Lloyd G. Greenwald

Mobile robot labs

There has been much interest in achieving educational and research objectives through the use of small, low-cost robot platforms. While our initial experiences with these platforms were similarly positive, we questioned whether these platforms could be pushed beyond their early uses and transitioned towards achieving substantial educational and research goals. This article reports initial results of this investigation - the construction and implementation of a series of detailed lab assignments using these platforms to tackle basic computer science, AI, robotics, and engineering problems. We first provide detailed descriptions of the labs we have developed and then discuss the robot platforms, including the progression of hardware issues encountered. Finally, we share what we have learned from this endeavor.

Evaluating ad hoc routing protocols with respect to quality of service

The ability of a mobile ad hoc network (MANET) to provide adequate quality of service (QoS) is limited by the ability of the underlying routing protocol to provide consistent behavior despite the inherent dynamics of a mobile computing environment. In this paper we study three MANET routing protocols: OLSR, DSR and AODV, with an emphasis on the effect they have on various QoS metrics. We describe and analyze how the protocols differ in the mechanisms they use to select paths, detect broken links, and buffer messages during periods of link outage. The effects of these differences are quantified in terms of packet delivery ratio, end-to-end hop count, end-to-end latency, and mechanism overhead. We show that the proactive protocol, OLSR, builds paths with consistently lower hop counts than the reactive protocols, AODV and DSR, a fact that leads to a reduction in end-to-end latency that assists a QoS model in meeting timing requirements and improves global network performance. We further show the impact of broken link detection latency on the packet delivery ratio. A routing protocol that can not quickly recover from link breakage caused by mobility renders a QoS model incapable of meeting delivery requirements. Finally, we analyze the effect of mobility on the distribution of end-to-end latencies. Traditionally, reactive protocols are criticized for buffering during the building of routes, however we also study buffering phenomenon caused by the proactive mechanisms of OLSR.

An exploration of statistical models for automated test case generation

In this paper, we develop methods that use logged user data to build models of a web application. Logged user data captures dynamic behavior of an application that can be useful for addressing the challenging problems of testing web applications. Our approach automatically builds statistical models of user sessions and automatically derives test cases from these models. We provide several alternative modeling approaches based on statistical machine learning methods. We investigate the effectiveness of the test suites generated from our methods by performing a preliminary study that evaluates the generated test cases. The results of this study demonstrate that our techniques are able to generate test cases that achieve high coverage and accurately model user behavior. This study provides insights into improving our methods and motivates a larger study with a more diverse set of applications and testing metrics.

Toward the use of local monitoring and network-wide correction to achieve QoS guarantees in mobile ad hoc networks

This paper presents the exploration of novel mechanisms toward providing quality of service (QoS) guarantees in mobile ad hoc networks. We study mechanisms that provide differentiated services to packets of varying priority traffic flows. These mechanisms do not require any central coordination and do not depend on any specific protocols at the physical, MAC, or network layers. Nodes independently monitor the rates of the highest priority flows and signal corrective mechanisms when these rates fall outside of specified local bounds. Triggering conditions for network-wide corrective mechanisms are designed to trade-off rapid reactive response to local QoS violations with control packet overhead. A range of corrective mechanisms are explored that attempt to maintain reactive response while improving total network utilization, including resources consumed by lower priority traffic. We provide simulation results that demonstrate the effectiveness of monitoring, reactive triggering, and basic and advanced corrective mechanisms. We discuss the extension of these novel mechanisms to a complete QoS solution for mobile ad hoc networks.