Research Of FFE-DEF Adaptive Equalization Technology In Long-line Transmission System

In the process of communication of the long-line transmission system, especially the long-line transmission without the repeaters, the base-band signal will be distortion and attenuation due to the long distance and high speed transmission. Because in this process, the transmission signal can be affected by many kinds of noise, and the common and serious noise is the inter-symbol interference(ISI).Besides that, the channel bandwidth is limited, so the high frequency part of the signal is badly weakened. Finally, all these caused the receiver cannot obtain the effective signal, the quality of transmission is affected and the transmission speed is limited. The adaptive equalizer is the most effective technique to eliminate the interference, correct and compensate the signal. The characteristic of the long-line transmission is equivalent to a low-pass filter. Its transfer function is a downward trend parabola, while the adaptive equalizer can get a transfer function of the rise parabola, which is contrary to the function of the long-line transmission. Thus, the adaptive equalizer can compensate and correct the signal at the receiving end. This paper is based on the research background of the system of deep well observation and the signal acquisition with thousands of meters cable. Aiming at the transmission problems in the long distance transmission with the coaxial, this article proposes a new adaptive equalizer with an improved variable step size algorithm.There are two kinds of commonly used equalizer, linear equalizer and nonlinear equalizer. The typical linear equalizer is forward feedback equalizer(FFE), its structure is simple, and it can deal with no severe inter-symbol interference. But its elimination ability is limited to serious symbol noise, and if increasing the number of the tap, the complexity is increased. Another commonly used equalizer is decision feedback equalizer(DFE), it is nonlinear equalizer, can deal with more serious inter-symbol interference. But it easy to cause the error diffusion and can not cancel the precursor ISI well.In this paper, taking the advantages and disadvantages of the above two kinds of equalizer into consideration, an improved adaptive equalizer is presented. Firstly, the structure of equalizer is improved. The new equalizer is a combination of 4-tap feed-forward equalizer and 5-tap decision-feedback equalizer.Two equalizers supplement each other in the operation, the structure is optimized, and the equalization effect is better. Then, this paper also proposes an improved variable step size algorithm based on arc-tan function. This algorithm can solve the problem of thecontradiction between the convergence speed and steady-state error more effective than the traditional least mean square algorithm(LMS) with fixed step size. The algorithm can rapidly converge at the initial stage, and at the steady state it can maintain a small step size to obtain smaller steady-state error. Compared with many other variable step size algorithms, the new one is more superiority and stability.This design is low complexity and does not need complex circuit, and the improved variable step size algorithm calculates quickly, and it has fast convergence and a good ability of channel tracking. The algorithm not only can accelerate the speed of data processing, but also can deal with the real-time change of the communication system. The experiments show that, the adaptive FFE-DFE equalization with the new variable step size algorithm can improve the performance of equalization and speed up the convergence. The inter-symbol interference and noise of the signal are also eliminated. The data transmission distance is greatly extended, and the data rate is improved too.