Power Allocation for Non-Coherent Multi-Carrier FSK Underwater Acoustic Communication Systems with Uneven Transmission Source Level

Non-coherent multi-carrier Frequency Shift Keying (FSK) is widely used in underwater acoustic communication systems due to its high reliability, low receiving complexity, and less susceptibility to channel conditions. In practical underwater acoustic communication systems, due to the wide band nature of the signal, the overall transmitting source level is usually not flat. Targeting at this issue, this paper proposes power allocation methods based on genetic algorithm and adaptive greedy algorithm, which maximize channel capacity and system robustness. Results based on simulated and pool experimental channels show that our methods can achieve higher channel capacity and lower bit error rate than power averaging scheme, therefore better system performance can be obtained.