Session S2

Session S2: Industry DSP, Communications and Sensors

Time:                 Monday, May 12, 10:30-12:00
Chair:                 Ghassan Maalouli, Raytheon Missile Systems
Co-Chair:          Issa Panahi, University of Texas at Dallas

S2-1: Music Injection For Subjective Speech Enhancement and the Psychoacoustic Pleasantness Analysis
Hua Bao, Issa Panahi, Philipos Loizou and Yi Hu University of Texas at Dallas, Dallas TX
In this paper we propose a method for subjective speech enhancement. Traditional noise reduction methods, although powerful in improving the SNR of speech signal, have the drawback of generating unpleasant residual noise, which makes the processed speech annoying. Music injection is adopted to improve the speech quality with respect to pleasantness. A prediction model of pleasantness based on four psychoacoustic measures (loudness, sharpness, roughness and tonality) is used to evaluate the proposed method. Three different types of instrumental music are taken to analyze the performance of our method. Subjective tests are also conducted to verify the objective pleasantness model.

S2-2: Incorporating Forward Error Correction Codes into FlexRay Communications
Fei Ren, Y. Rosa Zheng and Jagannathan Sarangapani University of Missouri-Rolla, Rolla MO
This paper proposes a new error handling method to the FlexRay automobile communications system. Currently, FlexRay signal frames employ two Cyclic Redundancy Check (CRC) codes for error detection. The proposed method applies Forward Error Correction (FEC) codes to the data field of FlexRay frames replacing the second CRC code. This method can be applied to short control messages that require stringent real-time processing. It is shown, using several BCH codes as examples, that this approach improves the probability of correct reception and reduces retransmission rate in hash Electromagnetic Interference (EMI) environments where burst errors are frequently encountered. Another advantage is that it only requires different processing at the codec without changing the frame structure.

S2-3: Extended Risk-Sensitive Filters and Its Applications to Tracking with Radar/Sonar/IR Sensor
Srinivasan Munisamy Koneru Lakshmaiah College of Engineering, India, S. Sadhu and Tapan Kumar Ghoshal Jadavpur University, India
This paper reviews different forms of Risk-Sensitive Filters (RSF) and generalizes them to nonlinear systems by approximating nonlinear processes and measurement functions using Taylor's series; they are called extended risk sensitive filters (ERSF). Three variants of ERSF algorithms which have differences in the propagation of the Riccati variable and the structure for the prior and a posterior estimates are applied to target tracking and its performance is compared with Cramer-Rao Lower Bound. To explore the superiority of the ERSF over the conventional EKF, some examples of target tracking with a radar/sonar sensor are considered. Robust tracking against initial uncertainty has been achieved by applying ERSF with proper tuning of risk factor. Various issues associated with minimum mean-squared, affine and robust recursive tracking based on the risk-sensitive approach are presented.

S2-4: Real-Time Acoustic Scene Characterization with Wireless Sensor Networks
Homin Kwon, Visar Berisha and Andreas Spanias Arizona State University, Tempe AZ
Real-time acoustic scene analysis has several applications such as homeland security, surveillance, and monitoring. The development of a collaborative networking infrastructure can be valuable in scene analysis since feature parameters can be extracted locally (at the node level) and combined at the base station. In this context, distributed and agile wireless sensor networks (WSNs) have been of particular interest recently. In this paper, we present hierarchy-based acoustic scene characterization algorithms for use in a real-time wireless sensor network. Voice scene analysis is accomplished using a speech discriminator, a gender classifier, a system for recognizing the state of emotion, an estimator of the number of speakers, and voice monitoring in an area of interest. Real-time implementations of these algorithms are accomplished using Crossbow motes and TI DSP boards, configured to operate in a wireless sensor network. A series of experiments are presented that characterize the performance of the algorithms under different conditions.

S2-5: Estimation of Time-Varying Multipath Channel Using a Complex Kalman Filter
Ryan Reed and Ghassan Maalouli Raytheon Missile Systems, Tucson AZ
The presence of multipath hinders robust communications in a wireless channel. Distortion is exacerbated when multipath is time-varying in nature. Several models have been proposed to represent time-varying channels. Some models are deterministic in nature, such as the Complex-Exponential, Basis-Expansion model (CE-BEM). Others are statistical such as the Rayleigh and Autoregressive (AR) channel models. The Kalman filter has been proposed as a method to estimate an AR multipath channel. In this paper, we present a treatment where the estimation of the AR channel is cast into a complex Kalman filter framework. We characterize the performance of the channel estimator in terms of mean-square-error. We demonstrate the sensitivity of the estimator to process noise.

S2-6: A Review of Digital Predistortion Algorithms
Robert Santucci Intel Corporation, Chandler AZ
This paper provides a summary of digital predistoriton techniques in use today. First, it introduces power amplifiers and provides a summary of their non-linearity. Next, it presents how this amplifier non-linearity has a detrimental effect on communications signals when amplitude modulation is present. After the requirement for amplifier linearization is established, it presents several different techniques for implementing linearization in the digital domain, known as digital predistortion. An emphasis is placed on techniques that adaptively generate signals to invert the amplifier non-linearity. Finally, a brief description of non-idealities in real systems that limit effectiveness is presented.

S2-7: Sensing and Processing With Orbital Angular Momentum
Harry Schmitt, Donald Waagen, Nitesh Shah Raytheon Missile Systems, Tucson AZ, A. D. Greentree, W. Moran and S. Skafidas University of Melbourne, Carlton VIC
This paper discusses potential defense and security benefits that might be realized through the exploitation of photon orbital angular momentum (OAM). While it was realized early on that beams with OAM can be “used to investigate all the analogues of polarized light [3],” efforts have been largely applied to quantum mechanical phenomena. This analogy is extended here to the classical realm by identifying a broad collection of sensor and signal processing applications. Optical Tagging is highlighted as an interesting example that exploits the unique signature afforded by classical OAM beams.

S2-8: An Unbiased Multipath Mitigating Delay-Lock-Loop Correlation Kernel for Unbalanced Direct Sequence Spreading Codes
Steven Miller Hemisphere GPS, Scottsdale AZ
Multipath is the dominant error source in precise positioning Global Navigation Satellite Systems (GNSS) such as the United States Global Positioning System (GPS). These systems utilize the satellite signal time of arrival estimates to solve for position. Multipath corrupts the time of arrival estimates by distorting the signal tracking phase discriminator; which results in a slowly time-varying phase bias. This bias ranges from several centimeters to tens of meters. The sub-meter bias is the most problematic for centimeter positioning systems. Moreover, in addition to multipath, the GPS spread spectrum code is unbalanced for certain space vehicles which can lead to a code tracking phase bias. Current multipath mitigating solutions include windowed delay-lock loop discriminators within the code tracking and synchronization block, but do not compensate for the unbalanced spreading codes. A new correlation kernel is proposed as part of a multipath mitigating delay-locked-loop code phase discriminator that does not suffer a tracking bias due to unbalanced spreading codes. This new algorithm performance is compared to existing techniques with respect to position bias, variance and robustness.