Study On Resource Allocation Strategy Of Underwater Acoustic OFDM Communication Nrtwork

In the 21 st century,the land resources of absence can not meet growing world economy,and each country strengthens the exploitation of marine resources.The research on underwater communication technology is a prerequisite for exploitation of marine resources.In the underwater network,electromagnetic wave propagation is severely attenuation,which is not suitable for underwater communication.Therefore,the sound wave is the only form of energy known to mankind so far to achieve underwater long-range transmission,and it also becomes the ideal and unique communication mode to realize the medium and long distance information transmission between nodes.Underwater acoustic communication networks face larger connatural challenges than radio communication networks such as narrow bandwidth,limited energy,and so on.Reasonably allocating communication resources to decrease the outage probability of the underwater communication networks is the main issue of this study.Firstly,this study briefly introduces the anti-interference role of OFDM technology in the underwater acoustic communication network.Various underwater communication nodes in the ocean are available.Nodes of different types have different communication characteristics and requirements.Therefore,the OFDM underwater acoustic heterogeneous network is established.Secondly,in the case of uncertainties,a robust joint channel and power allocation alternate iteration algorithm is proposed to solve a joint resource allocation(channel selection and power allocation)problem with multiple channels,users,and constraints in the underwater acoustic communication networks.The proposed algorithm employs robust theory to optimize the channel uncertainty and cope with the influence of channel uncertainty on network performance.The residual energy of nodes is considered in the proposed algorithm to balance their energy,thereby extending the working time of the entire network.The proposed algorithm considers the interference factor and can give priority to the appropriate channel for the node with large network influence to reduce thefrequent changes of the nodes,reduce the network interference,and improve the search efficiency.Simulation results show that the proposed algorithm can reduce the outage probability of network,extend the working time and improve the communication performance of the whole network.Finally,in the case that the channel gain is hard to obtain,a fully distributed adversarial multiplayer multi-armed MAB game is formulated in the OFDM underwater acoustic network.The proposed the G-EEL algorithm without any prior channel information and direct information exchange among nodes is proposed to cope with the game mode and improve the learning ability of the algorithm.In the improved algorithm,the proposed strategy updating machine can enlarge the set of the learning strategies.And the proposed algorithm can search the additional strategies other than the original fixed strategies.Essentially,the proposed learning algorithm overcomes the disadvantage that the traditional MAB method can only get the sub-optimal solution in the fixed finite set of strategies.The proposed strategy updating mechanism makes the learning algorithm more complex and the learning time increase.In addition,the dynamic learning mechanism is proposed to adjust the exploration and exploitation factor of the learning algorithm according to the characteristics of different learning stages.It can further improve the learning efficiency and accelerate the algorithm convergence.