An Improved T/2 Fractionally Spaced Blind Equalization Algorithm with Coordinate Transform

다국어 초록 (Multilingual Abstract)

To address the slow convergence and large residual error problems in the T/2 fractionally spaced equalization with constant modulus algorithm (T/2-FSE-CMA) for high-order QAM signal in wireless communication, we analyze the distribution characters of 16QAM signal constellations and propose an improved T/2 fractionally spaced blind equalization with coordinate transform and constant modulus algorithm (T/2-FSE-RCTCMA). In this algorithm, we equalize the real and imaginary parts of the input signal in the fractionally spaced equalizer separately. By using the coordinate transform the output signals are mapped into the same circle and the error function of constant modulus is achieved. Using the error function to adjust the weight vector of each sub-channel in the fractionally spaced equalizer, we may avoid the miss-detection caused by the T/2 FSE-CMA algorithm in multi-modulus high-order QAM signal equalization. The simulation results in wireless channel show that compared with T/2-FSE-CTCMA and T/2- FSE-CMA, the proposed algorithm has a faster convergence speed and a smaller residual error.

목차 (Table of Contents)

• Abstract
• 1. Introduction
• 2. Fractionally Spaced Blind Equalization Algorithm
• 3. Improved T/2 Fractionally Spaced Coordinates Transform BlindEqualization Algorithm
• 4. Algorithm Performance Analysis
• 4.1. Convergence Analysis
• 4.2. Analysis of Computation Load
• 5. Simulation Experiments
• 6. Conclusion
• References