Intelligent Networking Method For Ship Ad Hoc Networks

Network communication has great application prospects in the fields of marine environmental observation and protection,underwater resource exploration,marine disaster detection,maritime operation safety,and marine military support activities.Among them,the establishment of stable,reliable and cost-effective communication networks in the ocean is of great interest to academia and industry.At present,the common communication methods at sea include shore-based station communication,satellite and maritime radio communication;however,they do not satisfy the growing demand for cost-effective communication.Ship ad hoc networks(SANETs),which are built based on ships with wireless communication capabilities,has attracted attention because of its low deployment cost,reliability and flexibility.In a ship ad hoc network,communication is established without relying on fixed infrastructure and is mainly formed automatically with the help of mobile network nodes.However,this network has typical features such as heterogeneous and unstable communication nodes,ultra-long communication distance,and bottleneck effect of communication links,which makes the communication intermittent and causes low transmission efficiency;in addition,there may be multiple network scenarios in the network,making the existing routing protocols designed for specific scenarios cannot be directly applied.On the other hand,when the node distribution is too sparse to guarantee real-time availability between nodes,the connected mobile wireless ad hoc networks(MANETs)become Opportunistic networks(Opp Nets)with indirect connectivity.Opportunistic network causes redundant replicated packets due to broadcast mechanism,which may lead to network congestion and degrade network performance if not optimally controlled.To address the above issues,this paper concentrates on the critical technologies of ship ad hoc networks networking,and the research mainly involves the following four aspects:1)The link interruptions of the communication networks are mainly due to links instability.Therefore,the prediction of the link state can try to avoid the use of unstable links and sudden interruption of links to gain preparation time.Thus,a link availability prediction based on machine learning algorithm for opportunistic network in oceans is proposed.Its innovation lies in the adaptive prediction of dynamic changes in the time and space domain predictions of the link.The algorithm establishes a channel model suitable for maritime communication by addressing the complex,time-varying and unstable features of the maritime channel.Then,a intelligent learning algorithm based on the fusion of feature extractor and sequence learning predictor is developed to accurately capture the sequential pattern information of effective features,and the prediction results are controlled in [0,1]Interval.The results show that,the proposed prediction scheme improves the prediction precision and communication throughput of link status evaluation for opportunistic networks compared to the other classical algorithms.2)To address the problem that the existing routing schemes of Opp Nets mainly focus on node mobility and local relay optimization without considering the impact of node behavior on encounter opportunities and relay node forwarding capability,a novel routing protocol based on node attributes for opportunity network is proposed,which mainly includes the forward-looking prediction of relay nodes and the assessment of the ability to reach other nodes;a protocol-based combination of node delivery capability and forwarding utility objective function based on the protocol combined with the delivery capability and willingness of nodes and used for message forwarding decisions.Among them,delivery capability is constructed for predicting the reachability of nodes to reduce packet retransmission and save resources;forwarding willingness mechanism based on device capacity and node movement behavior is used to evaluate the possibility of nodes to participate in relay forwarding.The experiments show that the proposed protocol promotes the network performance in respect of delivery ratio,average delay and overhead rate compared to other protocols.3)Due to the mobility of marine communication nodes,SANETs may consist of both MANETs and Opp Nets,and these two types of networks distributed in different regions may change alternately,and it is difficult to predict when the network will switch from MANETs to Opp Nets and vice versa.In this case,traditional routing protocols designed for specific networks alone are not applicable to SANETs in complex network scenarios.therefore,this paper proposes an adaptive policy switching scheme to adapt to this changing network environment,which mainly includes real-time network selection with predictability and adaptive routing,switching the corresponding routing policies according to different network types.The results show that,the proposed scheme increases throughput by 28.96%,and reduces the packet loss rate by 19.75% with an additional latency of 4.82%,compared to other classical routing algorithms.4)For the opportunity network in SANETs,the routing scheme based on broadcast nature leads to unnecessary resource waste and low packet delivery rate,so an efficient prediction-based opportunity routing scheme is proposed,which mainly includes the efficient selection of relay nodes with predictability.The scheme first identifies an efficient candidate intermediate region to recognize the unavailability of some apparently qualified intermediate nodes;and then takes into account the packet reception ratio between nodes and relay advancement prediction,to improve packet delivery.The proposed scheme achieves performance improvements in SANETs in terms of packet loss ratio and throughput with a tolerable latency increase,compared to other schemes.The approaches presented in this study combine ship navigation characteristics,ocean channel structure,and artificial intelligence,which is practical and popularized for dynamic reliability assessment,fault diagnosis,and routing decision-making in ship communication.