BLE Beacon with User Traffic Awareness Using Deep Correlation and Attention Network

Abstract

Bluetooth Low Energy (BLE) beacon network is one of the essential infrastructures for many IoT and smart city applications due to the proliferation of Bluetooth-enabled devices. However, the BLE beacon network usually suffers from high maintenance costs due to the short battery lifetime. A recent work proposed duty-cycling BLE advertising interval subject to the detected existence of a user; if a user is detected, the beacon operates in a shorter advertising interval, otherwise operates in a longer advertising interval to reduce its energy consumption. However, since such reactive approach operates based on hardcoded advertising intervals for the two scenarios, the energy-efficiency is bound to be sub-optimal. To overcome this limitation, this paper proposes to predict the user traffic condition thereby adapting the optimal advertising interval based on the predicted user traffic condition. To this end, we introduce a novel neural network architecture that learns partial correlation and leverages attention mechanism to make accurate predictions with minimum prediction lag. The effectiveness of the proposed learning methods is verified by comprehensive simulations. It is proved that the proposed method can extend a beacon lifetime by at least 200% more than the state-of-the-art techniques.