Abstract: Accuracy and precision are importance parameters of self-localization systems and algorithms for wireless sensor networks.Based on the analysis of existed calibrations，Time Difference Of Arrival（TDOA） ranging of ultrasonic sensor and RF module and multilateration are used，a self-localization algorithm based on on-line calibration is presented to improve localization performance.Wireless sensor networks is generally indeterminacy deployed，and a densely deployed network provides sufficient amount of data redundancy，which is the basis of the proposed algorithm.Test shows that by applying the proposed algorithm，the error affection of localization is reduced efficiently.

摘要： 准确度和精确度是无线传感器网络定位系统重要的指标。分析现有的校正技术，在基于超声波传感器和无线射频模块的到达时间差（TDOA）测距技术和多边测量定位算法的基础上，结合校正技术，提出一种基于在线校正的定位算法。改进的算法针对传感器网络在实际应用中的不均匀性布撒，充分利用密集布撒下节点定位的冗余信息，以提高密集型网络的定位精度，从而提高整个网络的定位性能。实验表明，改进的定位算法有效地减小了网络的定位误差。