Between Dec 19, 2024 and Jan 2, 2025, datasets can be submitted to DRUM but will not be processed until after the break. Staff will not be available to answer email during this period, and will not be able to provide DOIs until after Jan 2. If you are in need of a DOI during this period, consider Dryad or OpenICPSR. Submission responses to the UDC may also be delayed during this time.

Browsing by Author "Yang, Jie"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Adaptive filter design for sparse signal estimation.
    (2011-12) Yang, Jie
    Recently, sparse signal estimation has become an increasingly important research area in signal processing due to its wide range of applications. Efficient adaptive algorithms have been developed for estimation of various sparse signals, and the approaches developed are usually application-specific. In this dissertation, we investigate the algorithm and system design for sparse signal estimation of several applications of practical interest, specifically echo cancellation, compressive sensing, and power amplifier pre-distortion. For echo cancellation, different approaches are considered to find the optimal solution. A series of algorithms are proposed to improve the performance and reduce the cost. Specifically, we describe novel adaptive tap algorithms with selective update criteria, a μ-law proportionate technique incorporated with efficient memorized proportionate Affine Projection Algorithms, and a new class of proportionate algorithms with gradient-controlled individual step sizes which can be implemented either in the time domain or the frequency domain. For compressive sensing algorithms with the l0 norm constraint, a sparse LMS algorithm with segment zero attractors is introduced. It can achieve significant convergence and error performance improvements while providing reduced computational cost, especially for large sparse systems with colored inputs. Such filters can also be combined with cascade or multistage realizations, thereby yielding even more efficient implementations. We also describe new results for the non-linear signal estimation problem in power amplifier (PA) pre-distortion with dynamic nonlinearities, where the signal can be represented using a Volterra series with sparse coefficients. An efficient solution using a power-indexed look-up table (LUT) based digital pre-distortion (DPD) is proposed to address the current challenge of poor dynamic performance of a PA operating with wideband signals. Experimental results obtained using a 2 GHz power amplifier driven by a 2-carrier WCDMA signal demonstrate very robust and stable performance for the PA in dynamic environments.
  • Thumbnail Image
    Item
    Development and Validation of a Physics Problem Difficulty Measure
    (2019-08) Yang, Jie
    Problem solving is both a primary goal and a standard teaching technique in introductory physics classes at the university level. To assess the utility of various pedagogical materials and techniques, it is necessary to determine student problem solving performance in the authentic situation of the course. However, this performance depends on both the student’s problem-solving skill and the problem difficulty. This dissertation proposes a technique for measuring the relative difficulty of the type of physics problems typically used in introductory physics courses for physical science and engineering students. Four categories, the problem context, the physics principles, the mathematical complexity, and the number of words in the problem, were constructed based on current cognitive theories. To test the validity of this measure, 3552 student grades on 20 final examination problems, spanning the full range of topics in a one-year introductory physics course, were compared to each problem difficulty rating. Only two categories, physics principles and mathematical complexity, were needed to account for most of the student problem solving variance. Using the average of those two categories, there was an 88% Pearson correlation between the difficulty score and the students’ problem solving grade with a P value < 0.001, with null hypothesis is the correlation coefficient between difficulty score and the students’ problem solving grade is not significantly different from zero. Three experts used the difficulty measure to test its reliability and had a pairwise Spearman correlation between their difficulty ratings of greater than 94%.