Fingerprint-based location positoning using improved KNN

Location estimation has become one of the most popular research areas for the wide application of Location Based Services (LBS). K nearest neighbors (KNN) algorithm is commonly used in fingerprinting approach, and it has been widely used for decades due to its simplicity and effectiveness. However, the main drawback of KNN algorithm is obvious. Theoretical behavior can hardly be obtained because KNN is sensitive to the value of K and it often fails to work well with inappropriate choice of distance metric or due to the presence of numerous irrelevant features. In order to fill this gap, an improved KNN algorithm is introduced. And this algorithm is beneficial to location estimation in a real GSM network.