Toward long-term quality of protection in mobile networks: a context-aware perspective

Sensor-equipped smartphones as well as wearable devices have undoubtedly become the predominant source of user-generated data in mobile networks. The proliferation of user-generated data has created a plethora of opportunities for personalized services based on the states of users and their surrounding environments. Those personalized services, although improving users' perceived quality of experience, have raised severe privacy concerns, as most of these applications aggressively collect users' personal data without providing clear statements on the usage and disclosure policies of such sensitive information. In order to sustain personalized services with long-term privacy preservation, disruptive paradigms are required. We envision that context awareness is a key pillar to providing long-term quality of protection (QoP) for individual privacy. In particular, users transit between different contexts, including mobility modes and social activities, and these contexts are temporally or logically correlated, which can be leveraged by adversaries to compromise users' privacy. In addition, users may have different QoP preferences in different contexts. With these salient features in mind, this article investigates context-aware QoP mechanism designs for personalized services in mobile networks. We discuss possible attacks and propose corresponding countermeasures. In particular, we develop a QoP framework that exploits context awareness to achieve better trade-offs between service quality and privacy protection in long-term services. Finally, we provide some implications for future context-aware QoP mechanism designs by conducting a case study on smartphone traces.

Introduction

Due to continuing advances in low-power sensors and actuators, we are entering an era of rapid expansion in wearable and pervasive computing. Nowadays, it is common for people to carry smart devices, such as sensor-equipped smartphones, smart watches, and healthcare devices. Those devices periodically measure users' physical activities (e.g., Fitbit) and physiology, such as Samsung Simband with electrocardiogram (ECG) and photoplethysmogram (PPG) sensors embedded, making it possible to continuously track users' contexts including mobility modes, social activities, and health conditions [1]. These devices are normally capable of wireless connectivity, which enables them to upload their sensor readings to their service providers. Examples of these service providers include healthcare providers that own electronic medical record systems for users' data collection and analysis, activity trackers that keep track of users' trajectories or mobility modes, and environment-based applications that offer personalized services based on the operating conditions of users and their surrounding environments.