Channel Equalizer Design for Continuous Wave Mud Pulse Based on OS-ELM

Mud pulse transmission technology is one of the core technologies for measurement while drilling. In the process of continuous wave mud pulse signal up to the ground, there will be large amplitude attenuation and inter-code interference caused by the multipath effect. Therefore, an adaptive decision feedback equalizer based on online sequential extreme learning machine (OS-ELM) is proposed in this paper, where the OS-ELM is used as the feedforward filter of the equalizer and the feedback filter is based on the least mean square algorithm for decision feedback. First, a data set is built using a simulated mud pulse equivalent baseband signal, which includes a useful distortion-free signal and a corresponding distorted signal. Then the designed equalizer is initialized and updated online using the created dataset to obtain equalizer parameters that have good distortion correction capability. Finally, tests are performed on the validation set samples with different signal-to-noise ratio (SNR). The test results show that the bit error rate (BER) of the mud pulse signal processed by the designed equalizer is greatly reduced compared to that of the conventional equalizer. It can be seen that the designed equalizer compensates and corrects the signal attenuation and inter-code interference significantly, which verifies the effectiveness and feasibility of the method in this paper.