Abstract:
PURPOSE: An efficient energy scheduling method of a single core processor for a real-time task with the amount of changeable calculation is provided to express the amount of uncertain calculation of real-time tasks by using a probability calculation amount model, thereby minimizing the sum of the amount of energy consumption of the real-time tasks. CONSTITUTION: A probability calculation amount model for predicting the amount of calculation which is performed before a deadline of real-time tasks is determined(S100). Execution speed used for estimating the amount of energy consumption of the probability calculation amount model is determined(S200). Conversion time of the execution speed for the deadline is determined(S300). Assignment time of the real-time tasks is determined in order that the sum of the amount of energy consumption of the real-time tasks is minimal based on the conversion time(S400). [Reference numerals] (S100) Determine a probability calculation amount model; (S200) Determine execution speeds to be used for each task; (S300) Determine the conversion time of the execution speeds determined for predetermined deadlines of each task; (S400) Determine assignment times of each task such that the sum of the energy consumption of each task becomes minimum
Abstract:
본 발명은 확률적 스케줄링 방법 및 이를 이용하는 저전력 멀티 코어 프로세서를 제공한다. 확률적 스케줄링 방법은 복수의 프로세서 코어들이 주어진 작업의 정확한 계산량을 실행 완료 이전에 알 수 없는 작업을 실행할 때, 불확실한 계산량을 확률적 계산량 모델로 표현하고 이 확률적 계산량 모델을 기준으로 하여 에너지 소모량의 확률적 기대값을 최소화하는 실행 속도의 변환 시점들을 결정하고, 각 코어 개수에 대하여 최소 에너지 소모량에 대한 기대값을 계산하고, 계산 결과에 기초하여 확률적 최소 기대값을 가지는 최적의 코어 개수를 선택하는 것을 특징으로 한다.
Abstract:
본 발명은 변동하는 계산량을 갖는 복수개의 주기적 실시간 작업을 멀티 코어 프로세서에서 수행할 때 누수 전력 소모를 고려한 확률적 최소 전력 스케줄링 방법에 관한 것으로서, 변동하는 계산량을 갖는 복수개의 주기적 실시간 작업을 위한 멀티 코어 프로세서의 누수 전력 소모를 고려한 확률적 최소 전력 스케줄링 방법에 있어서, 멀티 코어 프로세서의 전체 코어들 중에서 M개의 주기적 실시간 작업을 수행하는데 사용될 α개의 코어들을 결정하는 제1 단계; 상기 결정된 α개의 코어들을 제외한 나머지 코어들의 전원을 소등하는 제2 단계; 상기 결정된 α개의 코어들에게 M개의 주기적 실시간 작업들을 할당하는 제3 단계; 및 상기 α개의 각 코어에서 상기 M개의 주기적 실시간 작업들을 각각 수행할 때 전력 소모량 확률적 기대값을 최소화하는 실행 속도를 결정하는 제4 단계를 포함하는 스케줄링 방법을 제공한다. 본 발명에 의하면, 멀티 코어 프로세서상에서 변동하는 불확실한 계산량을 갖는 다수의 실시간 작업들의 수행할 때 누수 전력 소모를 고려하여 확률적 모델에 기초하여 전력 소모를 최소화할 수 있는 스케줄링 방법을 제공할 수 있는 효과가 있다.
Abstract:
PURPOSE: A connection-assured communication of wireless sensors is provided to secure connection time and maximize the communication amount per hour. CONSTITUTION: A number of communication available packets of a start node searched edge is changed by the predetermined value(S140). A maximum algorithm is applied about residual graph(S150). All available routes between receiving nodes and source nodes are searched. The residual graph is created by removing the searched route. The transmitting node and an unused edge which connects the receiving nodes repeat the previous process until the unused edge does not exist. The total traffic paths between the transmitting nodes and the receiving nodes are calculated(S160).
Abstract:
PURPOSE: A method and an apparatus for scheduling operation modes based on a probabilistic execution time are provided to efficiently utilize limited energy by changing a probabilistic execution time model of a performance target work in real time through scheduling. CONSTITUTION: A probabilistic execution time model of a target performance target is determined(S100), and an operation mode to perform is determined for each time based on the probabilistic execution time model in order to maximize utility of the target performance task until the task is completed(S200). The determined operation mode is changed and performed for each time until the task is completed(S300). The probabilistic execution time model is determined by the probability which the performance target task is performed in a specific time.
Abstract:
The present invention relates to a stochastic minimum-power scheduling method considering leakage power consumption when a multi-core processor performs periodic real-time tasks having variable computation. The stochastic minimum-power scheduling method considering the leakage power consumption of a multi-core processor for periodic real-time tasks having variable computation includes determining α cores to be used for performing M periodic real-time tasks among all cores of the multi-core processor; turning off remaining cores other than the α cores determined; allocating the M periodic real-time tasks to the determined α cores; and determining an execution speed minimizing the stochastic expectation value of power consumption when each of the α cores performs each of the M periodic real-time tasks. According to the present invention, by considering leakage power consumption when performing the real-time tasks having variable, uncertain computation on the multi-core processor, there is an effect in that it is possible to provide a scheduling method that may minimize power consumption based on a stochastic model.
Abstract:
PURPOSE: A pruning saw using a pneumatic motor is provided to facilitate cutting branches positioned high and low by controlling the length of a handle part. CONSTITUTION: A pruning saw using a pneumatic motor(520) comprises a handle part, a pneumatic motor, and a chain saw. The compacted air is supplied to the outside in a handle. The pneumatic motor is mounted on the handle part. The chain saw is connected to the pneumatic motor. A compressed air opening and closing unit is installed on the handle part. The compressed air opening and closing unit selectively opens and closes the compression air channel of the handle part.