The different network node densities made larger difference among average single-hop distance of different hop count values in traditional DV-Hop node localization algorithm,and the error increased with the increase of the hop count value. An improved DV-Hop algorithm was proposed to reduce the influence of the average single-hop distance difference on the node localization accuracy. Firstly,the strategy of hop count classification was proposed to classify the different hop counts in the network,so as to reduce the difference of average single-hop distance between different hop counts,and to enhance the accuracy of node localization. Then,by using the strategy of the improved weight coefficient,the weighted least squares estimation was improved to adapt to the nonlinear variation of the cu-mulative error,which could better control the weight of the different hop counts in the least squares estimation,and further enhance the accuracy of node localization. The experimental results show that the improved algorithm can ef-fectively reduce the influence of the average single-hop distance difference and the large hop count value on the node localization,its node localization performance is obviously superior to traditional DV-Hop node localization al-gorithm,compared with comparative literature also has a certain improvement,and it has better adaptability to differ-ent network node densities.%在传统DV-Hop节点定位算法中,不同的网络节点密度使得节点之间不同跳数的平均每跳距离差异较大,跳数越多误差越大.为了减小平均每跳距离差异对节点定位精度的影响,提出一种DV-Hop改进算法.改进算法首先提出跳数分类的策略对网络中不同的跳数进行分类,以减小不同跳数之间平均每跳距离差异的影响,提高节点的定位精度;然后对加权最小二乘估计进行改进,采用改进的权系数取值策略来适应累积误差的非线性变化,从而更好地控制不同跳数在最小二乘估计中的权重,以减小因跳数增加而产生的累积误差,进一步提高节点的定位精度.实验结果表明,改进算法可以有效地减小平均每跳距离差异以及高跳数对节点定位的影响,节点定位性能显著优于传统DV-Hop节点定位算法,相较于对比文献也有一定的提升,并且对不同的网络节点密度具有更好的适应性.
展开▼
