Received Signal Strength Prediction Using Generative Adversarial Networks for Indoor Localization

The popularity of WiFi networks has led to the adoption of fingerprint-based WiFi localization as one primary method for indoor location tracking. However, this technique requires a significant amount of time and effort to collect data at numerous reference points (RPs) to ensure accuracy. To reduce the cost and improve efficiency, generative models can be used to generate received signal strength (RSS) fingerprints. This paper proposes a Deep Convolutional Generative Adversarial Network (DCGAN) based model for building RSS fingerprint maps that can generate a comprehensive fingerprint database using only the location of a wireless access point. The proposed approach is expected to lower the cost and effort involved in the collection of RSS fingerprints while maintaining a high level of accuracy.