AN ADAPTIVE FILTERING ALGORITHM TO COMPENSATE FOR FREQUENCY-DEPENDENT IMAGE INTERFERENCE IN PRACTICAL WIRELESS RECEIVERS

    Authors

    Y. Liu;W. B. Mikhael

    Comments

    Authors: contact us about adding a copy of your work at STARS@ucf.edu

    Abbreviated Journal Title

    J. Circuits Syst. Comput.

    Keywords

    Frequency-dependent image interference; adaptive filter; non-data-aided; complex Taylor series expansion; I/Q IMBALANCE COMPENSATION; DIRECT-CONVERSION RECEIVERS; LOW-IF; RECEIVERS; MISMATCH COMPENSATION; CIRCULARITY; Computer Science, Hardware & Architecture; Engineering, Electrical &; Electronic

    Abstract

    Frequency-dependent image interference is an inevitable impairment in wideband quadrature receivers. To suppress this interference, this paper presents a non-data-aided adaptive compensation algorithm, optimal block adaptive filtering algorithm based on circularity (OBA-C). This technique exploits the concept that the image interference in wireless systems causes the received complex signal to lose its nature of circularity. Then the OBA-C algorithm restores the circularity of the signal to compensate for the image interference. To avoid manually selecting a step size, the presented algorithm employs the complex Taylor series expansion to optimally update the adaptive filter coefficients. This technique fully exploits the degrees of freedom of the system, and generates an individual update for each filter coefficient at each iteration. Computer simulations are carried out to test the performance of the OBA-C for practical levels of image interference. The simulation results illustrate that the OBA-C achieves fast convergence and excellent image rejection performance. Other advantages of OBA-C are also analyzed, including the robustness against radio frequency impairments and different levels of image interference.

    Journal Title

    Journal of Circuits Systems and Computers

    Volume

    22

    Issue/Number

    5

    Publication Date

    1-1-2013

    Document Type

    Article

    Language

    English

    First Page

    25

    WOS Identifier

    WOS:000318791500002

    ISSN

    0218-1266

    Share

    COinS