Research On Data Collection Schemes In Underwater Acoustic Sensor Networks

In recent years,underwater acoustic sensor networks have received extensive research in various scenarios.Underwater acoustic sensor network is a network that combines micro mechanical sensor technology and wireless technology,with intelligent sensing,computing,and communication.One of its core functions is to collect data for related operations.However,underwater environments such as large lakes and oceans are complex,and various influencing factors pose challenges to data collection.For example,the presence of an ocean thermocline affects the communication range of underwater sensor nodes in an underwater acoustic sensor network.Underwater acoustic communication transmits data through acoustic waves,which consumes high energy and has limited transmission distance.Generally,underwater nodes extend the transmission distance through multiple-hop,but this will increase the energy consumption of the nodes.Most nodes in underwater acoustic sensor networks are deployed underwater,and it is also difficult for each node to supplement energy.With the development of autonomous underwater vehicles and underwater glider technology,many research methods use them for data collection.However,although autonomous underwater vehicles can assist in data collection and reduce energy consumption of underwater nodes,their endurance time is relatively short.In large-scale network data collection,the lifetime of underwater acoustic sensor networks is relatively short.Many methods for underwater gliders to assist in data collection also do not consider issues such as the glider’s communication range and energy consumption.In addition,underwater image data is an important object in data collection.Existing methods for collecting data using autonomous underwater vehicles or underwater gliders rarely consider the timeliness of data(including image data)from generation to collection,i.e.,data collection delay.In view of the above problems,this paper proposes a layered data collection scheme for underwater acoustic sensor networks.The specific contents are as follows:(1)This paper proposes two image compression algorithms based on VQ-VAE-2(Vector Quantized Variable Auto Encoder 2),NVQ(No top decoder Vector Quantized)and VQI(Vector Quantized Index).The original image with large amount of information is encoded,compressed,transmitted,reconstructed and generated by convolution network.The NVQ model reduces the characteristics of the bottom quantization,and then merges with the top quantization result for quantization.VQI re-quantizes the subscript of the bottom quantization result,greatly reducing the amount of data.NVQ and VQI inherit the advantages of VQ-VAE-2.Pictures have good fidelity.NVQ and VQI greatly reduce the amount of data transmitted,improve the transmission speed of picture data,reduce transmission energy consumption,and reduce data collection delay.(2)This paper proposes a hierarchical data collection architecture based on the CCPPOA(Complete Coverage Path-planning Obstacle avoidance)algorithm,and uses image compression algorithms to compress data.In this scheme,the sea area is divided into upper layer,thermocline and lower layer.The communication range of underwater glider in different layers of the sea area is different.The glider’s glide angle and initial navigation point are determined by the communication range.The underwater glider collects data in the thermocline by constructing a full coverage path through the CCPPOA algorithm.The underwater glider mainly moves by adjusting its own net buoyancy,and the energy consumption is small.In the thermocline,the sensor node only needs to transmit data to the underwater glider in a single hop,instead of using multi-hop,which helps to reduce the energy consumption of the node.In the upper and lower layers,the nodes transmit data to the underwater glider through the sleep scheduling multi-hop mechanism,and the underwater glider performs only one ascent and descent in the upper and lower layers,which greatly reduces the glider’s glide path length and reduces the data collection delay.