Probabilistic registration for large-scale mobile participatory sensing

Abstract

Abstract-One of the main benefits of mobile participatory sensing becoming a reality is the increased knowledge it will provide about the real world while relying on a large number of mobile devices. Those devices can host different types of sensors incorporated in every aspect of our lives. However, given the increasing number of capable mobile devices, any participatory sensing approach should be, first and foremost, scalable. To address this challenge, we present an approach to decrease the participation of (sensing) devices in a manner that does not compromise the accuracy of the real-world information while increasing the efficiency of the overall system.To reduce the number of the devices involved, we present a probabilistic registration approach, based on a realistic human mobility model, that allows devices to decide whether or not to register their sensing services depending on the probability of other, equivalent devices being present at the locations of their expected path. We present the design and implementation of a registration middleware based on our techniques, using which mobile devices can base their registration decision. Through experiments performed on real and simulated datasets, we show that our approach scales, while not sacrificing significant amounts of sensing coverage.

Service-oriented middleware for large-scale mobile participatory sensing

“…However, instead of adding phantom traces to increase the MPC of the set of devices, we decided to restrict the area of focus and increase the sensing range (we omit the details due to space constraints. More information can be found in [9]). The chosen range is now 10 km.…”

“…The paper builds on the work presented in [9] where we first described the probabilistic registration solution. It extends the work by providing a deterministic solution that decreases the participation of mobile devices based on a more detailed knowledge of their future mobility.…”

“…However, this approach is costly and requires a computation time that increases linearly with the number of registered devices. To enhance the performance of our solution, we substitute the mobility paths information with a mobility model that allows us to estimate the mobility of registered devices and then compute a probability of the possible coverage (introduced in our previous work in [9]). Although this probabilistic approach is less accurate, it is faster and still meets required sensing coverage.…”

Dioptase: a distributed data streaming middleware for the future web of things

“…This graph is provided by the developer either directly or expressed as a query that is translated into a mashup graph. Using information provided by a discovery system (e.g., registry or distributed protocol [43]) that is aware of Things' locations and available resources, the logical mashup graph is automatically converted into a physical mashup graph (VP, EP), were each vp i ∈ VP is a pair (vl, n) that maps a component vl onto a host device n, as depicted in Figure 1(b). In particular, depending on its capabilities, a Thing can be assigned either a single component or an entire subgraph.…”

Novel sampling algorithm for Levy-walk based mobile phone sensing

“…We use LW mobility model given in [12] for the simulation of proposed sampling algorithm. Authors of [13] discusses reducing number of participatory users plus improving coverage by assuming that users are aware of their path. Authors of [4] presents collaborative mobile phone sensing using cloud network.…”