Research On Intelligent Optimization Design Technology Of SPMA Protocol

With the development of communication technology,modern warfare has gradually become a joint information warfare by sea,land and air.Information interaction in combat has also expanded from simple instructions to multimedia data such as text,reconnaissance images and video,and the delay and accuracy of information transmission have put forward higher requirements.The emergence of data link technology has greatly improved the coordination between various military equipment and personnel.Aircraft,ships,satellites,and the military can be combined into a whole to enhance combat capabilities.Therefore,data link technology plays a decisive role in modern warfare,and the Medium Access Control(MAC)layer protocol that manages and allocates channel resources is one of the key technologies.Tactical Targeting Network Technology(TTNT)is the latest generation of tactical data link technology.Its MAC layer uses Statistical Priority-based Multiple Access(SPMA)protocol to build an on-demand ad hoc tactical network.The SPMA protocol supports multi-priority,which can limit and delay the transmission of low-priority packets when the traffic load is heavy,and provide high transmission success rate guarantee and great delay performance for high-priority packets.However,the existing parameter setting schemes and backoff strategies of the protocol are no longer optimal in a network environment where the topology and transmission rate change dynamically,which limits the performance of the network.This paper studies the intelligent optimization technology of the SPMA protocol,and designs an optimization algorithm based on queuing theory and deep reinforcement learning to improve the network’s performance such as delay,throughput and transmission success rate.This article first introduces the characteristics,applications and main performance indicators of tactical wireless networks,analyzes the commonly used MAC protocols in tactical data links and their application scenarios,and introduces existing improvements to these protocols.After that,the SPMA protocol’s multi-priority access mechanism,channel state judgment,frequency-hopping and time-hopping mechanism and flow control mechanism are presented,and the problems and shortcomings of the SPMA protocol are analyzed.Then it describes the basic principles of reinforcement learning,the commonly used reinforcement learning and deep reinforcement learning algorithms in communication protocol design.The basic principles of queuing theory are also introduced.Next,this paper studies the priority-threshold optimization method of SPMA protocol,and proposes an adaptive priority-threshold setting strategy.Aiming at the characteristics of ad hoc network transmission rate adaptation and network topology dynamic change,combined with probability theory and queuing theory,the relationship between priority-threshold setting and the first time success rate of the highest priority packets is theoretically deduced.The priority-thresholds are adaptively modified when the rate changes and the number of nodes changes.Through simulation,the correctness of the strategy is verified,and the performance improvement of the SPMA network brought by the strategy is analyzed.When the network supports a higher transmission rate or the number of nodes is reduced,the transmission delay is reduced and throughput is increased;when the network only supports low-rate transmission or the number of nodes increases,the first time success rate of the highest priority packets is guaranteed.Finally,this paper studies the backoff strategy of the SPMA protocol,combined with the Deep Deterministic Policy Gradient(DDPG)algorithm to design an intelligent backoff strategy for the SPMA protocol.Taking the priority of the packet and the channel occupancy statistics in a certain time slot in the past as input,the DDPG algorithm is used to select the backoff time in the continuous action space.Simulation shows that this strategy can effectively reduce the delay of low-priority packets,improve network throughput and transmission success rate,and reduce backoff times.