No abstract

No abstract

No abstract

Proposed is a light-weight unsupervised decision tree based classification method to detect the user's postural actions, such as sitting, standing, walking and running as user states by analysing the data from a smartphone accelerometer sensor. The proposed method differs from other approaches by applying a sufficient number of signal processing features to exploit the sensory data without knowing any a priori information. Experiments show that the proposed method still makes a solid differentiation in user states (e.g. an above 90% overall accuracy) even when the sensor is operated under slower sampling frequencies.Introduction: Human activity recognition (HAR) is achieved by analysing the contextual information collected from heterogeneous and unobtrusive wearable/built-in mobile device sensors, where an emerging topic is to infer the user's postural actions, such as sitting, standing, walking and running [1,2]. It uses either statistical tool based (e.g. hidden Markov models [3]) or pattern recognition based (e.g. Gaussian mixture models (GMMs) [4], k-nearest neighbours [3]) classification models while detecting physical activities. However, the former methods mostly require predefined and user-manipulated system parameter settings, such as arbitrary formation of the state transition matrix, or building filtering coefficients; while, the latter rely on first creating high dimensional feature vectors to exploit signal characteristics (e.g. mean, standard deviation, correlation, frequency and wavelet transform models) of sensory data, and then clustering these vectors according to user-manipulated (e.g. mostly visually observed) training data classes. The drawback of these algorithms also stems from the offline decision process due to the computational complexity and non-instant decision-making capability, which makes HAR systems almost impossible to run real-time on resource constrained mobile devices.Towards this end, this Letter proposes a solution (as shown in Fig. 1) to perfectly exploit acceleration signals within the fast decision tree (DT) classifiers. The solution applies a sufficient number of signal processing and statistical techniques (light-weightiness) without receiving any a priori information related to user state classes, and setting any predefined/fixed thresholds over any specific acceleration spaces (unsupervised learning) in order to differentiate user activities.

New-generation mobile devices will inevitably be employed within the realm of ubiquitous sensing. In particular, smartphones have been increasingly used for human activity recognition (HAR)-based studies. It is believed that recognizing human-centric activity patterns could accurately enough give a better understanding of human behaviors. Further, such an ability could have a chance to assist individuals to enhance the quality of their lives. However, the integration and realization of HAR-based mobile services stand as a significant challenge on resourceconstrained mobile-embedded platforms. In this manner, this paper proposes a novel discrete-time inhomogeneous hidden semi-Markov model (DT-IHS-MM)-based generic framework to address a better realization of HAR-based mobile context awareness. In addition, we utilize power-efficient sensor management strategies by providing three intuitive methods and constrained Markov decision process (CMDP), as well as partially observable Markov decision process (POMDP)-based optimal methods. Moreover, a feedback control mechanism is integrated to balance the tradeoff between accuracy in context inference and power consumption. In conclusion, the proposed sensor management methods achieve a 40% overall enhancement in the power consumption caused by the physical sensor with respect to the overall 85-90% accuracy ratio due to the provided adaptive context inference framework.Index Terms-Context-aware framework, human activity recognition (HAR), optimal sensing, power efficiency. I. INTRODUCTIONT HE ever-increasing technical advances in embedded systems, together with the proliferation of growing development and deployment in small-size sensor technologies, have enabled smartphones to be repurposed to recognize daily oc-Manuscript

No abstract

No abstract

Energy consumption is a major concern in context-aware smartphone sensing. This paper first studies mobile device-based battery modeling, which adopts the kinetic battery model (KiBaM), under the scope of battery non-linearities with respect to variant loads. Second, this paper models the energy consumption behavior of accelerometers analytically and then provides extensive simulation results and a smartphone application to examine the proposed sensor model. Third, a Markov reward process is integrated to create energy consumption profiles, linking with sensory operations and their effects on battery non-linearity. Energy consumption profiles consist of different pairs of duty cycles and sampling frequencies during sensory operations. Furthermore, the total energy cost by each profile is represented by an accumulated reward in this process. Finally, three different methods are proposed on the evolution of the reward process, to present the linkage between different usage patterns on the accelerometer sensor through a smartphone application and the battery behavior. By doing this, this paper aims at achieving a fine efficiency in power consumption caused by sensory operations, while maintaining the accuracy of smartphone applications based on sensor usages. More importantly, this study intends that modeling the battery non-linearities together with investigating the effects of different usage patterns in sensory operations in terms of the power consumption and the battery discharge may lead to discovering optimal energy reduction strategies to extend the battery lifetime and help a continual improvement in context-aware mobile services.