Application of Privacy Principles to Wireless Sensor Networks

Networks of small, wireless computers equipped with sensing hardware that can detect environmental characteristics like light, temperature, vibration, humidity, or gas concentration are becoming increasingly widespread. Researchers have proposed that such sensor networks be used to monitor a variety of environments, and there is particular in-terest in monitoring in indoor building environments [6, 10]. Possible benefits of such networks include increased efficiency of heating and cooling systems, increased awareness for building occupants of their surroundings (e.g. which conference rooms are in use, whether a fellow employee is at his desk), improved security, safety, and alarm systems, and so on. Unfortunately, such networks also pose substantial privacy risks to users of buildings: continu-ous monitoring and data collection, especially when combined with personally identifying information that can be gleaned from door security systems, cameras, or databases correlating particular offices with employees, make it possible for businesses and governments to know the whereabouts and activities of individuals at all times. In this paper, we discuss possible privacy violations that can arise from wireless sensor networks set up to monitor buildings and present some simple techniques that could be used to reduce privacy risks while still maintaining the ability to preserve many of the benefits of sensor-based monitoring. We assume that all data collection is done through a unified database-like interface (discussed briefly in Section 1), and that it is feasible to augment such an interface with some set of privacy aware constraints. We further assume that occupants of buildings have som