**数学与系统科学研究院**

**计算数学所学术报告**

**报告人**：
Yinyu Ye

Stanford University, U.S.A

**报告题目**
Semidefinite Programming for Euclidean Distance Geometry Problems

**Abstract:**
We describe an semidefinite programming (SDP) based model and method for the
position estimation problem in wireless sensor network localization.
The optimization problem is set up so as to minimize the error in sensor
positions to fit incomplete and noisy distance measures. We use the SDP duality
theory to derive conditions on whether a network is localizable or not
when the distance measures are accurate. We also present probabilistic
analyses of the SDP solution when the distances are noisy. In all cases,
observable gauges are developed to check the quality of the point estimation
of all sensors and to detect erroneous sensors. Moreover, the SDP model also
provides an effective method for solving other distance geometry problems.

**报告时间**：2004年8月2日 上午9：00-11：00

下午3：00-5：00

**报告地点**：科技综合楼三层报告厅