Mobile Crowd Sensing
Mobile Crowd Sensing (MCS) arises as a new sensing paradigm based on the power of the crowd jointly with the sensing capabilities of various mobile devices, such as smartphones or wearable devices. The increasing popularity of smartphones, already equipped with multiple sensors from
GPS to microphone and camera, paired with the inherent mobility of their owners enables the ability to acquire local knowledge from the individual’s surrounding environment through the mobile device’s sensing properties or even from the individual itself. This local knowledge ranges from location information to more specialized data such as pollution levels going through a longer list of personal and surrounding
context, noise levels or traffic awareness among others.
However, MCS presents numerous and unique research challenges most of them based on the fact that human participation
is in the loop and range from participatory and opportunistic data collection, proper incentive mechanisms, transient network communication and big data processing. Nonetheless, human participation raises singular issues regarding the privacy and security of data, as sensitive information
such as human voice or location may be revealed. Furthermore, the quality and trustworthiness of the contributed data (e.g. counterfeit data contributed by malicious users) should also be addressed.
In a nutshell, MCS is about relying on the crowd to perform sensing task through their sensor-enabled devices as shown in the following image.
Our research focus mainly on three of the most important challenges of MCS, namely:
- User participation
- Data sensing quality
- User anonymity
MCS systems typically involve a very large number of users or crowd sensors in the sensing tasks by collecting and sending local data obtained through their sensor-enabled mobile devices to a data collection center. The performance and usefulness of such sensor networks heavily depends on the crowd sensor’s willingness to participate in the data collection process. Therefore, incentive mechanisms are of utmost importance in MCS scenarios to engage as many crowd sensors and provide to the data collection center a considerable wealth of data.
Data sensing quality
In MCS systems there is no control over the crowd sensors and hence, we cannot control their behavior or assume they will be equally honest. Therefore, the overall quality of the sensor readings may deteriorate if counterfeit data is received from malicious users. Then, the obvious question is how to validate the sensing data that crowd sensors provide to the system. A commonly used approach is to validate the data
depending on the trust level of the crowd sensor that reports it.
An important aspect of MCS scenarios is the collection of potentially sensitive information pertaining to individuals. For instance, GPS sensor readings can be use to track users movements and profile them for other purposes besides their crowd sensing tasks.
A popular approach for preserving users privacy is removing any user identifying attributes from the sensing data before sending it to the data collection center. Obviously, this approach can be applied only in those MCS applications that are solely interested in the actual data and no further interaction with the crowd sensors is necessary. Another approach is to use pseudonyms when sending sensed data to the data collection
center. These pseudonyms are typically randomly generated and bare no relation with the individual’s real identity. In this case we say that individuals benefit from pseudo-anonymity since we cannot infer their real identity but we can still identify subsequent sensor readings as reported by the same user through the same pseudonym.
C. Tanas, J. Herrera. Users as Smart Sensors: A Mobile Platform for Sensing Public Transport Incidents. In Citizen in Sensor Networks, Springer Berlin Heidelberg NY vol. 7685, pp: 81-93. (July 2012) DOI: 10.1007/978-3-642-36074-9_8.
C. Tanas, C. Perez, J. Herrera. Security and privacy challenges in smart sensor networks. In Actas de la XII Reunión Española de Criptología y Seguridad de la Información (RECSI 2012), pp: 423-428. (September 2012).
C. Tanas, J. Herrera. Crowdsensing Simulation using ns-3. In Citizen in Sensor Networks, Springer International Publishing, pp: 47-58. (December 2013) DOI: 10.1007/978-3-319-04178-0_5.
C. Tanas, J. Herrera. When users become sensors: can we trust their readings?. International Journal of Communication Systems, pp: n/a. (October 2013). ISSN: 10745351. DOI: 10.1002/dac.2689.