Design And Implementation Of Costas Arrays Search System Based On Distributed Architecture

Costas arrays has ideal autocorrelation and is widely used in telemetry,wireless communication and cryptography,etc.However,there have not been more in-depth research results on the existence of Costas arrays of some special orders.In order to study the high-order Costas arrays,the paper studies the characteristics and construction methods of the Costas arrays,discusses the difference between the algebraic construction method and the exhaustive method,and focuses on the efficient implementation of the exhaustive method.The problem of search efficiency of first-order sequence puts forward the idea of using distributed computing to construct the search system of Costas arrays.Aiming at the realization of the distributed search system,the research content of this thesis is mainly divided into the following two parts,and the experimental data proves that the search system designed in the paper can significantly improve the search speed of Costas arrays.The first part optimizes the single-machine search efficiency of Costas arrays from the definition of Costas arrays,and proposes a divide-and-conquer recursive multi-threaded scheduling algorithm based on task stealing mechanism according to the characteristics of Costas arrays search task.For the traversal search in a multi-threaded environment,due to the uneven distribution of Costas arrays,the traditional multi-threaded search method can only complete all calculation tasks by a fixed thread after specifying the task,and the calculation amount of different tasks is different,resulting in a waste of thread resources.When the amount of thread tasks is large,the tasks are cut through divide and conquer recursion,and the task stealing mechanism is used between the threads to balance the thread load in the thread pool,so as to maximize the use of thread resources.The second part proposes to implement Costas arrays search system based on distributed architecture to solve the bottleneck problem of stand-alone performance.Because the high-order Costas arrays traversal search requires a large amount of calculation,the single-machine search for high-order sequences consumes a long time.Therefore,the research uses a distributed architecture to decompose high-level search tasks,and uses network middleware,such as databases,caches,message queues,etc.,to reliably distribute and aggregate computing tasks,and multi-machine collaboration to achieve efficient search.The structure of the search system adopts a management end/consumer end model,the management end is responsible for task decomposition and aggregation,and the consumer end is responsible for executing tasks and returning calculation results.The management end and the consumer end are highly decoupled.Under the premise of sufficient performance of the management end,the number of consumer ends can be dynamically increased or decreased through reasonable parameter configuration to achieve dynamic expansion and contraction of the system’s computing power.The management side and the consumer side ensure the reliable transmission of task information through middleware,and all three support cross-platform and cluster deployment,and finally realize an efficient high-end Costas arrays search system.