Optimal Design Of Dynamic Thread Pool Considering External Factors And Its Implementation

With the continuous development of electronic products, multi-threaded technology iswidely used in order to meet the complex task scheduling of software. Thread pooltechnology is used to manage multi-thread, and the research on thread pool includes thesize and design pattern of thread pool. Currently, software development engineers set thesize of thread pool through their development experience, and this setting is certainlyblindness. The main design patterns of thread pool mainly focus on the internal factors, lessabout the external factors.Aiming at the incompleteness of thread pool research, the requested object file ofthread is chosen for the external factors of thread pool in this thesis. The relationshipbetween the requested object file and the size of thread pool is transformed into therelationship between the requested object file and threaded timeout by threaded timeout,and the relationship between the requested object file and threaded timeout is positiveaccording mathematical modeling and solving methods, thus the relationship betweenthreaded timeout and the size of thread pool is reversed. With the larger target file ofthreads, these design methods using target file segmentation improve the operationalefficiency of the thread pool. The requested object file is used as the standard of thepriority of thread pool. This design pattern reduces the amount of waited threads, andsaves system resources.Multi-threaded download applications on the mobile terminals of IOS achieve theoptimum design methods in this thesis, and data obtained by the program verify thereversed relationship between the requested object file and the size of thread pool,moreover verify results provide a theoretical basis for the thread pool size settings. Thedesign patterns which target file optimizes the size of dynamic thread pool reduce theamount of CPU occupancy on the system in a short thread, which improve applicationintegrated operating performance. Software industry is a rapidly growing industry，andMulti-threading technology and the thread pool technology update constantly, thereforethread pool also requires in-depth study.