抄録J-032
Grouping Smartphone Users based on Wi-Fi Signal Strengths
Grouping Smartphone Users based on Wi-Fi Signal Strengths
◎Jose P. Alvarez・Teemu Leppanen（東大）・岩井将行（電機大）・小林博樹・瀬崎 薫（東大）
Wi-Fi access point signal strengths are commonly used for indoor localization. We extend their use by proposing a method to group smartphone users into flocks, by considering the cosine similarity between these measurements.
The raw measurements are divided into same size groups and local fingerprints are obtained by averaging the power, which we believe is the most representative measure of their temporal vicinity. We conducted experiments with this method using Android smartphones in our mobile agent-based distributed applications framework. The experiment results yielded similarities with an average gap of 40% for distant smartphones.
The raw measurements are divided into same size groups and local fingerprints are obtained by averaging the power, which we believe is the most representative measure of their temporal vicinity. We conducted experiments with this method using Android smartphones in our mobile agent-based distributed applications framework. The experiment results yielded similarities with an average gap of 40% for distant smartphones.
