This paper proposes a single instruction multiple data (SIMD) processor, which is programmed with high-level OpenCL language. The low-power processor is customized for executing multiple-input-multiple-output (MIMO) detection algorithms at a high performance while consuming very little power making it suitable for software-defined radio (SDR) applications. The novel combination of SIMD operations on a transport programmed multicore datapath allows saving power on both the execution front end and the back end, leading to very good energy efficiency with a compiler programmable design. We demonstrate the feasibility of the architecture with the layered orthogonal lattice detector and minimum mean-square-error MIMO algorithms, which can be used as a software-defined radio implementation of the 3GPP local thermal equilibrium r11 standard. Compared to other state-of-the-art SDR architectures, the proposed design adds features that improve programmer productivity with an insignificant power and area impact.

In this paper we investigate and enhance the user-pairing probability in a system employing multi-user superposition transmission (MUST). In order to improve the multiuser pairing probability, we propose an additional feedback of the second best channel state information (CSI) consisting of channel quality indicator (CQI) and precoding matrix indicator (PMI). We analyze the pairing-probability and show that the additional feedback increases significantly the pairing possibilities at the scheduler. By system level simulations we confirm that the proposed enhanced feedback is improving significantly MUST performance in the context of MUST operation on the same-beam with Gray-mapped super-constellation. In addition, we suggest a simple link-to-system mapping for maximum likelihood (ML) MUST receiver, which can re-use legacy mutual-information-to-block-error-rate tables.

The choice of basis for linear matrix modulation (linear space-time code with linear combination constellation) is considered. Unitarily invariant polynomials of square matrices are discussed, the full spectrum of invariants interpolating between the well-known trace and determinant. These give the full spectrum of space-time code design criteria. The diagonal dominance (expansion around the trace) of these invariants is considered. Using this, it is shown that minimizing the self-interference, or equivalently, maximizing the second order expansion coefficient of the mutual information around SNR=0, is required when maximizing the mutual information and/or optimizing performance at any SNR. As an example, symbol rate 3 schemes for 4 transmit antennas are considered.

The article discusses soft-decision decoding of binary linear block codes using the t-algorithm and its variants. New variants of the basic algorithm are presented that reduce the decoding complexity using a threshold adaptive to the signal-to-noise ratio and address the variable decoding complexity by either limiting the memory or using a generalized M-algorithm with a nonconstant state profile.

This paper proposes a new and computationally efficient soft-decision decoding algorithm for binary linear block codes. It uses the tree representation of a systematically encoded block code. The tree representation is traversed through in a breadth-first fashion and it is adaptively pruned during the decoding using the following principle: if the metric difference between the best partial path and any other partial path at the same level is greater than a threshold value, then such a path is not extended any further. The threshold value is chosen based on a signal-to-noise ratio estimate. The efficiency of the pruning procedure is enhanced by constructing the tree so that the most reliable symbols are processed first. We show that practically optimum bit error rate performance in the additive white Gaussian noise channel can be efficiently achieved.

The authors demonstrate that the self-organizing maps (SOMs) of Kohonen can be used as speech feature extractors that are able to take temporal context into account. They have investigated two alternatives for using SOMs as such feature extractors, one based on tracing the location of highest activity on a SOM, the other on integrating the activity of the whole SOM for a period of time. The experiments indicated that an improvement is achievable by using these methods.>

Self-organizing feature maps (SOFMs) are used as speech feature extractors followed by a classifier based on multilayer feedforward networks. Usually SOFMs have been used in speech recognition as static pattern classifiers or vector quantizers, ignoring their property of preserving the local topology of input pattern space. Here, the topological ordering of the acoustic speech data in the SOFM is utilized to form trajectories in the map which are then fed into a classifier. Viewing the trajectories at multiple resolution levels, feature vectors are formed that take contextual information into account. Experiments with such feature vectors indicate that better accuracies can be obtained than by using a simple SOFM classifier based on instantaneous acoustic features.>

This paper focuses on comparing three approaches to improve the accuracy of classifying short-time speech frames into phoneme classes by taking into account the classifications of nearby frames, also individually classified. We investigate whether this improvement has an effect to the accuracy of transcribing speech into phoneme sequences using two different decoding schemes, one based on simple durational rules, and the other on hidden Markov models (HMMs). The experiments indicate that recognition accuracies can indeed be improved significantly by taking the local context into account. 1 INTRODUCTION "More is to be gained by discovering suboptimal ways of handling context than by discovering optimal ways of handling the local structure", Haralick stated in [2]. In this paper we explore some options to take advantage of local context in improving classification accuracy. Our framework is a phonemic speech recognizer based on classifying shorttime feature vectors. The viewpoint taken i...

Abstract A new approach to construct phonemic transcriptions of spoken utterances is described. The Self-Organizing Feature Maps by Kohonen are first applied to vector-quantize speech into a sequence of phoneme labels a centisecond apart. This code sequence is converted into a phoneme string using a multi-layered feed-forward network trained with error back propagation. The trained network acts as a filter removing undesired transitional and coarticulatory effects from the code sequence. This makes it almost a trivial task to convert the code sequence into a phoneme sequence. The need for any statistical speech models, such as Hidden Markov Models, is thus eliminated. The new approach is compared to an existing one being used in a speech recognition system, in which simple durational rules are used for the same transformation task. The accuracy of produced phonemic transcriptions is 4.8 per cent units better using the proposed multi-layered network approach (88.4% opposed to 83.6%).