摘要
QoE测量点选址问题是选择尽可能少的测量点来准确反映网络中用户获取服务的情况。本文基于失效概率已知的QoE测量点选址模型,用区间描述失效概率的不确定性,建立了QoE测量点选址的鲁棒模型,并将其转化为混合整数线性规划求解。测试结果表明了鲁棒选址模型对考虑节点失效的QoE测量点选址问题的有效性,算例分析表明覆盖率和失效个数对选址方案有不同程度的影响。
Facility location issue of QoE test points is to accurately reflect the obtained service of all network users with as less test points as possible. We established a robust model of QoE test points location with an interval to indicate the uncertainty of failure possibility based on QoE test points location model of given failure possibility. We then converted it to a mixed integer linear programming model. Test results show that the model is effective for QoE test points location issue with node failure. Case analysis demonstrates that coverage rate and failure number have different impact on location scheme.
出处
《山东科学》
CAS
2016年第4期80-86,共7页
Shandong Science
基金
国家重点基础研究发展计划(973计划)(2011CB706900)
国家高技术研究发展计划(863计划)(2011AA01A102)
国家自然科学基金(11571015
11331012)
中国科学院战略性先导科技专项(XDA06010302)
中国科学院大数据挖掘与知识管理重点实验室开放课题
华为技术有限公司资助
关键词
设施选址
鲁棒优化
节点失效
QOE
facility location
lobust optimization
node failule
quality of experience