Extending lifetime of sensor network by introducing mobile nodes and its evaluation using real devices

在本研究中，我们提出了一种使k覆盖目标域的无线传感器网络的生存期最大化的方法，即使最小数量的节点运行，并使其他节点处于睡眠状态。为此，我们首先制定问题，确定每个传感器节点的工作模式和每次的多跳通信路由，以最大化k覆盖时间。该问题是一个典型的组合优化问题，很难在实际时间内找到最优解。因此，我们提出了一种方法，通过细分整个时间段，周期性地找到每个较短时间段的解。我们提出了两种方法；第一种方法是根据活动节点对域k-覆盖的贡献程度依次选择活动节点，第二种方法是找到每个节点1-覆盖域的层，并为k-覆盖选择k个活动层。第二种方法被认为容错性更强。为了评估我们的算法在k覆盖时间方面的有效性，我们将我们的方法与禁用某些提议特征的方法进行了比较。通过基于仿真的比较，我们证实，与其他方法相比，无论k和节点数量如何，所提出的方法的寿命都要长1.1到1.7倍。

In this study, we proposed a method for maximizing the lifetime of a wireless sensor network which k-covers the target field, by making minimum number of nodes operate and putting other nodes in sleep. For this objective, we first formulated the problem to decide the operating mode for each sensor node and the multi-hop communication route at each time to maximize the k-coverage time. This problem is a typical combinatorial optimization problem, and thus it is hard to find the optimal solution for practical time. So, we proposed a method which periodically finds the solution for each of shorter time periods, which are made by subdividing the entire time period. We proposed two methods; the first one is to choose active nodes in order of their contribution degree to the field k-coverage, and the second one is to find layers that each of them 1-covers the field, and choose k active layers for k-coverage. The second method is considered to be more fault-tolerant. In order to evaluate the effectiveness of our algorithms in terms of k-coverage time, we compared our methods with methods in which some of the proposed features are disabled. Through simulation-based comparison, we confirmed that the proposed methods achieve 1.1 to 1.7 times longer lifetime regardless of k and the number of nodes, than the other methods.