Huabei Yin

A weight-based map matching method in moving objects databases

In location management, the trajectory represents the motion of a moving object in 3D space-time, i.e., a sequence (x, y, t). Unfortunately, location technologies, cannot guarantee error-freedom. Thus, map matching (a.k.a. snapping), matching a trajectory to the roads on the map, is necessary. We introduce a weight-based map matching method, and experimentally show that, for the offline situation, on average, our algorithm can get up to 94% correctness depending on the GPS sampling interval.

Search-and-Discover in Mobile P2P Network Databases

In this paper we propose a novel algorithm called Rank-Based Broadcast (RBB) for discovery of local resources in mobile P2P networks. With RBB, each moving object periodically broadcasts the most relevant resource reports and queries it knows to its neighbors, and the contribution is in determining how to rank the reports and queries in terms of their relevance, when to broadcast them, and how many to broadcast. A major difference between RBB and many existing algorithms in the resource discovery and publish/subscribe literature is that RBB does not rely on any pre-established routing structure, and therefore is able to adapt to both high mobility environments. In the paper we experimentally compare RBB with flooding and PSTree, a publish/subscribe algorithm for wireless ad-hoc networks. The results show that RBB by far outperforms the other two algorithms.

Accuracy and Resource Consumption in Tracking and Location Prediction

Tracking is an enabling technology for many location based services. Given that the location of a moving object changes continuously but the database cannot be updated continuously, the research issue is how to accurately maintain the current location of a large number of moving objects while minimizing the number of updates. The traditional approach used in existing commercial transportation systems is for the moving object or the cellular network to periodically update the location database; e.g. every 2 miles. We introduce a new location update policy, and show experimentally that it is superior to the simplistic policy currently used for tracking; the superiority is up to 43% depending on the uncertainty threshold. We also introduce a method of generating realistic synthetic spatio-temporal information, namely pseudo trajectories of moving objects. The method selects a random route, and superimposes on it speed patterns that were recorded during actual driving trips.

Extracting Semantic Location from Outdoor Positioning Systems

With help of context, computer systems and applications could be more user-friendly, flexible and adaptable. With semantic locations, applications can understand users better or provide helpful services. We propose a method that automatically derives semantic locations from user’s trace. Our experimental results show that the proposed method identities up to 96% correct semantic locations.

Dissemination of spatial-temporal information in mobile networks with hotspots

In this paper we examine the dissemination of reports about resources in mobile networks with hotspots, where hotspots, vehicles and sensors communicate with each other via short-range wireless transmission. Each disseminated report represents information about a spatial-temporal event, such as the availability of a parking slot at a particular time or the detection of an injured in an earthquake damaged building. We propose an opportunistic dissemination paradigm, in which a moving object transmits the reports it carries to encountered peers and obtains new reports in exchange. We address two issues in such an environment. First, we develop an architecture that allows a moving object to receive resource reports opportunistically. Second, we study how the received reports are used by a consumer to reduce resource discovery time. The proposed system has the potential to create a completely new information marketplace.

Resource discovery using spatio-temporal information in mobile ad-hoc networks

In this paper we examine the benefit of reports about resources in mobile ad-hoc networks. Each disseminated report represents information about a spatio-temporal event, such as the availability of a parking slot or a cab request. Reports are disseminated by a peer-to-peer broadcast paradigm, in which an object periodically broadcasts the reports it carries to encountered objects. We evaluate the value of resource information in terms of how much time is saved when using the information to discover resources, compared to the case when the information is not used.

Searching Local Information in Mobile Databases

A mobile ad-hoc network (MANET) is a set of moving objects that communicate with each other via unregulated, short-range wireless technologies such as IEEE 802.11, Bluetooth, or Ultra Wide Band (UWB). No fixed infrastructure is assumed or relied upon. An important application domain of MANET’s is local resource discovery. In a local resource discovery application, a user finds local resources that satisfy specified criteria. For example, a driver finds an available parking slot in a region by receiving information generated by the parking meter, or gets the traffic conditions on a highway segment a mile ahead; a cab driver finds a near-by customer, or a participant at a convention finds another participant with a matching profile.