Hongchi Shi

An intelligent distributed environment for active learning

Active learning is an effective learning approach. In this article we present an intelligent agent-assisted environment for active learning to better support the student-centered, selfpaced, and highly interactive learning approach. The environment uses the students learningrelated profile such as learning style and background knowledge in selecting, organizing, and presenting learning material, and it adopts a new approach to course content organization and delivery based on smart instructional components that can be integrated into a wide range of courses. The environment is being implemented using the prevalent Internet, Web, digital library, and multiagent technologies.

A survey on network protocols for wireless sensor networks

Recent advances in MEMS (micro-electromechanical systems), processor, radio, and memory technologies have dramatically enabled development of wireless sensor networks. A sensor network is a large network of small sensor nodes, capable of sensing, communication, and computation. It can be deployed to sense some physical phenomenon for a wide variety of applications. During recent years, research in wireless sensor networks has become more and more active. Network protocols developed for sensor networks are of great importance to meet specific design goals of sensor networks. We present a survey of recent work addressing network protocols, including routing and information dissemination algorithms, for wireless sensor networks. We evaluate them in terms of design goals, assumptions, operation models, energy models, and performance metrics.

SHARP: a new approach to relative localization in wireless sensor networks

For wireless sensor networks, localization is crucial due to the dynamic nature of deployment. In relative localization, nodes use the distance measurements to estimate their positions relative to some coordinate system. In absolute localization, a few nodes (called anchors) need to know their absolute positions, and all the other nodes are absolutely localized in the coordinate system of the anchors. Relative and absolute localization methods differ in both the performance and the cost. We present a new approach to relative localization that we refer to as: simple hybrid absolute-relative positioning (SHARP). In SHARP, a relative localization method (M1) is used to relatively localize N/sub r/ reference nodes. Then, an absolute localization method (M2) uses these N/sub r/ nodes as anchors to localize the rest of the nodes. Choosing N/sub r/, M1, and M2 gives a wide range of performance-cost tuning. We have done extensive simulation using the multidimensional scaling (MDS) method as M1 and the ad-hoc positioning system (APS) method as M2. While previous research shows that MDS gives better localization results than APS, our simulation shows that SHARP outperforms MDS if both the localization error and the cost are considered.

A sorted RSSI quantization based algorithm for sensor network localization

Range estimation is essential in many sensor network localization algorithms. Although wireless sensor systems usually have available received signal strength indication (RSSI) readings, this information has not been effectively used in the existing localization algorithms. In this paper, we present a sensor network localization algorithm based on a sorted RSSI quantization scheme that can improve the range estimation accuracy when distance information is not available. The range level used in the quantization process can be decided by each node using an adaptive quantization scheme. The new algorithm can be implemented in a distributed way and has significant improvement over existing range-free algorithms. The performance advantage for various sensor networks is shown with experimental results from our extensive simulation with a more realistic radio model.

A general model for bidirectional associative memories

This paper proposes a general model for bidirectional associative memories that associate patterns between the X-space and the Y-space. The general model does not require the usual assumption that the interconnection weight from a neuron in the X-space to a neuron in the Y-space is the same as the one from the Y-space to the X-space. We start by defining a supporting function to measure how well a state supports another state in a general bidirectional associative memory (GBAM). We then use the supporting function to formulate the associative recalling process as a dynamic system, explore its stability and asymptotic stability conditions, and develop an algorithm for learning the asymptotic stability conditions using the Rosenblatt perceptron rule. The effectiveness of the proposed model for recognition of noisy patterns and the performance of the model in terms of storage capacity, attraction, and spurious memories are demonstrated by some outstanding experimental results.

Analysis of Energy Consumption in Clustered Wireless Sensor Networks

Clustering is one of the most important approaches used in wireless sensor networks to save energy. Transmission range affects the clustering and in turn affects the energy consumption in the network. We establish an energy consumption model for clustered wireless sensor networks and solve the optimal transmission range problem. Using this model, the total energy consumption can be estimated beforehand based on the traffic pattern, energy model, and network deployment parameters. This model provides an insight into the energy consumption behavior in clustered wireless sensor networks and the relationship among major factors. The optimal transmission range for energy consumption is a function of the traffic load and the node density, but the effect of node density is very limited

A partial-range-aware localization algorithm for ad-hoc wireless sensor networks

Current wireless sensor network localization algorithms are categorized either as range-free or range-aware algorithms based on whether they use the range (i.e., distance) information. Although wireless sensor systems usually have available received signal strength (RSS) readings, this useful information has not been effectively used in the existing localization algorithms. The existing range-free algorithms do not use this information, and the range-aware algorithms require sophisticated range techniques to estimate ranges. In this paper, we introduce and develop a partial-range-aware localization scheme utilizing RSS readings. The scheme can be implemented in a distributed way and can be used in any hop-based range-free algorithms to improve their localization accuracy. Through extensive simulation, the results show that the partial-range-aware algorithm can improve the localization accuracy by reducing position errors up to 50% of the previous range-free localization results.

An agent enabling personalized learning in e-learning environments

In this work we focus on the potential of intelligent agents in e-learning environments. A central problem is the modeling of human learners so that the agent can facilitate personalized learning. The solution presented here is based on a connectionist approach. An intelligent agent, which replaces the human instructor, controls an e-learning environment and exploits a self-organizing map (SOM) learner behavioral model in order to achieve the learning goal. The learner SOM model has been implemented. The experiments support the argument that e-learning environments are feasible and can significantly assist dissemination of knowledge within modern educational settings.

Parallel machine translation: principles and practice

Parallel machine translation (PMT) is a new machine translation paradigm that takes advantage of high-speed and large-memory computers and existing machine translation systems with different characteristics to solve the difficult machine translation problem. PMT is based on technologies of parallel computing, machine translation, and artificial intelligence. A PMT system consists of many machine translation procedures running in parallel, coordinated by a controller to dissolve various ambiguities in machine translation. We have designed and implemented a PMT system based on the above approach at a coarse parallel level. The system consists of four independent machine translation subsystems. Each subsystem is implemented using an existing machine translation technique and has its own characteristics. We present the principles and practice of PMT. We also present some results of experiments with our experimental PMT system and point out some future research on PMT.

Energy-efficient routing for 2D grid wireless sensor networks

Wireless sensor networks have become possible because of the on-going improvements in smart sensor technology. One important issue in wireless sensor networks is to achieve energy-efficient operation in order to extend network lifetime. For many applications, placement of sensing nodes can be pre-determined. In this paper, we propose several power-efficient routing algorithms for a sensor network with 2D grid topology, demonstrating that optimal routings are operation-specific and energy consumption can be minimized by optimally selecting data paths and base station location.