Near-optimal diagnosis system deployment in wireless sensor networks

Wireless sensor networks (WSNs) have been extensively applied in many important fields with larger scale and more complex structure. The applications of WSNs are regarded as a sustainable solution to provide ongoing and efficient monitoring services in the real world. When such an application of WSNs faces poor performance or unexpected condition, the administrator needs to deploy a diagnosis system to diagnose the task. One possible way is to transform some original motes as diagnosis motes by using reprogramming technique. However, the challenge is how to achieve best efficiency in the diagnosis nodes selection processing. Moreover, the required network evidence is always distributed in multidimensional data spaces. The existing approaches fail to consider this multidimensional feature of evidence in diagnosis nodes selection problem. To address the above issues, we propose a solution for the multidimensional diagnosis nodes selection problem and give the corresponding upper bound. Lastly our experimental results demonstrated that our approach is scalable and applicable for real WSNs.