Showing papers by "Haifeng Shen published in 2015"

An Efficient and Reliable Geographic Routing Protocol Based on Partial Network Coding for Underwater Sensor Networks

Abstract: Efficient routing protocols for data packet delivery are crucial to underwater sensor networks (UWSNs). However, communication in UWSNs is a challenging task because of the characteristics of the acoustic channel. Network coding is a promising technique for efficient data packet delivery thanks to the broadcast nature of acoustic channels and the relatively high computation capabilities of the sensor nodes. In this work, we present GPNC, a novel geographic routing protocol for UWSNs that incorporates partial network coding to encode data packets and uses sensor nodes’ location information to greedily forward data packets to sink nodes. GPNC can effectively reduce network delays and retransmissions of redundant packets causing additional network energy consumption. Simulation results show that GPNC can significantly improve network throughput and packet delivery ratio, while reducing energy consumption and network latency when compared with other routing protocols.

Extending Attention Span of ADHD Children through an Eye Tracker Directed Adaptive User Interface

Abstract: Attention Deficit Hyperactivity Disorder (ADHD) is a common neurobiological condition affecting school-age children. One of the main symptoms is short attention span, which is a key factor of poor academic performance, especially in tasks requiring high amount of concentration time. In this position paper, we outline our research plan of extending the attention span of children with ADHD by designing and developing a computer software application that adapts its user interface for incorporating multiple stimuli. The proposed adaptive user interface is directed by an eye tracker, which constantly monitors the user's attention state.

Partial Selection of Agile Software Requirements

Abstract: Prioritization And Selection (PAS) of requirements is an important concern of Agile software development. Existing PAS techniques are primarily binary selection of requirements and postponing lower-priority requirements to the future. As a result, lower-priority requirements may be constantly delayed until they never have a chance to be implemented at all. However, ignoring lower-priority requirements may threaten the quality of software. In this paper, we propose the concept of partial selection as an alternative to binary selection of Agile software requirements rather than ignoring them altogether or postponing them to the future releases. We further contribute a goal-oriented framework that allows requirements to be partially satisfied when tolerated. The framework specifies the fuzzy priority of requirements and accomplishes partial selection through relaxing the satisfaction level of requirements with respect to their priority in the release backlog of software. The experimental results show that partial selection of requirements will reduce the number of neglected requirements when compared to binary selection.

ClasSense: A Mobile Digital Backchannel System for Monitoring Class Morale

Abstract: useful for lecturers to gather real-time online feedback from students in a large lecture environment. Nevertheless, the fast pace of a lecture relative to a large audience can make it difficult for the lecturer to process and respond to large amount of feedback in real time. To address this issue, we propose a mobile digital backchannel system ClasSense that allows students to express their emotions during a class and analyses the sentiments of their feedback in real time so that the lecturer can continuously monitor the morale of the student population and respond to the most important concerns students have in common. This paper presents the key design considerations and a pilot study of the proposed system.

A Lightweight Solution to Version Incompatibility in Service-Oriented Revision Control Systems

Abstract: Service-Oriented Revision Control (SORC) is a novel distributed Software Configuration Management (SCM) model specifically designed for effectively supporting collaborative programming of Web applications, where each developer's code is exposed to their peers as Web services. As developers often work on their services in a continuous and evolutionary manner, newer versions of the same service are constantly created and as such incompatibility between versions becomes a major issue in service versioning that leads to hurdles in the interactions between the service provider and its consumers. In this paper, we propose a lightweight solution to version incompatibility in SORC by allowing the service provider to automatically generate a compatibility message when committing a new version, which is to be retrieved by consumers when updating their local proxies to the service without relying on heavyweight service registries.

Designing a Mobile Digital Backchannel System for Monitoring Sentiments and Emotions in Large Lectures

Abstract: Digital backchannel systems have been proven useful for gathering real-time students' feedback in a large lecture environment. However, it is difficult for lecturers to process and respond to large amount of feedback in real time while they are teaching in a large lecture. To address this issue, we propose a mobile digital backchannel system ClasSense that allows students to give feedback during a class and extract the students' sentiments and emotions in their feedback in real time so that the lecturer can continuously monitor the morale of the student population and respond to the most important concerns students have in common. This position paper presents the key design considerations and outlines our research plan.