Abstract
The acceleration of applications, running on a general purpose processor (GPP), by mapping parts of their execution to reconfigurable hardware is an approach which does not involve program's source code and still ensures program portability over different target reconfigurable fabrics. However, the problem is very challenging, as suitable sequences of GPP instructions need to be translated/mapped to hardware, possibly at runtime. Thus, all mapping steps, from compiler analysis and optimizations to hardware generation, need to be both efficient and fast. This paper introduces some of the most representative approaches for binary acceleration using reconfigurable hardware, and presents our binary acceleration approach and the latest results. Our approach extends a GPP with a Reconfigurable Processing Unit (RPU), both sharing the data memory. Repeating sequences of GPP instructions are migrated to an RPU composed of functional units and interconnect resources, and able to exploit instruction-level parallelism, e.g., via loop pipelining. Although we envision a fully dynamic system, currently the RPU resources are selected and organized offline using execution trace information. We present implementation prototypes of the system on a Spartan-6 FPGA with a MicroBlaze as GPP and the very encouraging results achieved with a number of benchmarks.

Abstract
This paper describes and evaluates a fully digital circuit for one-way master-to-slave, highly precise time synchronization in a low-power wearable system equipped with a set of sensor nodes. These sensors are connected to each other in a mesh topology, with conductive yarns used as one-wire bidirectional communication links. The circuit is designed to perform synchronization in the MAC layer, so that the deterministic part of the clock skew between nodes is kept constant and compensated with a single message exchange. In each sensor node, the synchronization circuit provides a programmable clock signal and a real-time counter for time stamping. Experimental results from a fabricated ASIC (in a CMOS 0.35 mu m technology) show that the circuit keeps the one-hop average clock skew below 4.6 ns and that the skew grows linearly as the hop distance to the reference node increases. The sub-microsecond average clock skew achieved by the proposed solution satisfies the requirements of many wearable sensor network applications.

Abstract
The goal of this study was to analyse perceptually and acoustically the voices of patients with Unilateral Vocal Fold Paralysis (UVFP) and compare them to the voices of normal subjects. These voices were analysed perceptually with the GRBAS scale and acoustically using the following parameters: mean fundamental frequency (F0), standard-deviation of F0, jitter (ppq5), shimmer (apq11), mean harmonics-to-noise ratio (HNR), mean first (F1) and second (F2) formants frequency, and standard-deviation of F1 and F2 frequencies. Statistically significant differences were found in all of the perceptual parameters. Also the jitter, shimmer, HNR, standard-deviation of F0, and standard-deviation of the frequency of F2 were statistically different between groups, for both genders. In the male data differences were also found in F1 and F2 frequencies values and in the standard-deviation of the frequency of F1. This study allowed the documentation of the alterations resulting from UVFP and addressed the exploration of parameters with limited information for this pathology.

Abstract
Objectives/Hypothesis. This article aims to establish correlations between acoustic and audio-perceptual measures using the GRBAS scale with respect to four different voice analysis software programs. Study Design. Exploratory, transversal. Methods. A total of 90 voice records were collected and analyzed with the Dr. Speech (Tiger Electronics, Seattle, WA), Multidimensional Voice Program (Kay Elemetrics, NJ, USA), PRAAT (University of Amsterdam, The Netherlands), and Voice Studio (Seegnal, Oporto, Portugal) software programs. The acoustic measures were correlated to the audio-perceptual parameters of the GRBAS and rated by 10 experts. Results. The predictive value of the acoustic measurements related to the audio-perceptual parameters exhibited magnitudes ranging from weak (R-a(2) = 0.17) to moderate (R-a(2) = 0.71). The parameter exhibiting the highest correlation magnitude is B (Breathiness), whereas the weaker correlation magnitudes were found to be for A (Asthenia) and S (Strain). The acoustic measures with stronger predictive values were local Shimmer, harmonics-to-noise ratio, APQ5 shimmer, and PPQ5 jitter, with different magnitudes for each one of the studied software programs. Conclusions. Some acoustic measures are pointed as significant predictors of GRBAS parameters, but they differ among software programs. B (Breathiness) was the parameter exhibiting the highest correlation magnitude.

Abstract
Narrow band parametric speech coding and wideband audio coding represent opposite coding paradigms involving audible information, namely in terms of the specificity of the audio material, target bit rates, audio quality and application scenarios. In this paper we explore a new avenue addressing parametric coding of wideband speech, using the potential and accuracy provided by frequency-domain signal analysis and modeling techniques that typically belong to the realm of high-quality audio coding. A first analysis-synthesis validation framework is described that illustrates the decomposition, parametric representation and synthesis of perceptually and linguistically relevant speech components while preserving naturalness and speaker specific information.

Abstract
Real-time monitoring applications may generate delay sensitive traffic that is expected to be delivered within a firm delay boundary in order to be useful. In this context, a previous work proposed an End-to-End Delay (EED) estimation mechanism for Wireless Sensor Networks (WSNs) to preview potential useless packets, and to early discard them in order to save processing and energy resources. Such estimation mechanism accounts delays using timers that make use of an Exponentially Weighted Moving Average (EWMA) function where the smoothing factor is a constant defined prior to the WSN deployment. Later experiments showed that, in order to enhance the estimation results, such smoothing factor should be defined as a function of the network load. The current work proposes an optimization of the previous estimation mechanism that works by evaluating the network load and by adapting the smoothing factor of the EWMA function accordingly. Results show that this optimization leads to a more accurate EED estimation for different network loads.