Xiujun Ma

An On-Road Wireless Sensor Network Approach for Urban Traffic State Monitoring

Wireless sensor networks are expected to be deployed on urban roadways to monitor the traffic continuously. One of the requirements of traffic monitoring is displaying the traffic states of the front roadways, which can guide the drivers to choose the right way and avoid potential traffic congestions. In this scenario, the information of traffic state changes should be refreshed as early as possible. We propose an adaptive segmentation of the traffic flow based on discrete Fourier transform, which responses timely to traffic state changes without inducing large error. On the other hand, considering the limited power of wireless sensor networks, we propose a novel algorithm for in-network aggregation of the traffic flow time-series, which reduces the communication cost between the sensor nodes and base station significantly. The proposed algorithm scales well with the size of the sensor networks. Our methods are computationally efficient and suitable to be implemented on sensor nodes. The primary experiments on PeMS data demonstrate the effectiveness and energy efficiency of our approach.

Inferring properties and revealing geographical impacts of intercity mobile communication network of China using a subnet data set

This article provides a novel and practical approach for investigating the characteristics of intercity telecommunication network whose overall and complete information is unavailable. Using a mobile phone call data set covering 4.39 million subscribers registered in a particular region, we construct two intercity mobile communication subnets and infer characteristics of the whole intercity mobile communication network of China. Results confirm that intercity communication intensity is characterized by the gravity model. The communication intensity based on mobile call number decreases along the distance with a scaling exponent 0.5, whereas the scaling exponent for the communication intensity based on mobile call duration is 0.4. Moreover, we uncover the rank-size distribution of tie strength (mobile call number and duration) between a city and its neighbours. The rank-size law of tie strengths between cities is mainly determined by the rank-size distribution of cities. The distance between cities plays a less decisive role than the size distribution in the network, but significantly impacts mobile communication patterns. The call duration of individual intercity mobile communication is generally positively correlated to the communication distance, explaining why the distance decay of communication intensity based on call durations is slower than that based on call numbers. The contribution of this research is twofold. First, we identify the distance decay effect in intercity mobile communications of China and uncover the dominant impact of the rank-size distribution of cities. Second, a method for estimating the properties of the whole network according to the observed interactions of its subnets is developed.

Using contour lines to generate digital elevation models for steep slope areas: a case study of the Loess Plateau in North China

In soil erosion models digital elevation models (DEMs) play an important role. Most interpolations from contour lines fail to address multi-value cells (MVCs) problems and therefore find it difficult to deal with steep slopes. These interpolation methods randomly assign a single value for MVCs and use this value in interpolation for nearby cells. This approach is very inaccurate. The frequency of MVCs and the problem caused by them was investigated, and the errors created by using the random values for MVCs was analyzed in this paper. A special treatment for MVCs and practical solution to create more accurate interpolation cell values in DEM building was developed. The new approach involves storing additional information of contour lines that go through the MVCs, such as maximum height and minimum height values, and the cardinal orientation relationship of the contour lines. A special MVC filter kernel was developed to decide the appropriate elevation value for interpolation. The computation method is based on raster data. An area on the Loess Plateau in North China was selected as an example to demonstrate the problems of the previously common used approach and to show results of the new method.

A novel approach to estimate human space-time path based on mobile phone call records

The spatio-temporal behavior of people is of significant importance for a variety of social and economic applications. The wide use of mobile phone provides a new way to obtain the space and time information of citizens by means of the positioning mechanism of cellular network. In the conceptual framework of time geography, this kind of data could be used for generalize a space-time path of human movement behavior. However, phone call records for every user are very parse, so it calls for reasonable estimation methodologies, and few pilot studies could utilize the real dataset to put it into practice. This paper proposed a statistical approach to estimate a persons space-time path. Location Stability Index is defined to measure the regularity and mobility stability of a person. The approach is adaptive to the spatio-temporal distribution of calling activity, showing good robustness and flexibility for different kinds of users. The location stability concept is also applied to analyze real data performance and parameter determination. We realized the algorithm and did experiments on real cell phone call dataset of a Chinese city, and the results showed that our methodology performed well according to the real situation.

An online approach based on locally weighted learning for short-term traffic flow prediction

Traffic flow prediction is a basic function of Intelligent Transportation System. Due to the complexity of traffic phenomenon, most existing methods build complex models such as neural networks for traffic flow prediction. As a model may lose effect with time lapse, it is important to update the model on line. However, the high computational cost of maintaining a complex model puts great challenge for model updating. The high computation cost lies in two aspects: computation of complex model coefficients and huge amount training data for it. In this paper, we propose to use a nonparametric approach based on locally weighted learning to predict traffic flow. Our approach incrementally incorporates new data to the model and is computationally efficient, which makes it suitable for online model updating and predicting. In addition, we adopt wavelet analysis to extract the periodic characteristic of the traffic data, which is then used for the input of the prediction model instead of the raw traffic flow data. The primary experiments on real data demonstrate the effectiveness and efficiency of our approach.

Multi-Agent cooperative reinforcement learning in 3d virtual world

A virtual world is an online community in the form of a computer-based simulated environment, through which users can interact with one another and use and create objects The non-player characters (NPC) in virtual world are following a fixed set of pre-programmed behaviors and lack the ability to adapt with the changing surrounding Reinforcement learning agent is a way to deal with this problem However, in a cooperative social environment, NPC should learn not only by trial and error, but also through cooperation by sharing information The key investigation of this paper is: modeling the NPCs as multi-agent, and enable them to conduct cooperative learning, then speeding up the learning process By using a fire fighting scenario in Robocup Rescue, our research shows that sharing information between cooperative agents will outperform independent agents who do not communicate during learning The further work and some important issues of multi-agent reinforcement learning in virtual world will also be discussed in this paper.

A replication strategy based on swarm intelligence in spatial data grid

In the spatial data grid, the distribution of query and the data is unevenly some resource become hotspot and the hotspots are changing over time, which may cause the global load unbalanced, this dynamic problem becomes a key challenge in Data Grid. Data replication is a way to deal with this problem, which improves data availability, reduces latency and increases throughput. In this paper, we present a new replication approach which is adaptive, completely decentralized, and based on swarm intelligence which is intrinsically a bottom-up approach. Every site in the grid system has a single agent, which is serving as containers for data, following simple rules of behavior and without knowing any global information. The strategy that agents follow includes which data to create replica and where the replica is locating. The local interactions and simple action between agents give a fairly optimal replication location solution globally. We carried the experiments using OptorSim for the EU Data Grid Testbed 1. Experimental results show that our approach performs better than No replication and when the scale of jobs is big, our method will outperform the Economic Model, but the space consumption is proportional.

Locally kernel regression adapting with data distribution in prediction of traffic flow

Prognosis of traffic flow is a basic part of intelligent transportation research. Due to the extremely complexity of vehicular traffic, efficient models should be constructed to do accurate simulation and prediction of real traffic, such as locally kernel models. However, locally kernel regression fails when the traffic data points are sparse, and the data distribution should be considered seriously. Moreover, the spatiotemporal features of real traffic make pure locally kernel regression inapplicable. This paper proposes a locally kernel regression mechanism adapting with data distribution for the prediction of traffic flow. This mechanism is also explained by Three-Phase Traffic Theory. Experimental studies show the feasibility and efficiency of our approach.

A spatial clustering algorithm for line objects based on extended Hausdorff distance

Spatial clustering is one of the most commonly used approaches to spatial data mining. This study proposes an algorithm for clustering spatial line objects. Proximity between lines is measured using the extended Hausdorff distance. Lines are clustered using an improved K-means procedure. The procedure defines the kernel line of a cluster and the kernel lines of all clusters are updated in each iteration The algorithm is applied to tropical cyclone tracks of 20 years in the western Northern Pacific. Results show that the algorithm can partition lines to clusters that agree with human cognition.

Agent-based spatial resource management with non-center controller in distributed environment

The resource management among autonomous and heterogeneous spatial database system in a distributed environment is a challenge work. In this paper, we propose an agent framework to manage spatial resource dynamically without a center controller which could have the whole information of each node in the distributed system. The BDI rational engine and the FIPA Contract Net Interaction Protocal can help spatial query agents to dynamically select the appropriate nodes to execute the spatial query efficiently. The agents can adjust the distribution of spatial resource depanding on the history records.