A Maintainability Evaluation Method of Large Scale Wireless Sensor Networks Based on Sample Entropy

The evaluation of maintainability for large scale wireless sensor networks (WSNs) is crucial. However little work has been done in such area. In this paper, a maintainability evaluation method is proposed, which is based on sample entropy First of all, we define several parameters, such as affect factor, quality allocation matrix, quality-cost vector, and contribution weight. Each affect factor has a quality-allocation vector. The quality allocation vectors at the same layer of the affect factors can form a quality allocation matrix. The contribution weight, quality allocation matrix, and quality-cost vector work together to construct the system cost function. The quality-cost vector and contribution weight can be subjective or objective. The subjectiveness means that it is given from the expert experience. On the other hand, the objectiveness means that it is derived from the data collected from the on-site field. In the proposed method, the subjectiveness and objectiveness are combined to achieve better system performance. The proposed evaluation method is verified by the data collected in an application field at SIAT dorm. It shows that our proposed method has good performance.