The Least Mean Squares Blind Adaptive Multi-user Detection algorithm, the Recursive Least Squares Blind Adaptive Multi-user Detection algorithm, and the Kalman Filter Blind Adaptive Multi-user Detection algorithm, for under-ice communication networks with rising and diving interfering users, are investigated in this paper. Under-ice random asynchronous multi-user communication experiments with rising and diving interfering users were carried out (7 users, 15.7 bit/s per user, 1.2 km under-ice acoustic channels in the Bohai Sea). Good experimental results verified the feasibility and effectiveness of the proposed algorithms in mitigating inter-symbol interference and multiple access interference.