This thesis investigates distributed router architectures and IP networks with centralized control. While the current trend in IP-router architectures is towards decentralized design, there have also been research proposals for centralizing the control functions in IP networks. With continuous evolution of routers and IP networks, we believe that eventually IP networks in an autonomous system (AS) and a distributed router might converge into one network system. This system, which can be considered both as a distributed router and a centrally-controlled IP network, is divided into a control plane and a forwarding plane. The control plane is responsible for routing, management and signalling protocols, while the forwarding plane is responsible for forwarding packets. The work in this thesis covers both the forwarding and control planes. In the forwarding plane, we study network processor systems that function as forwarding elements in a distributed router. We introduce a system model and a simulation tool based on the model. Using the simulation tool, we investigate network processor system design by studying throughput, utilization, queueing behavior and packet delays. In addition to network processor systems, we study IP-address lookup, which is one of the key packet processing functions in Internet routers. Our work in IP-address lookup contains an efficient lookup algorithm, a scheme to divide the lookup procedure into two-stages in a distributed router, and an approach to perform efficient lookup on a router supporting multiple virtual routers. In the control plane, we study three emerging research issues with centralized control. We provide a thorough study of the routing convergence process in networks with centralized control, and compare it with decentralized link-state routing protocols. We propose an efficient approach to perform traffic engineering and routing in networks with centralized control, and compare it with an approach using optimized link weights. Finally, we present an approach to perform loop-free updates of forwarding tables when the forwarding paths change. This loop-free update approach is particularly useful in networks with centralized control. The results presented in this thesis are useful for building next-generation routers and IP networks with centralized control that might eventually converge into one network system. / QC 20100726

Not available / text

International Telemetering Conference Proceedings / October 18-21, 2004 / Town & Country Resort, San Diego, California / Sensor networks can demand large amounts of bandwidth if the raw sensor data is transferred to a central location. Feature recognition and sensor fusion algorithms can reduce this bandwidth. Unfortunately the designers of the system, having not yet seen the data which will be collected, may not know which algorithms should be used at the time the system is first installed. This paper describes a flexible architecture which allows the deployment of data reduction algorithms throughout the network while the system is in service. The network of sensors approach not only allows for signal processing to be pushed closer to the sensor, but helps accommodate extensions to the system in a very efficient and structured manner.

The ZigBee standard is a wireless networking standard created and maintained by the ZigBee Alliance. The standard aims to provide an inexpensive, reliable, and efficient solution for wirelessly networked sensing and control products. The ZigBee Alliance is composed of over 300 member companies making use of the standard in different ways, ranging from energy management and efficiency, to RF remote controls, to health care products. Home automation is one market that greatly benefits from the use of ZigBee. With a focus on conserving home electricity use, a sample design is created to test a home automation network using Freescale's ZigBee platform. Multiple electrical designs are tested utilizing sensors ranging from proximity sensors to current sense transformers. Software is fashioned as well, creating a PC application that interacts with two ZigBee transceiver boards performing different home automation functions such as air conditioner and automatic lighting control. / by Michael G. DiBenedetto. / Thesis (M.S.C.S.)--Florida Atlantic University, 2009. / Includes bibliography. / Electronic reproduction. Boca Raton, Fla., 2009. Mode of access: World Wide Web.

Recursive Inter-Network Architecture (RINA) networks have a shorter protocol stack than the current architecture (the Internet) and rely instead upon separation of mech- anism from policy and recursive deployment to achieve large scale networks. Due to this smaller protocol stack, fewer networking mechanisms, security or otherwise, should be needed to secure RINA networks. This thesis examines the security proto- cols included in the Internet Protocol Suite that are commonly deployed on existing networks and shows that because of the design principles of the current architecture, these protocols are forced to include many redundant non-security mechanisms and that as a consequence, RINA networks can deliver the same security services with substantially less complexity.

<p>We present an indirection architecture for the Internet called SelNet. SelNet provides a uniform indirection mechanism for controlling the route that packets take through the network and which functions are invoked to process these packets. In the current Internet, at least for the majority of users, there is only one way that a packet can go and that is to the default route. Whilst this is sufficient for many applications, numerous applications have arisen which require alternative routes or processing to be present not only at the application-layer of the protocol stack, but at the network-layer itself. Solutions to such scenarios attempt to place an indirection point between the communicating end-systems either with a middlebox (such as a proxy) or by altering one or more of the Internet's naming systems. However these approaches lead to an application-specific network, which is against the Internet's design goals. We argue for a uniform approach to indirection instead of building multiple, partially overlapping structures as is the current trend. SelNet differs from existing indirection approaches in that it is function-orientated, rather than node-orientated and that it provides an explicit, controllable resolution mechanism for resolving host names and services. The motivation behind our approach is to create efficient indirection structures for supporting new applications which have indirection requirements. We present a detailed design and specification of SelNet. We then go on to describe implementation work with the LUNAR ad-hoc routing protocol and the Janus middleware for accessing sensor networks systems. The purpose of this implementation work is to demonstrate the feasibility of SelNet and its ability to reach its goals.</p>

Due to the advances in content distribution and data center technologies, the traffic inside metropolitan area network (MAN) becomes more and more distributed. The current MAN architecture that only plays the role of bridge becomes insufficient to handle such new traffic patterns. Hence, the demand for a new MAN architecture is inevitable. It is expected to be both efficient and cost-effective. In this thesis, we proposed Optical Buffer Ring (OBR) as the solution of next generation MAN. It combines both OBS's low end-to-end delay and RPR's low loss rate. And compared to more advanced network architectures based on optical packet switching, OBR is of lower cost and thus more feasible in the near future. To evaluate the performance of OBR, we conducted simulation study over large set of parameters. According to the results, the performance of OBR indeed coincides with our expectation. Furthermore, OBR scales better than both OBS and RPR in that the end-to-end delay of OBR decreases as network size increases while the loss rate increases little with network size. Index Terms¡ÐMAN, WDM, OBS, OBR, RPR, HORNET, LightRing, Diffserv, optical packet switching.

We present an indirection architecture for the Internet called SelNet. SelNet provides a uniform indirection mechanism for controlling the route that packets take through the network and which functions are invoked to process these packets. In the current Internet, at least for the majority of users, there is only one way that a packet can go and that is to the default route. Whilst this is sufficient for many applications, numerous applications have arisen which require alternative routes or processing to be present not only at the application-layer of the protocol stack, but at the network-layer itself. Solutions to such scenarios attempt to place an indirection point between the communicating end-systems either with a middlebox (such as a proxy) or by altering one or more of the Internet's naming systems. However these approaches lead to an application-specific network, which is against the Internet's design goals. We argue for a uniform approach to indirection instead of building multiple, partially overlapping structures as is the current trend. SelNet differs from existing indirection approaches in that it is function-orientated, rather than node-orientated and that it provides an explicit, controllable resolution mechanism for resolving host names and services. The motivation behind our approach is to create efficient indirection structures for supporting new applications which have indirection requirements. We present a detailed design and specification of SelNet. We then go on to describe implementation work with the LUNAR ad-hoc routing protocol and the Janus middleware for accessing sensor networks systems. The purpose of this implementation work is to demonstrate the feasibility of SelNet and its ability to reach its goals.

This thesis presents a scalable service-oriented architecture for the demand-driven deployment of location-neutral software services, using an end-to-end or ‘holistic’ approach to address identified shortcomings of the traditional Web Services model. The architecture presents a multi-endpoint Web Service environment which abstracts over Web Service location and technology and enables the dynamic provision of highly-available Web Services. The model describes mechanisms which provide a framework within which Web Services can be reliably addressed, bound to, and utilized, at any time and from any location. The presented model eases the task of providing a Web Service by consuming deployment and management tasks. It eases the development of consumer agent applications by letting developers program against what a service does, not where it is or whether it is currently deployed. It extends the platform-independent ethos of Web Services by providing deployment mechanisms which can be used independent of implementation and deployment technologies. Crucially, it maintains the Web Service goal of universal interoperability, preserving each actor’s view upon the system so that existing Service Consumers and Service Providers can participate without any modifications to provider agent or consumer agent application code. Lastly, the model aims to enable the efficient consumption of hosting resources by providing mechanisms to dynamically apply and reclaim resources based upon measured consumer demand.

Human health is increasingly threatened by the emergence of multiply drug resistant malignant organisms. Yet, our understanding of the numerous ways by which such resistance arises is modest. Here, we present evidence of a bet hedging strategy in the budding yeast, Saccharomyces cerevisiae, to counter the effects of cytotoxic drugs through the action of Pdr5, an ATP-binding cassette transporter. We have employed flow cytometry and fluorescent activated cell sorting to probe the expression levels of a GFP-tagged version of PDR5 in individual cells. The results obtained from these experiments demonstrate that each yeast population is variable in the levels of Pdr5 production, and a small subpopulation of cells produces this efflux pump at much higher quantities than the population average. Consequently, cells with high and low levels of Pdr5 grow differentially in presence and absence of cycloheximide, a cytotoxic drug. These properties are highly suggestive of a bet hedging strategy mediated by Pdr5 levels. We further link this bet hedging strategy to the transcriptional regulatory network of PDR5 consisting of two major transcription factors, Pdr1 and Pdr3. Our analysis suggests that a self-activating feedback loop acting on Pdr3 plays an important role in generation of the aforementioned subpopulation. Furthermore, our results point to a large difference in the activity of these two regulators wherein Pdr3 is notably stronger than Pdr1. The disparity in their activity could indicate a mechanism for generation of the observed proportions of subpopulations with regards to the level of Pdr5.