针对基于双曲线定位的DV-Hop算法中误差项的异方差性引起的定位误差大的问题，提出了一种基于加权双曲线定位的DV-Hop改进算法。算法分析了基于双曲线定位的DV-Hop算法模型中误差项的异方差性，用加权最小二乘法对异方差性进行纠正，对加权最小二乘法中的权值矩阵进行了推导并得到与跳数相关的最佳权值矩阵，使得误差项满足同方差性，所得估计值更加接近最佳线性无偏估计。仿真实验结果表明，所提算法在节点定位精度上较现有的基于双曲线定位的DV-Hop算法都有一定提高。(Aiming at the problem that the heteroscedasticity of observation errors leads to low locating accuracy in DV-Hop algorithm based on hyperbolic positioning, this paper proposes an improved DV-Hop algorithm which based on weighted hyperbolic positioning. The algorithm analysis the heteroscedasticity of the error term in the model and takes advantage of weighted least square method which corrects heteroscedasticity straightforwardly. Then, the weight matrix related to least hops between nodes is deduced, the estimators thus obtained are more close to the best linear unbiased estimator. The simulation results show that the location performance of the improved algorithm is better than that of the exsiting DV-Hop algorithm which based on hyperbolic positioning.)