Progressive Image Transmission Of Underwater Acoustic Network Based On Cross-layer Design And Unequal Error Protection

The underwater acoustic network is an important technical means for the exploration and development of current marine resources.In some important application scenarios such as underwater ecological monitoring,marine resource exploitation,underwater military monitoring and natural disaster warning,underwater acoustic network communication technology is of great significance.The underwater acoustic network uses sound waves as the communication carrier,which is very different from wireless communication on land.Underwater network communication has the following challenges: underwater acoustic channel has shortcomings such as large delay and dynamic change of delay;in addition,the available bandwidth of underwater nodes is small and the limited energy supply of the battery limits the working time and communication coverage of the node;Network protocol design is difficult,and it is necessary to redesign the protocol or modify the network protocol on land according to the characteristics of underwater acoustic communication.The research on underwater image and video transmission mainly includes image compression processing technology and underwater acoustic channel transmission anti-interference technology.The former focuses on the optimization of image processing algorithms,and reduces the amount of data to be transmitted when image quality is guaranteed.The latter focuses on the reliability of data transmission in underwater networks,improving the delivery rate of data transmission in the network,and reducing latency.In this paper,by analyzing the problems of image transmission in underwater acoustic network,a progressive image transmission method based on cross-layer design and unequal error protection is proposed.The scheme uses SPIHT algorithm to perform progressive compression processing on images,and compressed output bit streams with different degrees of importance are encoded using fountain codes with unequal error protection property.The application layer marks the parameter data packets for decoding as critical data packets,and enforces protection at the MAC layer using hop-by-hop acknowledgment and retransmission policies.Adopting a transmission strategy with no acknowledgement and no retransmission for a large number of ordinary packets with lower importance levels,reducing the collision between packets and reducing the load of the underwater acoustic network.In this paper,there are two ideas of unequal error protection.On one hand we use fountain code with unequal protection performance in the application layer to encode image compression data,on the other hand crosslayer interaction between application layer and MAC layer is adopted to pass the different important levels of packets,so packets take different transport protection strategies.In this paper,the underwater network simulation platform based on hierarchical state machine is used to simulate the network transmission,and the network performance such as network throughput and packet delivery rate and the final image receiving quality are analyzed.The simulation is compared with the image transmission scheme of hop-by-hop confirmation retransmission.The simulation experiments show that in the multi-hop water network,the scheme shows better performance in receiving image quality,data packet delivery rate and transmission delay.