针对资源预留过程中的＂预留违约风险＂问题,提出一种可量化分析收益与风险均衡的资源预留策略.该策略基于真实预留请求日志的统计特性,采用效用函数将预留模型的多目标规划问题转化为二次规划问题,通过求解该二次规划问题来确定预留方案.通过理论分析给出了策略模型的有效性证明和求解方法,并通过仿真实验在较大规模的模拟网格系统中检验了所提出的预留策略的性能.结果表明,该策略在有效保障任务服务质量需求的前提下,能依据网格任务不同的风险偏好来确定最优的资源预留组合策略,从而有效解决＂预留违约风险＂和＂资源利用率＂之间的矛盾.

In order to prevent the violation risk existing in the resource reservation,a novel resource reservation strategy based on the quantified balance between benefits and risks is proposed.The strategy takes into consideration the statistical logs of actual reservation services and uses the utility function to change the multi-objective programming problem of the reservation model into a quadratic one.By solving the quadratic programming problem,a suitable reservation scheme can be determined.Moreover,theoretical analysis is performed to prove the validity and to present the solution to the model,and some simulations are carried out to verify the effectiveness of the proposed strategy in a large-scale simulated grid system.Simulation results indicate that the proposed strategy effectively guarantees the quality of service and provides optimal source reservation strategy according to the risk preferences of different grid services.Thus,the conflict between the reservation violation risk and the resource utilization rate is successfully prevented.