| A polling system,as a widely used anti-conflict protocol,can avoid data collision during transmission,eliminate energy waste to a great extent,and promote efficient information collection.Simultaneously,in the practical environment,the customer transmission failure caused by channel quality can’t be ignored.Therefore,to reduce the delay,improve the quality of service and conform to the actual channel application environment,this paper proposes a polling system based on multi-server and retransmission mechanisms.First,to solve the problem of limited load capacity of single-server polling mechanism and network congestion due to the rapid increase of network concurrency,a multi-server polling access control system is proposed.By analyzing the various states and working mechanisms of the system and establishing a theoretical model,the performance characteristics of the system in the synchronous mode are analyzed and verified by simulation.In addition,due to the complex and computationally intensive problems of traditional polling research methods,and XGBoost prediction model is also established to detect multiple servers in the system and learn the situation of each server to further predict the system performance,providing a new research method for polling systems.Secondly,to solve the problem of large concurrent users and low transmission efficiency in WSN,a blockchain-based synchronous control polling MAC(BSCP-MAC)protocol is proposed.Simultaneously,the model and its data distribution method are analyzed by providing services for customers within each block according to the synchronization method.By verifying the protocol in each of the three service strategies of Gated,Exhaustive,and Limited-1,the results demonstrate that BSCP-MAC is applicable in all three fundamental polling strategies.Comparing the three strategies shows that the Exhaustive service has the lowest customer latency and the fastest information processing efficiency.Thirdly,a reserved polling scheduling method with an error retransmission mechanism is proposed to adapt to the actual communication environment and to address the massive user demand at the same time.By analyzing the specific polling scheme and working mechanism of the system,the model is established by the mathematical analysis method,and the theoretical expressions of the average length and delay of customers are further derived.Then through simulation experiments,the correctness and reliability of the theoretical model are verified.Finally,to realize the application of the polling system in WSN,the nesC language is used to further implement an Exhaustive service polling system with a retransmission mechanism on CC2530 nodes by hierarchically dividing all sensor nodes by clustering based on Ting OS operating platform.In addition,a smart home fire alarm system is designed based on the implemented polling mechanism.The system can not only monitor the indoor environmental information in real-time through sensors but also add an alarm device to meet the actual application needs of users. |
