Hf Channel Blind Decision Feedback Equalizer Technology Research

This paper studies the blind decision feedback equalization which is a very important technique for shortwave communication and signal analysis, especially for 8PSK signal receival and for military signal reconnaissance and interception.A variety of algorithms of blind decision feedback equalization in this field are analysed, and performance comparison among them is also provided. However, due to the disadvantages, such as higher computational complexity, low convergence rate and stability, they may not be suitable for shortwave application.Based on the above discussion and background, several improved blind decision feedback equalization algorithms of DD-LMS type are proposed, including step-size-variable normalized DD-LMS (DD-VSSNLMS) algorithm, signed DD-LMS (such as DD-SELMS, DD-SSLMS and DD-SENLMS) and leaky DD-LMS (DD-LLMS) algorithm, etc.. These DD-LMS type algorithms preserve the same merits in computational complexity and implementation as that of LMS algorithm, however possess improved performances in convergence rate, residual error, stability and tracking ability. Simulation results prove the feasibility of above proposed algorithms; therefore they are suitable for shortwave channels.In this paper, the blind decision feedback equalization algorithms based on Kalman filtering are also discussed. Simulation results show that a modified square root Kalman algorithm could be used for blind equalization of shortwave channels. Finally, a simulation platform design for the blind decision feedback equalization algorithms is provided.