently, a quasi-orthogonal space-time block code (QSTBC) capable of achieving a significant fraction of the outage mutual information of a multiple-input-multiple output (MIMO) wireless communication system for the case of four transmit and one receive antennas was proposed. We generalize these results to $n_T=2^n$ transmit and an arbitrary number of receive antennas $n_R$. Furthermore, we completely characterize the structure of the equivalent channel for the general case and show that for all $n_T=2^n$ and $n_R$ the eigenvectors of the equivalent channel are fixed and independent from the channel realization. Furthermore, the eigenvalues of the equivalent channel are independent identically distributed random variables each following a noncentral chi-square distribution with $4n_R$ degrees of freedom. Based on these important insights into the structure of the QSTBC, we derive an analytical lower bound for the fraction of outage probability achieved with QSTBC and show that this bound is tight for low signal-to-noise-ratios (SNR) values and also for increasing number of receive antennas. We also present an upper bound, which is tight for high SNR values and derive analytical expressions for the case of four transmit antennas. Finally, by utilizing the special structure of the QSTBC we propose a new transmit strategy, which decouples the signals transmitted from different antennas in order to detect the symbols separately with a linear ML-detector rather than joint detection, an up to now only known advantage of orthogonal space-time block codes (OSTBC). | Complete Characterization of the Equivalent MIMO Channel for
Quasi-Orthogonal Space-Time Codes | 6,852 |
This paper deals with arbitrarily distributed finite-power input signals observed through an additive Gaussian noise channel. It shows a new formula that connects the input-output mutual information and the minimum mean-square error (MMSE) achievable by optimal estimation of the input given the output. That is, the derivative of the mutual information (nats) with respect to the signal-to-noise ratio (SNR) is equal to half the MMSE, regardless of the input statistics. This relationship holds for both scalar and vector signals, as well as for discrete-time and continuous-time noncausal MMSE estimation. This fundamental information-theoretic result has an unexpected consequence in continuous-time nonlinear estimation: For any input signal with finite power, the causal filtering MMSE achieved at SNR is equal to the average value of the noncausal smoothing MMSE achieved with a channel whose signal-to-noise ratio is chosen uniformly distributed between 0 and SNR. | Mutual Information and Minimum Mean-square Error in Gaussian Channels | 6,853 |
A new lower bound on the error probability of maximum likelihood decoding of a binary code on a binary symmetric channel was proved in Barg and McGregor (2004, cs.IT/0407011). It was observed in that paper that this bound leads to a new region of code rates in which the random coding exponent is asymptotically tight, giving a new region in which the reliability of the BSC is known exactly. The present paper explains the relation of these results to the union bound on the error probability. | On the asymptotic accuracy of the union bound | 6,854 |
We introduce the idea of distortion side information, which does not directly depend on the source but instead affects the distortion measure. We show that such distortion side information is not only useful at the encoder, but that under certain conditions, knowing it at only the encoder is as good as knowing it at both encoder and decoder, and knowing it at only the decoder is useless. Thus distortion side information is a natural complement to the signal side information studied by Wyner and Ziv, which depends on the source but does not involve the distortion measure. Furthermore, when both types of side information are present, we characterize the penalty for deviating from the configuration of encoder-only distortion side information and decoder-only signal side information, which in many cases is as good as full side information knowledge. | Source Coding With Encoder Side Information | 6,855 |
We consider transmitting a source across a pair of independent, non-ergodic channels with random states (e.g., slow fading channels) so as to minimize the average distortion. The general problem is unsolved. Hence, we focus on comparing two commonly used source and channel encoding systems which correspond to exploiting diversity either at the physical layer through parallel channel coding or at the application layer through multiple description source coding. For on-off channel models, source coding diversity offers better performance. For channels with a continuous range of reception quality, we show the reverse is true. Specifically, we introduce a new figure of merit called the distortion exponent which measures how fast the average distortion decays with SNR. For continuous-state models such as additive white Gaussian noise channels with multiplicative Rayleigh fading, optimal channel coding diversity at the physical layer is more efficient than source coding diversity at the application layer in that the former achieves a better distortion exponent. Finally, we consider a third decoding architecture: multiple description encoding with a joint source-channel decoding. We show that this architecture achieves the same distortion exponent as systems with optimal channel coding diversity for continuous-state channels, and maintains the the advantages of multiple description systems for on-off channels. Thus, the multiple description system with joint decoding achieves the best performance, from among the three architectures considered, on both continuous-state and on-off channels. | Source-Channel Diversity for Parallel Channels | 6,856 |
In this work we have considered formal power series and partial differential equations, and their relationship with Coding Theory. We have obtained the nature of solutions for the partial differential equations for Cycle Poisson Case. The coefficients for this case have been simulated, and the high tendency of growth is shown. In the light of Complex Analysis, the Hadamard Multiplication's Theorem is presented as a new approach to divide the power sums relating to the error probability, each part of which can be analyzed later. | Application of Generating Functions and Partial Differential Equations
in Coding Theory | 6,857 |
Existing quantum key distribution schemes need the support of classical authentication scheme to ensure security. This is a conceptual drawback of quantum cryptography. It is pointed out that quantum cryptosystem does not need any support of classical cryptosystem to ensure security. No-cloning principal can alone provide security in communication. Even no-cloning principle itself can help to authenticate each bit of information. It implies that quantum password need not to be a secret password. | No-cloning principal can alone provide security | 6,858 |
We study the repeated use of a monotonic recording medium--such as punched tape or photographic plate--where marks can be added at any time but never erased. (For practical purposes, also the electromagnetic "ether" falls into this class.) Our emphasis is on the case where the successive users act independently and selfishly, but not maliciously; typically, the "first user" would be a blind natural process tending to degrade the recording medium, and the "second user" a human trying to make the most of whatever capacity is left. To what extent is a length of used tape "equivalent"--for information transmission purposes--to a shorter length of virgin tape? Can we characterize a piece of used tape by an appropriate "effective length" and forget all other details? We identify two equivalence principles. The weak principle is exact, but only holds for a sequence of infinitesimal usage increments. The strong principle holds for any amount of incremental usage, but is only approximate; nonetheless, it is quite accurate even in the worst case and is virtually exact over most of the range--becoming exact in the limit of heavily used tape. The fact that strong equivalence does not hold exactly, but then it does almost exactly, comes as a bit of a surprise. | Thermodynamics of used punched tape: A weak and a strong equivalence
principle | 6,859 |
For practical wireless DS-CDMA systems, channel estimation is imperfect due to noise and interference. In this paper, the impact of channel estimation errors on multiuser detection (MUD) is analyzed under the framework of the replica method. System performance is obtained in the large system limit for optimal MUD, linear MUD and turbo MUD, and is validated by numerical results for finite systems. | Impact of Channel Estimation Errors on Multiuser Detection via the
Replica Method | 6,860 |
Communications in dispersive direct-sequence code-division multiple-access (DS-CDMA) channels suffer from intersymbol and multiple-access interference, which can significantly impair performance. Joint maximum \textit{a posteriori} probability (MAP) equalization and multiuser detection with error control decoding can be used to mitigate this interference and to achieve the optimal bit error rate. Unfortunately, such optimal detection typically requires prohibitive computational complexity. This problem is addressed in this paper through the development of a reduced state trellis search detection algorithm, based on decision feedback from channel decoders. The performance of this algorithm is analyzed in the large-system limit. This analysis and simulations show that this low-complexity algorithm can obtain near-optimal performance under moderate signal-to-noise ratio and attains larger system load capacity than parallel interference cancellation. | Low Complexity Joint Iterative Equalization and Multiuser Detection in
Dispersive DS-CDMA Channels | 6,861 |
In this paper, the performance of a binary phase shift keyed random time-hopping impulse radio system with pulse-based polarity randomization is analyzed. Transmission over frequency-selective channels is considered and the effects of inter-frame interference and multiple access interference on the performance of a generic Rake receiver are investigated for both synchronous and asynchronous systems. Closed form (approximate) expressions for the probability of error that are valid for various Rake combining schemes are derived. The asynchronous system is modelled as a chip-synchronous system with uniformly distributed timing jitter for the transmitted pulses of interfering users. This model allows the analytical technique developed for the synchronous case to be extended to the asynchronous case. An approximate closed-form expression for the probability of bit error, expressed in terms of the autocorrelation function of the transmitted pulse, is derived for the asynchronous case. Then, transmission over an additive white Gaussian noise channel is studied as a special case, and the effects of multiple-access interference is investigated for both synchronous and asynchronous systems. The analysis shows that the chip-synchronous assumption can result in over-estimating the error probability, and the degree of over-estimation mainly depends on the autocorrelation function of the ultra-wideband pulse and the signal-to-interference-plus-noise-ratio of the system. Simulations studies support the approximate analysis. | Performance Evaluation of Impulse Radio UWB Systems with Pulse-Based
Polarity Randomization | 6,862 |
Sensor networks in which energy is a limited resource so that energy consumption must be minimized for the intended application are considered. In this context, an energy-efficient method for the joint estimation of an unknown analog source under a given distortion constraint is proposed. The approach is purely analog, in which each sensor simply amplifies and forwards the noise-corrupted analog bservation to the fusion center for joint estimation. The total transmission power across all the sensor nodes is minimized while satisfying a distortion requirement on the joint estimate. The energy efficiency of this analog approach is compared with previously proposed digital approaches with and without coding. It is shown in our simulation that the analog approach is more energy-efficient than the digital system without coding, and in some cases outperforms the digital system with optimal coding. | Energy-Efficient Joint Estimation in Sensor Networks: Analog vs. Digital | 6,863 |
The effect of Rician-ness on the capacity of multiple antenna systems is investigated under the assumption that channel state information (CSI) is available only at the receiver. The average-power-constrained capacity of such systems is considered under two different assumptions on the knowledge about the fading available at the transmitter: the case in which the transmitter has no knowledge of fading at all, and the case in which the transmitter has knowledge of the distribution of the fading process but not the instantaneous CSI. The exact capacity is given for the former case while capacity bounds are derived for the latter case. A new signalling scheme is also proposed for the latter case and it is shown that by exploiting the knowledge of Rician-ness at the transmitter via this signalling scheme, significant capacity gain can be achieved. The derived capacity bounds are evaluated explicitly to provide numerical results in some representative situations. | On the Capacity of Multiple Antenna Systems in Rician Fading | 6,864 |
The problem of scheduling sensor transmissions for the detection of correlated random fields using spatially deployed sensors is considered. Using the large deviations principle, a closed-form expression for the error exponent of the miss probability is given as a function of the sensor spacing and signal-to-noise ratio (SNR). It is shown that the error exponent has a distinct characteristic: at high SNR, the error exponent is monotonically increasing with respect to sensor spacing, while at low SNR there is an optimal spacing for scheduled sensors. | A Large Deviations Approach to Sensor Scheduling for Detection of
Correlated Random Fields | 6,865 |
Estimating the number of sources impinging on an array of sensors is a well known and well investigated problem. A common approach for solving this problem is to use an information theoretic criterion, such as Minimum Description Length (MDL) or the Akaike Information Criterion (AIC). The MDL estimator is known to be a consistent estimator, robust against deviations from the Gaussian assumption, and non-robust against deviations from the point source and/or temporally or spatially white additive noise assumptions. Over the years several alternative estimation algorithms have been proposed and tested. Usually, these algorithms are shown, using computer simulations, to have improved performance over the MDL estimator, and to be robust against deviations from the assumed spatial model. Nevertheless, these robust algorithms have high computational complexity, requiring several multi-dimensional searches. In this paper, motivated by real life problems, a systematic approach toward the problem of robust estimation of the number of sources using information theoretic criteria is taken. An MDL type estimator that is robust against deviation from assumption of equal noise level across the array is studied. The consistency of this estimator, even when deviations from the equal noise level assumption occur, is proven. A novel low-complexity implementation method avoiding the need for multi-dimensional searches is presented as well, making this estimator a favorable choice for practical applications. | Estimation of the Number of Sources in Unbalanced Arrays via Information
Theoretic Criteria | 6,866 |
Convex relaxations of the optimal finger selection algorithm are proposed for a minimum mean square error (MMSE) Rake receiver in an impulse radio ultra-wideband system. First, the optimal finger selection problem is formulated as an integer programming problem with a non-convex objective function. Then, the objective function is approximated by a convex function and the integer programming problem is solved by means of constraint relaxation techniques. The proposed algorithms are suboptimal due to the approximate objective function and the constraint relaxation steps. However, they can be used in conjunction with the conventional finger selection algorithm, which is suboptimal on its own since it ignores the correlation between multipath components, to obtain performances reasonably close to that of the optimal scheme that cannot be implemented in practice due to its complexity. The proposed algorithms leverage convexity of the optimization problem formulations, which is the watershed between `easy' and `difficult' optimization problems. | Optimal and Suboptimal Finger Selection Algorithms for MMSE Rake
Receivers in Impulse Radio Ultra-Wideband Systems | 6,867 |
One of the features characterizing almost every multiple access (MA) communication system is the processing gain. Through the use of spreading sequences, the processing gain of Random CDMA systems (RCDMA), is devoted to both bandwidth expansion and orthogonalization of the signals transmitted by different users. Another type of multiple access system is Impulse Radio (IR). In many aspects, IR systems are similar to time division multiple access (TDMA) systems, and the processing gain of IR systems represents the ratio between the actual transmission time and the total time between two consecutive ransmissions (on-plus-off to on ratio). While CDMA systems, which constantly excite the channel, rely on spreading sequences to orthogonalize the signals transmitted by different users, IR systems transmit a series of short pulses and the orthogonalization between the signals transmitted by different users is achieved by the fact that most of the pulses do not collide with each other at the receiver. In this paper, a general class of MA communication systems that use both types of processing gain is presented, and both IR and RCDMA systems are demonstrated to be two special cases of this more general class of systems. The bit error rate (BER) of several receivers as a function of the ratio between the two types of processing gain is analyzed and compared under the constraint that the total processing gain of the system is large and fixed. It is demonstrated that in non inter-symbol interference (ISI) channels there is no tradeoff between the two types of processing gain. However, in ISI channels a tradeoff between the two types of processing gain exists. In addition, the sub-optimality of RCDMA systems in frequency selective channels is established. | On The Tradeoff Between Two Types of Processing Gain | 6,868 |
In this work, a non-cooperative power control game for multi-carrier CDMA systems is proposed. In the proposed game, each user needs to decide how much power to transmit over each carrier to maximize its overall utility. The utility function considered here measures the number of reliable bits transmitted per joule of energy consumed. It is shown that the user's utility is maximized when the user transmits only on the carrier with the best "effective channel". The existence and uniqueness of Nash equilibrium for the proposed game are investigated and the properties of equilibrium are studied. Also, an iterative and distributed algorithm for reaching the equilibrium (if it exists) is presented. It is shown that the proposed approach results in a significant improvement in the total utility achieved at equilibrium compared to the case in which each user maximizes its utility over each carrier independently. | A Non-Cooperative Power Control Game for Multi-Carrier CDMA Systems | 6,869 |
Transmission of information over a discrete-time memoryless Rician fading channel is considered where neither the receiver nor the transmitter knows the fading coefficients. First the structure of the capacity-achieving input signals is investigated when the input is constrained to have limited peakedness by imposing either a fourth moment or a peak constraint. When the input is subject to second and fourth moment limitations, it is shown that the capacity-achieving input amplitude distribution is discrete with a finite number of mass points in the low-power regime. A similar discrete structure for the optimal amplitude is proven over the entire SNR range when there is only a peak power constraint. The Rician fading with phase-noise channel model, where there is phase uncertainty in the specular component, is analyzed. For this model it is shown that, with only an average power constraint, the capacity-achieving input amplitude is discrete with a finite number of levels. For the classical average power limited Rician fading channel, it is proven that the optimal input amplitude distribution has bounded support. | The Noncoherent Rician Fading Channel -- Part I : Structure of the
Capacity-Achieving Input | 6,870 |
Transmission of information over a discrete-time memoryless Rician fading channel is considered where neither the receiver nor the transmitter knows the fading coefficients. The spectral-efficiency/bit-energy tradeoff in the low-power regime is examined when the input has limited peakedness. It is shown that if a fourth moment input constraint is imposed or the input peak-to-average power ratio is limited, then in contrast to the behavior observed in average power limited channels, the minimum bit energy is not always achieved at zero spectral efficiency. The low-power performance is also characterized when there is a fixed peak limit that does not vary with the average power. A new signaling scheme that overlays phase-shift keying on on-off keying is proposed and shown to be optimally efficient in the low-power regime. | The Noncoherent Rician Fading Channel -- Part II : Spectral Efficiency
in the Low-Power Regime | 6,871 |
In this paper, optimal power allocation and capacity regions are derived for GSIC (groupwise successive interference cancellation) systems operating in multipath fading channels, under imperfect channel estimation conditions. It is shown that the impact of channel estimation errors on the system capacity is two-fold: it affects the receivers' performance within a group of users, as well as the cancellation performance (through cancellation errors). An iterative power allocation algorithm is derived, based on which it can be shown that the total required received power is minimized when the groups are ordered according to their cancellation errors, and the first detected group has the smallest cancellation error. Performace/complexity tradeoff issues are also discussed by directly comparing the system capacity for different implementations: GSIC with linear minimum-mean-square error (LMMSE) receivers within the detection groups, GSIC with matched filter receivers, multicode LMMSE systems, and simple all matched filter receivers systems. | Capacity Regions and Optimal Power Allocation for Groupwise Multiuser
Detection | 6,872 |
Application of the turbo principle to multiuser decoding results in an exchange of probability distributions between two sets of constraints. Firstly, constraints imposed by the multiple-access channel, and secondly, individual constraints imposed by each users' error control code. A-posteriori probability computation for the first set of constraints is prohibitively complex for all but a small number of users. Several lower complexity approaches have been proposed in the literature. One class of methods is based on linear filtering (e.g. LMMSE). A more recent approach is to compute approximations to the posterior probabilities by marginalising over a subset of sequences (list detection). Most of the list detection methods are restricted to non-singular systems. In this paper, we introduce a transformation that permits application of standard tree-search methods to underdetermined systems. We find that the resulting tree-search based receiver outperforms existing methods. | A Tree Search Method for Iterative Decoding of Underdetermined Multiuser
Systems | 6,873 |
We consider the pulse design problem in multicarrier transmission where the pulse shapes are adapted to the second order statistics of the WSSUS channel. Even though the problem has been addressed by many authors analytical insights are rather limited. First we show that the problem is equivalent to the pure state channel fidelity in quantum information theory. Next we present a new approach where the original optimization functional is related to an eigenvalue problem for a pseudo differential operator by utilizing unitary representations of the Weyl--Heisenberg group.A local approximation of the operator for underspread channels is derived which implicitly covers the concepts of pulse scaling and optimal phase space displacement. The problem is reformulated as a differential equation and the optimal pulses occur as eigenstates of the harmonic oscillator Hamiltonian. Furthermore this operator--algebraic approach is extended to provide exact solutions for different classes of scattering environments. | A Group-Theoretic Approach to the WSSUS Pulse Design Problem | 6,874 |
A new design method for high rate, fully diverse ('spherical') space frequency codes for MIMO-OFDM systems is proposed, which works for arbitrary numbers of antennas and subcarriers. The construction exploits a differential geometric connection between spherical codes and space time codes. The former are well studied e.g. in the context of optimal sequence design in CDMA systems, while the latter serve as basic building blocks for space frequency codes. In addition a decoding algorithm with moderate complexity is presented. This is achieved by a lattice based construction of spherical codes, which permits lattice decoding algorithms and thus offers a substantial reduction of complexity. | Space Frequency Codes from Spherical Codes | 6,875 |
In this paper have written the results of the information analysis of structures. The obtained information estimation (IE) are based on an entropy measure of C. Shannon. Obtained IE is univalent both for the non-isomorphic and for the isomorphic graphs, algorithmically, it is asymptotically steady and has vector character. IE can be used for the solution of the problems ranking of structures by the preference, the evaluation of the structurization of subject area, the solution of the problems of structural optimization. Information estimations and method of the information analysis of structures it can be used in many fields of knowledge (Electrical Systems and Circuit, Image recognition, Computer technology, Databases and Bases of knowledge, Organic chemistry, Biology and others) and it can be base for the structure calculus. | Information estimations and analysis of structures | 6,876 |
This paper presents a stochastic algorithm for iterative error control decoding. We show that the stochastic decoding algorithm is an approximation of the sum-product algorithm. When the code's factor graph is a tree, as with trellises, the algorithm approaches maximum a-posteriori decoding. We also demonstrate a stochastic approximations to the alternative update rule known as successive relaxation. Stochastic decoders have very simple digital implementations which have almost no RAM requirements. We present example stochastic decoders for a trellis-based Hamming code, and for a Block Turbo code constructed from Hamming codes. | Stochastic Iterative Decoders | 6,877 |
We consider the problem of nonlinear dimensionality reduction: given a training set of high-dimensional data whose ``intrinsic'' low dimension is assumed known, find a feature extraction map to low-dimensional space, a reconstruction map back to high-dimensional space, and a geometric description of the dimension-reduced data as a smooth manifold. We introduce a complexity-regularized quantization approach for fitting a Gaussian mixture model to the training set via a Lloyd algorithm. Complexity regularization controls the trade-off between adaptation to the local shape of the underlying manifold and global geometric consistency. The resulting mixture model is used to design the feature extraction and reconstruction maps and to define a Riemannian metric on the low-dimensional data. We also sketch a proof of consistency of our scheme for the purposes of estimating the unknown underlying pdf of high-dimensional data. | A complexity-regularized quantization approach to nonlinear
dimensionality reduction | 6,878 |
The Gallager bound is well known in the area of channel coding. However, most discussions about it mainly focus on its applications to memoryless channels. We show in this paper that the bounds obtained by Gallager's method are very tight even for general sources and channels that are defined in the information-spectrum theory. Our method is mainly based on the estimations of error exponents in those bounds, and by these estimations we proved the direct part of the Slepian-Wolf theorem and channel coding theorem for general sources and channels. | Some Extensions of Gallager's Method to General Sources and Channels | 6,879 |
We show how to construct an algorithm to search for binary idempotents which may be used to construct binary LDPC codes. The algorithm, which allows control of the key properties of sparseness, code rate and minimum distance, is constructed in the Mattson-Solomon domain. Some of the new codes, found by using this technique, are displayed. | Idempotents, Mattson-Solomon Polynomials and Binary LDPC codes | 6,880 |
Cycle codes are a special case of low-density parity-check (LDPC) codes and as such can be decoded using an iterative message-passing decoding algorithm on the associated Tanner graph. The existence of pseudo-codewords is known to cause the decoding algorithm to fail in certain instances. In this paper, we draw a connection between pseudo-codewords of cycle codes and the so-called edge zeta function of the associated normal graph and show how the Newton polyhedron of the zeta function equals the fundamental cone of the code, which plays a crucial role in characterizing the performance of iterative decoding algorithms. | Pseudo-Codewords of Cycle Codes via Zeta Functions | 6,881 |
It is shown that some well-known and some new cyclic codes with orthogonal parity-check equations can be constructed in the finite-field transform domain. It is also shown that, for some binary linear cyclic codes, the performance of the iterative decoder can be improved by substituting some of the dual code codewords in the parity-check matrix with other dual code codewords formed from linear combinations. This technique can bring the performance of a code closer to its maximum-likelihood performance, which can be derived from the erroneous decoded codeword whose euclidean distance with the respect to the received block is smaller than that of the correct codeword. For (63,37), (93,47) and (105,53) cyclic codes, the maximum-likelihood performance is realised with this technique. | Near Maximum-Likelihood Performance of Some New Cyclic Codes Constructed
in the Finite-Field Transform Domain | 6,882 |
An algorithm of improving the performance of iterative decoding on perpendicular magnetic recording is presented. This algorithm follows on the authors' previous works on the parallel and serial concatenated turbo codes and low-density parity-check codes. The application of this algorithm with signal-to-noise ratio mismatch technique shows promising results in the presence of media noise. We also show that, compare to the standard iterative decoding algorithm, an improvement of within one order of magnitude can be achieved. | Improved Iterative Decoding for Perpendicular Magnetic Recording | 6,883 |
Based on the ideas of cyclotomic cosets, idempotents and Mattson-Solomon polynomials, we present a new method to construct GF(2^m), where m>0 cyclic low-density parity-check codes. The construction method produces the dual code idempotent which is used to define the parity-check matrix of the low-density parity-check code. An interesting feature of this construction method is the ability to increment the code dimension by adding more idempotents and so steadily decrease the sparseness of the parity-check matrix. We show that the constructed codes can achieve performance very close to the sphere-packing-bound constrained for binary transmission. | GF(2^m) Low-Density Parity-Check Codes Derived from Cyclotomic Cosets | 6,884 |
We present a unified large system analysis of linear receivers for a class of random matrix channels. The technique unifies the analysis of both the minimum-mean-squared-error (MMSE) receiver and the adaptive least-squares (ALS) receiver, and also uses a common approach for both random i.i.d. and random orthogonal precoding. We derive expressions for the asymptotic signal-to-interference-plus-noise (SINR) of the MMSE receiver, and both the transient and steady-state SINR of the ALS receiver, trained using either i.i.d. data sequences or orthogonal training sequences. The results are in terms of key system parameters, and allow for arbitrary distributions of the power of each of the data streams and the eigenvalues of the channel correlation matrix. In the case of the ALS receiver, we allow a diagonal loading constant and an arbitrary data windowing function. For i.i.d. training sequences and no diagonal loading, we give a fundamental relationship between the transient/steady-state SINR of the ALS and the MMSE receivers. We demonstrate that for a particular ratio of receive to transmit dimensions and window shape, all channels which have the same MMSE SINR have an identical transient ALS SINR response. We demonstrate several applications of the results, including an optimization of information throughput with respect to training sequence length in coded block transmission. | Unified Large System Analysis of MMSE and Adaptive Least Squares
Receivers for a class of Random Matrix Channels | 6,885 |
Generalisations of the bent property of a boolean function are presented, by proposing spectral analysis with respect to a well-chosen set of local unitary transforms. Quadratic boolean functions are related to simple graphs and it is shown that the orbit generated by successive Local Complementations on a graph can be found within the transform spectra under investigation. The flat spectra of a quadratic boolean function are related to modified versions of its associated adjacency matrix. | Generalised Bent Criteria for Boolean Functions (I) | 6,886 |
In the first part of this paper [16], some results on how to compute the flat spectra of Boolean constructions w.r.t. the transforms {I,H}^n, {H,N}^n and {I,H,N}^n were presented, and the relevance of Local Complementation to the quadratic case was indicated. In this second part, the results are applied to develop recursive formulae for the numbers of flat spectra of some structural quadratics. Observations are made as to the generalised Bent properties of boolean functions of algebraic degree greater than two, and the number of flat spectra w.r.t. {I,H,N}^n are computed for some of them. | Generalised Bent Criteria for Boolean Functions (II) | 6,887 |
We present an efficient, low-cost implementation of time-hopping impulse radio that fulfills the spectral mask mandated by the FCC and is suitable for high-data-rate, short-range communications. Key features are: (i) all-baseband implementation that obviates the need for passband components, (ii) symbol-rate (not chip rate) sampling, A/D conversion, and digital signal processing, (iii) fast acquisition due to novel search algorithms, (iv) spectral shaping that can be adapted to accommodate different spectrum regulations and interference environments. Computer simulations show that this system can provide 110Mbit/s at 7-10m distance, as well as higher data rates at shorter distances under FCC emissions limits. Due to the spreading concept of time-hopping impulse radio, the system can sustain multiple simultaneous users, and can suppress narrowband interference effectively. | A low-cost time-hopping impulse radio system for high data rate
transmission | 6,888 |
We present an interleaving scheme that yields quasi-cyclic turbo codes. We prove that randomly chosen members of this family yield with probability almost 1 turbo codes with asymptotically optimum minimum distance, i.e. growing as a logarithm of the interleaver size. These interleavers are also very practical in terms of memory requirements and their decoding error probabilities for small block lengths compare favorably with previous interleaving schemes. | On quasi-cyclic interleavers for parallel turbo codes | 6,889 |
In this paper, a class of nonlinear MMSE multiuser detectors are derived based on a multivariate Gaussian approximation of the multiple access interference. This approach leads to expressions identical to those describing the probabilistic data association (PDA) detector, thus providing an alternative analytical justification for this structure. A simplification to the PDA detector based on approximating the covariance matrix of the multivariate Gaussian distribution is suggested, resulting in a soft interference cancellation scheme. Corresponding multiuser soft-input, soft-output detectors delivering extrinsic log-likelihood ratios are derived for application in iterative multiuser decoders. Finally, a large system performance analysis is conducted for the simplified PDA, showing that the bit error rate performance of this detector can be accurately predicted and related to the replica method analysis for the optimal detector. Methods from statistical neuro-dynamics are shown to provide a closely related alternative large system prediction. Numerical results demonstrate that for large systems, the bit error rate is accurately predicted by the analysis and found to be close to optimal performance. | Nonlinear MMSE Multiuser Detection Based on Multivariate Gaussian
Approximation | 6,890 |
The performance of second order statistics (SOS) based semi-blind channel estimation in long-code DS-CDMA systems is analyzed. The covariance matrix of SOS estimates is obtained in the large system limit, and is used to analyze the large-sample performance of two SOS based semi-blind channel estimation algorithms. A notion of blind estimation efficiency is also defined and is examined via simulation results. | Analysis of Second-order Statistics Based Semi-blind Channel Estimation
in CDMA Channels | 6,891 |
The achievable information rate of finite-state input two-dimensional (2-D) channels with memory is an open problem, which is relevant, e.g., for inter-symbol-interference (ISI) channels and cellular multiple-access channels. We propose a method for simulation-based computation of such information rates. We first draw a connection between the Shannon-theoretic information rate and the statistical mechanics notion of free energy. Since the free energy of such systems is intractable, we approximate it using the cluster variation method, implemented via generalized belief propagation. The derived, fully tractable, algorithm is shown to provide a practically accurate estimate of the information rate. In our experimental study we calculate the information rates of 2-D ISI channels and of hexagonal Wyner cellular networks with binary inputs, for which formerly only bounds were known. | On the Achievable Information Rates of Finite-State Input
Two-Dimensional Channels with Memory | 6,892 |
We define multilevel codes on bipartite graphs that have properties analogous to multilevel serial concatenations. A decoding algorithm is described that corrects a proportion of errors equal to half the Blokh-Zyablov bound on the minimum distance. The error probability of this algorithm has exponent similar to that of serially concatenated multilevel codes. | Multilevel expander codes | 6,893 |
The problems of sensor configuration and activation for the detection of correlated random fields using large sensor arrays are considered. Using results that characterize the large-array performance of sensor networks in this application, the detection capabilities of different sensor configurations are analyzed and compared. The dependence of the optimal choice of configuration on parameters such as sensor signal-to-noise ratio (SNR), field correlation, etc., is examined, yielding insights into the most effective choices for sensor selection and activation in various operating regimes. | Sensor Configuration and Activation for Field Detection in Large Sensor
Arrays | 6,894 |
Spectral properties and performance of multi-pulse impulse radio ultra-wideband systems with pulse-based polarity randomization are analyzed. Instead of a single type of pulse transmitted in each frame, multiple types of pulses are considered, which is shown to reduce the effects of multiple-access interference. First, the spectral properties of a multi-pulse impulse radio system is investigated. It is shown that the power spectral density is the average of spectral contents of different pulse shapes. Then, approximate closed-form expressions for bit error probability of a multi-pulse impulse radio system are derived for RAKE receivers in asynchronous multiuser environments. The theoretical and simulation results indicate that impulse radio systems that are more robust against multiple-access interference than a "classical" impulse radio system can be designed with multiple types of ultra-wideband pulses. | Impulse Radio Systems with Multiple Types of Ultra-Wideband Pulses | 6,895 |
Density evolution (DE) is one of the most powerful analytical tools for low-density parity-check (LDPC) codes on memoryless binary-input/symmetric-output channels. The case of non-symmetric channels is tackled either by the LDPC coset code ensemble (a channel symmetrizing argument) or by the generalized DE for linear codes on non-symmetric channels. Existing simulations show that the bit error rate performances of these two different approaches are nearly identical. This paper explains this phenomenon by proving that as the minimum check node degree $d_c$ becomes sufficiently large, the performance discrepancy of the linear and the coset LDPC codes is theoretically indistinguishable. This typicality of linear codes among the LDPC coset code ensemble provides insight into the concentration theorem of LDPC coset codes. | On the Typicality of the Linear Code Among the LDPC Coset Code Ensemble | 6,896 |
This paper investigates decoding of binary linear block codes over the binary erasure channel (BEC). Of the current iterative decoding algorithms on this channel, we review the Recovery Algorithm and the Guess Algorithm. We then present a Multi-Guess Algorithm extended from the Guess Algorithm and a new algorithm -- the In-place Algorithm. The Multi-Guess Algorithm can push the limit to break the stopping sets. However, the performance of the Guess and the Multi-Guess Algorithm depend on the parity-check matrix of the code. Simulations show that we can decrease the frame error rate by several orders of magnitude using the Guess and the Multi-Guess Algorithms when the parity-check matrix of the code is sparse. The In-place Algorithm can obtain better performance even if the parity check matrix is dense. We consider the application of these algorithms in the implementation of multicast and broadcast techniques on the Internet. Using these algorithms, a user does not have to wait until the entire transmission has been received. | A New Non-Iterative Decoding Algorithm for the Erasure Channel :
Comparisons with Enhanced Iterative Methods | 6,897 |
We propose a technique to derive upper bounds on Gallager's cost-constrained random coding exponent function. Applying this technique to the non-coherent peak-power or average-power limited discrete time memoryless Ricean fading channel, we obtain the high signal-to-noise ratio (SNR) expansion of this channel's cut-off rate. At high SNR the gap between channel capacity and the cut-off rate approaches a finite limit. This limit is approximately 0.26 nats per channel-use for zero specular component (Rayleigh) fading and approaches 0.39 nats per channel-use for very large specular components. We also compute the asymptotic cut-off rate of a Rayleigh fading channel when the receiver has access to some partial side information concerning the fading. It is demonstrated that the cut-off rate does not utilize the side information as efficiently as capacity, and that the high SNR gap between the two increases to infinity as the imperfect side information becomes more and more precise. | Duality Bounds on the Cut-Off Rate with Applications to Ricean Fading | 6,898 |
A binary linear error correcting codes represented by two code families Kronecker products sum are considered. The dimension and distance of new code is investigated. Upper and lower bounds of distance are obtained. Some examples are given. It is shown that some classic constructions are the private cases of considered one. The subclass of codes with equal lower and upper distance bounds is allocated. | On a Kronecker products sum distance bounds | 6,899 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.