A Novel Method Of Bussgang-Type Blind Equalization In High-order QAM Systems

The blind equalization has been studied for more than30years. In this stage, various blind equalization algorithms have been proposed continuously. However, the development of modern communications technology and the concepts of multiple communications make the signal transmission rate increase greatly and the time-varying characteristics of the wireless transmission channel enhancement, which are bringing more stringent requirements for blind equalization algorithms. Diversity and denseness of the modulation constellation is becoming an inevitable trend. The dense QAM modulation is most likely to become the mainstream of modulation in the future, which makes it will be meaningful to research these blind equalization algorithms of dense QAM signals systems.This thesis proposes a new family of Bussgang blind equalization algorithm based on analyzing several types of traditional Bussgang algorithms. The proposed algorithms is named Expand Multi-Modulus algorithms (EMMA). Compared with traditional Bussgang blind equalization algorithms, EMMA has stronger ability to detect signals with high-dense square and non-square QAM. The main research work is as follows:(1) Through the analysis of the error function effects on convergence of algorithms, this paper presents a calculation method of constellation characteristic constant R.(2) In order to further improve the performance of EMMA algorithm, through combination of the new cost function and Constellation-Matched Error (CME) functions, this paper puts forward instantaneous dual-mode switched hybrid algorithm.(3) For the application, with the output signals of the early equalizer, this paper also proposes an approximate calculation method of constellation characteristic constant R. This computational method does not rely on any a priori knowledge of constellations.Finally, simulation results show the feasibility and correctness of these algorithms.