Sensor node localization using weighted and iterative maximum likelihood

We propose a localization method based on an enhanced maximum likelihood (ML) method, which uses the probability density function of radio signal strength indicators (RSSIs). ML is a method used for estimating node location with high accuracy. However, it often requires a large number of anchor nodes, which results in a high cost. To solve this problem, we introduce two key features into the ordinary ML method: iterative multilateration and certainty weight. The former enables the localization of nodes, which cannot directly obtain sufficient information from anchors to estimate their own location, by using the neighboring nodes with already estimated locations as pseudo-anchors (PAs). The latter provides high accuracy by weighting the location of the PA depending on its certainty. The proposed method performs localization in a distributed manner and has high scalability. The evaluation results show that our method can estimate many node locations with a higher accuracy than the original ML.