WalkCompass: Finding Walking Direction Leveraging Smartphone's Inertial Sensors

Determining moving direction with smartphone’s inertial sensors is a well known problem in the field of location service. Compass alone cannot solve this problem because smartphone’s compass cannot adapt with the orientation change. GPS, works successfully in outdoor, is not suitable in indoor scenario. Another well known approach called dead-reckoning needs to know phones initial orientation and over time it keeps accumulating errors. An ideal system should be able to calculate the heading direction without any user intervention and it should be accurate and light-weight. Moreover in order to be able to work as a generic system, the system should not be restricted by any strict requirement for its function. Considering all the limitations of the conventional solutions and all the requirements of a generic solution, we propose a smartphone based solution called WalkCompass designed especially for pedestrians. Our system focuses on the variation of force during normal human walk and captures this property with the help of the smartphone’s inertial sensors. Therefore, the algorithm is inherently free from errors generated by the orientation of the phone. The performance of the system does not depend on the holding style or location of the phone in the body. The algorithm can work fast enough to determine the direction of movement in real time and because of its low complexity, the complete system can be implemented in a regular smartphone. WalkCompass does not need any bootstrapping and can produce results at each step of a walk. The heading direction estimated by WalkCompass has an average error of 6 degrees, which is half of the contemporary solutions.