Research On Key Technologies Of Reliable Data Collection In Underwater Wireless Sensor Network

Recently,Underwater Wireless Sensor Networks(UWSNs)have been gradually applied to the field of underwater data collection,and related research has received increasing attention.In UWSNs,reliable data collection is the key technology that determines whether the underwater application can function stably,mainly involving the research of reliable links and routing in underwater.Compared with terrestrial,the long propagation delay of acoustic signals and the time-varying network topology in UWSNs may cause information asymmetry between the sender and receiver of the link,thus increasing the probability of packet collision and triggering the problem of point-to-point transmission failure.At the same time,the above features can also make the maintenance of global routing information more difficult,while the transmission of relying on local routing information alone is prone to fall into routing voids,leading to end-to-end packet loss.To solve the above problems,there are currently two types of technical.Layered design improvement solutions design separate network protocols for each layer,which has the advantages of being easier to implement and highly scalable.However,the layered design makes it difficult to share information between layers,increasing information redundancy and reducing the utilisation of effective information in the network.Cross-layer design,instead,facilitates decision-making by using information exchange between different layers,such as link information helps to update and maintain local routing tables,and routing information can guide channel allocation and thus reduce channel congestion.Cross-layer design therefore allows for flexible use of each layer’s functionality and is more adaptable to the dynamically changing underwater environment.However,the cross-layer design affects the independence and robustness of each layer,and it is a challenge to use the inter-layer information effectively to improve the reliability of links and routes while ensuring the original functions of each layer.In addition,some work has been done to ensure link reliability of underwater data collection by deploying Autonomous Underwater Vehicles(AUVs)to measure the Value of Information(VoI)of the data in terms of timeliness and to maximise VoI as an optimization objective for AUV path planning.However,data integrity is also considered to be an important measure of VoI in some underwater applications:VoI of data from missed areas during the collection will increase over time.Considering only data timeliness while maximise Vol as an optimization objective will lead to a continuous accumulation of data that is difficult to be covered by AUVs under the speed limit,and eventually data that exceeds the node storage space will be discarded leading to a significant loss of VoI.Therefore,to enhance the reliability of AUV-aided data collection,the second challenge is how to plan the path of AUV reasonably to reduce the overall loss of data value while taking into account data timeliness and integrity.To address both of these challenges,this thesis provides an in-depth analysis of existing work and carries out three main tasks,the details of which and the main contributions of the research are as follows.(1)To address the problem of point-to-point transmission failure due to packet collision at the link layer in the hop-by-hop collection scenario of UWSNs,this thesis proposes a Link-Efficiency multi-channel Transmission Protocol(LETP)with simultaneous conflict avoidance,high bandwidth utilization and low signaling overhead.To achieve channel conflict avoidance,we propose Link-Efficiency Channel Allocation(LECA)by using cross-layer routing information at the link layer.The link information is also used to assist the routing layer in proposing a probabilistic forwarding node selection method with low signalling overhead to improve bandwidth utilisation.Experimental results show that LETP improves the packet delivery rate to over 97%and improves the channel utilisation by about 10%compared to the comparison algorithm.(2)To address the packet loss problem caused by routing voids at the network layer in the hop-by-hop collection scenario of UWSNs,this thesis proposes a Cross-layer Design based Transmission Protocol(CDTP).CDTP leverages link information to predict the movement of nodes to build a stable layered network topology,and proposes a dynamic routing strategy by analysing the competition between neighbours,effectively avoiding packets falling into a void region.To reduce end-to-end delays and signalling overheads,CDTP enables reliable routing and link selection between nodes through only one round of communication by sharing cross-layer information.Experimental results show that the packet delivery rate of CDTP is nearly 98%,and the average end-to-end delay is improved by about 20%compared to the comparison work.(3)To address the problem of Vol loss due to the difficulty of balancing data timeliness and integrity in UWSNs under AUV speed constraints,this thesis proposes an approximate algorithm for AUV path planning that maximises information value collection.To unify the impact of data timeliness and integrity Vol loss on collection reliability,the algorithm dynamically adjusts the priority of AUV access based on the accumulation of data at each node over time and the decay of timeliness Vol.The AUV path planning problem is then modelled as a variant of the Travelling Salesman Problem(TSP),and a novel polynomial-time path planning approximation algorithm is proposed to effectively reduce the overall Vol loss.Experimental results show that the data loss under our algorithm is about 40%of the comparison work and has a lower running time.