| 摘要: |
| 针对传统时差定位算法在量测噪声较大情况下定位性能不佳的缺点，提出了一种基
于二阶锥规划的新时差定位算法。该算法通过凸松弛和引入惩罚项，将难以求解的用户位置
最大似然估计问题转换为一个易于求解的二阶锥规划问题，并将松弛问题的最优解作为用户
位置的初始估计，利用传统的泰勒级数展开法得到最终定位结果。仿真给出了不同基站数目
及量测噪声下算法的定位性能。仿真结果表明，在量测噪声较大的情况下，新算法的定位精
度
仍可以逼近理论克拉美罗下限, 而且算法中惩罚因子的选取范围易于确定。 |
| 关键词: 到达时差；定位算法；最大似然估计 泰勒级数展开；二阶锥规
划；惩罚因子 |
| DOI： |
|
| 基金项目: |
|
| A new TDOA location algorithm based on second order cone programming |
| JIN Jia-bao,ZHANG Song,YANG Jing-shu |
| () |
| Abstract: |
| The traditional TDOA(Time Difference of Arrival) location algorithms have large
performance loss as
the measurement noise is high. To overcome this drawback, this paper proposes a
new effective TDOA location algorithm based on second order cone programming(SO
CP). B
y introducing a penalty term and relaxing the equality constrains, the nonlinear
and nonconvex maximum likelihood estimation problem for user position is
transformed into a convex optimization problem, named second order cone programm
ing that can be efficiently solved by modern interior point methods. The optimal
solution of relaxed problem is used as the initial guess for traditional Taylor
method to estimate the user position. The simulation provides the location perf
ormance versus measurement noise under different numbers of base station. Simula
t
ion results show that the performance of proposed algorithm can attain the Crame
r-Rao lower bound as the noise variance is high. The intervals of penalty factor
are also discussed in this paper. |
| Key words: TDOA location algorithm maximum likelihood estimator Taylor series expansion sec
ond order cone programming penalty factor |
