A deadlock resolution strategy based on spiking neural P systems
Journal of Ambient Intelligence and Humanized Computing(2019)
摘要
Deadlock resolution is a classic problem faced by operating and database systems. It not only affects the utilization of system resources, but also may bring unpredictable consequences to the events corresponding to deadlock processes. Spiking neural P systems (SN P systems) are neural networks built upon membrane computing models inspired by neuron communication via excitatory and inhibitory spikes. We explore the application of SN P systems to the optimal revocation of deadlock processes in order to overcome deficiencies in traditional deadlock resolution methods such as exhaustive method and time cost method. Our system determines the optimal order of revocation by calculating the time cost associated with the deadlock release process. Compared to the exhaustive method and time cost method, our system significantly reduces the time complexity of the deadlock revocation process, improving deadlock resolution efficiency and offering more opportunities for parallelism.
更多查看译文
关键词
Membrane computing,Spiking neural P systems,Deadlock,Time complexity
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要