Research On Aggregation Traffic Intelligent Control Method Of Heterogeneous Links In Command And Control Network

The iteration of operational concepts,the innovation of operational styles,and the emergence of operational forces have led to the rapid transformation of warfare from networkcentric warfare to intelligent warfare.Intelligent warfare,characterized by joint all-domain operations,requires a new generation of command and control networks with wide-area coverage and broadband interconnection as the core,extending to on-demand access and intelligent ubiquitous sky-ground integration access networks.Through efficient aggregation of heterogeneous links such as wired,optical fiber,shortwave,ultrashortwave,microwave,regional broadband,and satellite,it tightly connects resources from land,sea,air,space,and cyberspace domains to support cross-domain adaptive combinations of operational forces in perception,decision-making,engagement,and support,realizing "weapon system decoupling and dynamic element recombination" and achieving the capability goal of "any sensor to the optimal shooter." As a result,the quantity,heterogeneity,and complexity of battlefield elements will rapidly increase,and the decision-making space will increase exponentially,posing new challenges to the real-time and accuracy requirements of command and control.This paper focuses on the core issue of ’ how to provide intelligent and efficient network service quality guarantee for the construction of optimal kill chain driven by different combat missions under the condition of decoupling of weapon system and dynamic reorganization of elements ’,and explores the intelligent control method of heterogeneous link aggregation traffic in command and control network.The specific research contents and innovations are summarized as follows :(1)After decoupling the weapon systems and dynamically reorganizing the elements,combat units such as sensor nodes,command and control nodes,and firepower nodes in various operational domains achieve link aggregation and service interconnection through more than ten heterogeneous links,including wired,fiber optic,shortwave,ultrashort wave,microwave,local area network,and satellite,within the command and control network.The intense bursts and high concurrency of operational tasks,as well as the high dynamics and weak connectivity of the network topology,will result in even more instantaneous surges of aggregated traffic over heterogeneous links.This traffic exhibits non-stationary statistical characteristics,making it difficult to predict accurately.Therefore,this paper proposes an intelligent prediction method for aggregated traffic over heterogeneous links in command and control network,which integrates spatiotemporal features.This method models the high-order characteristics of network traffic from two dimensions: temporal correlation and spatial correlation.Temporal correlation is used to characterize the inherent correlation in the traffic sequence of target nodes,while spatial correlation is used to describe the external influence of neighboring nodes’ traffic sequence on the target node’s traffic sequence.Based on this,a network traffic intelligent prediction model is constructed by utilizing recurrent neural networks to capture the temporal features of network traffic,graph convolutional neural networks to capture the spatial features of network traffic,and attention mechanisms to enhance the model’s generalization ability in a dynamically changing environment.Experimental results demonstrate that compared to models such as HA,ARIMA,SVR,GRU,and T-GCN,the proposed intelligent prediction method for aggregated traffic over heterogeneous links in command and control networks significantly reduces prediction errors and improves prediction accuracy,whether for single-step or multistep predictions.This method establishes a solid foundation for the timely control of aggregated traffic over heterogeneous links in command and control networks and the pre-allocation of cache resources in aggregation nodes.(2)After decoupling the weapon systems and dynamically reorganizing the elements,more than ten types of heterogeneous links including wired,fiber optic,shortwave,ultrashort wave,microwave,area bandwidth network,and satellite are converged onto the network control equipment,connecting multiple domains of the battlefield space.The transmission rates of the services range from Kbps to Gbps,while the timeliness of the services varies from milliseconds to minutes.The fluctuations in the arrival rates of the services and the diversity of service quality requirements make the aggregation of traffic over heterogeneous links more complex and difficult to finely control.To address this,this paper proposes an adaptive control method for the aggregation of traffic over heterogeneous links in command and control networks,aiming at differentiated services.This method combines the long-term and instantaneous states of the queues to construct a queue state determination model,providing criteria for determining five queue states: burst tolerance,burst suppression,decay tolerance,decay suppression,and queue stability.Based on this,a queue management algorithm that incorporates the queue state is designed,and together with a hierarchical and graded dynamic scheduling mechanism,the adaptive control of traffic aggregation is achieved.Simulation results demonstrate that compared to traffic control methods based on algorithms such as ARED,URED,TRED,and MDPRED,the proposed adaptive control method for the aggregation of traffic over heterogeneous links in command and control networks shows significant advantages in terms of queue length stability,queuing delay,delay jitter,packet loss rate,and throughput.It provides technical support for ensuring reliable service quality for delay-sensitive,bandwidth-sensitive,and best-effort delivery of heterogeneous traffic.(3)After decoupling the weapon systems and dynamically reorganizing the elements,heterogeneous services such as voice,data,images,and videos in the multi-domain battlefield space are aggregated and propagated through the traffic control of heterogeneous links in the command and control network.The process involves complex and instantaneous surges of traffic arrivals and hierarchical,graded,and dynamically scheduled node services.As a result,the queue lengths in the aggregation nodes of the heterogeneous links become more closely coupled,and the allocation of cache resources becomes challenging.To address this issue,this paper proposes a cache resource management method for aggregation nodes of heterogeneous links in command and control networks under Quality of Service(Qo S)constraints.This method characterizes the arrival process of aggregated traffic over heterogeneous links in command and control networks using traffic models.It introduces the theory of large deviations to equivalently represent the hierarchically coupled traffic control system as multiple independent single-server single-queue subsystems.Based on this,the paper derives the equivalent relationships between queue length distribution,packet dropping probability,and cache size under Gaussian input conditions,and establishes a model for computing the cache boundary of aggregated traffic.Simulation and numerical analysis results demonstrate that the proposed cache resource management method for aggregation nodes of heterogeneous links in command and control networks achieves reliable theoretical results in three typical application scenarios.This method provides theoretical support for the precise allocation of cache resources for delay-sensitive,bandwidth-sensitive,and best-effort delivery of heterogeneous traffic.(4)After decoupling the weapon systems and dynamically reorganizing the elements,the command and control network constructs different end-to-end transmission paths by flexibly combining and reconfiguring the dispersed,heterogeneous,and cross-domain combat forces to meet the end-to-end performance requirements of different combat missions.In a highly adversarial and dynamic battlefield environment,the time-varying nature of link states due to factors such as electromagnetic interference and multipath propagation introduces complexity and variability in the end-to-end delay of aggregated traffic over heterogeneous links,making it difficult to quantitatively analyze.To address this issue,this paper proposes an end-to-end delay analysis method for the aggregation of traffic over heterogeneous links in command and control networks based on network calculus.This method utilizes stochastic arrival curves to describe the constrained boundaries of the complex and instantaneous surges of aggregated traffic arrival processes.It also uses stochastic service curves to describe the constrained boundaries of the service capabilities of aggregation nodes in a time-varying link state environment.Based on this,the paper considers two scenarios: with and without traffic control.It establishes a relationship model among end-to-end delay,stochastic arrival curves,and stochastic service curves,enabling quantitative analysis of end-to-end delay for delay-sensitive,bandwidth-sensitive,and best-effort delivery of heterogeneous traffic.Numerical verification results demonstrate that the end-to-end delay of aggregated traffic is positively correlated with the number of concatenated nodes and the degree of traffic bursts,and negatively correlated with the violation probability.Moreover,the traffic control strategy designed in this paper exhibits superior performance in end-to-end delay guarantee,laying a solid theoretical and technical foundation for the construction of optimal kill chain from sensor to shooter.