Adaptive Nearest Neighbor Queries in Travel Time Networks * Wei-Shinn Ku, Roger Zimmermann, Haojun Wang, and Chi-Ngai Wan Computer Science Department University of Southern California Los Angeles, California 90089 USA {wku, rzimmerm, haojunwa, cwan}@usc.edu ABSTRACT Nearest neighbor (NN) searches represent an important class of queries in geographic information systems (GIS). Most nearest neigh- bor algorithms rely on static distance information to compute NN queries (e.g., Euclidean distance or spatial network distance). How- ever, the final goal of a user when performing an NN search is of- ten to travel to one of the points of the search result. In this case, finding the nearest neighbors in terms of travel time is more im- portant than the actual distance. In the existing NN algorithms dy- namic real-time events (e.g., traffic congestions, detours, etc.) are usually not considered and hence the pre-computed nearest neigh- bor objects may not accurately reflect the shortest travel time. In this paper we propose a novel travel time network that integrates both spatial networks and real-time traffic event information. Based on this foundation of the travel time network, we develop a local- based greedy nearest neighbor algorithm and a global-based adap- tive nearest neighbor algorithm that both utilize real-time traffic in- formation to provide adaptive nearest neighbor search results. We have performed a theoretical analysis and simulations to verify our methods. The results indicate that our algorithms remarkably re- duce the travel time compared with previous nearest neighbor so- lutions. Categories and Subject Descriptors H.3.3 [Information Storage and Retrieval]: Information Search and Retrieval—Search Process; H.2.8 [Database Management]: Database Application—spatial databases and GIS General Terms Algorithms Keywords Nearest neighbor query, travel time network, location-based ser- vices, advanced traveler information systems. 1. INTRODUCTION * This research has been funded in part by NSF ITR grant CMS- 0219463, equipment gifts from Intel and Hewlett-Packard, unre- stricted cash grants from the Lord Foundation and by the Integrated Media Systems Center, a National Science Foundation Engineering Research Center, Cooperative Agreement No. EEC-9529152. Nearest neighbor (NN) queries are of significant interest for ap- plications that work with spatial data. A sample query could be to “find the nearest gas station from my current location.” Previous work [12, 6] has resulted in efficient techniques to compute NN queries in Euclidean space. More recently, novel algorithms [7, 15, 11] have been proposed to compute NN queries in spatial networks. These methods extend NN queries by considering the spatial net- work distance, which provides more realistic measure for applica- tions were objects are constrained in their movements. However, these existing techniques only consider static models of spatial net- works: pre-defined road segments with fixed road conditions are used in computing nearest neighbors. Thus any real-time events (e.g., detours, traffic congestions, etc.) affecting the spatial net- work cannot be reflected in the query result. For example, a traffic jam occurring on the route to the computed nearest neighbor most likely elongates the total driving time. More drastically, the closure of a restaurant which was found as the nearest neighbor might even invalidate a query result. This motivates the need for new algo- rithms which extend existing NN query techniques by integrating real time event information. Recent advances in personal locater systems (e.g., GPS), wire- less communication technologies (e.g., 802.11x), and peer-to-peer networks (P2P) have created an innovative environment that allows the exchange of real time traffic information between peers. By leveraging ad-hoc networks, traffic information can be shared in a P2P manner among mobile hosts (MH) and thus local traffic in- formation (e.g., driving speed of vehicles) can be considered when computing NN queries. Furthermore, cellular communication en- ables remote traffic information server (TIS) access such that col- lecting and disseminating traffic information for a much wider area becomes possible. In this paper we propose two novel adaptive nearest neighbor query algorithms which incorporate real time traffic information. Compared with existing work, our design leverages the communi- cation among mobile hosts and traffic information servers to adap- tively compute more accurate nearest neighbor results. The contri- butions of our work are: • We introduce the concept of travel time networks (TTN). A travel time network integrates the real time traffic informa- tion with a static data of a spatial network. The length of each edge in the travel time network represents the driving time of each road segment. As a result, NN queries are computed on travel time networks more realistically than on spatial road networks. • We propose a Local-based Adaptive Nearest Neighbor query algorithm (LANN). LANN utilizes the peer-to-peer commu-