Pawel Gburzynski

Load balancing for parallel forwarding

Workload distribution is critical to the performance of network processor based parallel forwarding systems. Scheduling schemes that operate at the packet level, e.g., round-robin, cannot preserve packet-ordering within individual TCP connections. Moreover, these schemes create duplicate information in processor caches and therefore are inefficient in resource utilization. Hashing operates at the flow level and is naturally able to maintain per-connection packet ordering; besides, it does not pollute caches. A pure hash-based system, however, cannot balance processor load in the face of highly skewed flow-size distributions in the Internet; usually, adaptive methods are needed. In this paper, based on measurements of Internet traffic, we examine the sources of load imbalance in hash-based scheduling schemes. We prove that under certain Zipf-like flow-size distributions, hashing alone is not able to balance workload. We introduce a new metric to quantify the effects of adaptive load balancing on overall forwarding performance. To achieve both load balancing and efficient system resource utilization, we propose a scheduling scheme that classifies Internet flows into two categories: the aggressive and the normal, and applies different scheduling policies to the two classes of flows. Compared with most state-of-the-art parallel forwarding schemes, our work exploits flow-level Internet traffic characteristics.

A New Aggregate Local Mobility (ALM) Clustering Algorithm for VANETs

We present a beacon-based clustering algorithm aimed at prolonging the cluster lifetime in VANETs. We use a new aggregate local mobility criterion to decide upon cluster re-organisation. The scheme incorporates a contention method to avoid triggering frequent re-organisations when two clusterheads encounter each other for a short period of time. Simulation results show a significant improvement of cluster lifetime and reduced node state/role changes compared to previous popular clustering algorithms.

Hidden Problems with the Hidden Node Problem

We discuss a few problems introduced by the RTS/CTS mechanism of collision avoidance and focus on the virtual jamming problem, which allows a malicious node to effectively jam a large fragment of a wireless network at a minimum expense of power. We propose a solution to this problem and provide experimental data illustrating the impact of virtual jamming and the effectiveness of our proposed solution

Controlled flooding in wireless ad-hoc networks

We show how flooding can be adopted as a reliable and efficient routing scheme in ad-hoc wireless mobile networks. It turns out that, with the assistance of some tunable heuristics, flooding is not necessarily inferior to sophisticated point-to-point forwarding schemes, at least for some classes of wireless applications. We discuss a reactive broadcast-based ad-hoc routing protocol in which flooding exhibits a tendency to converge to a narrow strip of nodes along the shortest path between source and destination. The width of this strip can be adjusted automatically or by the user, e.g., in response to varying node density and mobility patterns. Finally, we point out a certain deficiency inherent in the IEEE 802.11 family of collision avoidance schemes and show how to fix it to provide better service to broadcast-based routing schemes represented by our variant of controlled flooding.

Software engineering for health education and care delivery systems: The Smart Condo project

Providing affordable, high-quality healthcare to the elderly while enabling them to live independently longer is of critical importance, as this is an increasing and expensive demographic to treat. Sensor-network technologies are essential to developing assisted living environments. In our Smart Condo project, we have deployed a sensor network with a variety of sensor types in an 850 square-foot condominium. The sensor network records a variety of events and environmental parameters and feeds the related data into our web-based system. This system is responsible for inferring higher-order information about the activities of the condos occupant and supporting the visualization of the collected information in a 2D Geographic Information System (GIS) and a 3D virtual world, namely Second Life (SL).

Fighting the spam wars: A remailer approach with restrictive aliasing

We present an effective method of eliminating unsolicited electronic mail (so-called spam) and discuss its publicly accessible prototype implementation. A subscriber to our system is able to obtain an unlimited number of aliases of his/her permanent (protected) E-Mail address to be handed out to parties willing to communicate with the subscriber. It is also possible to set up publishable aliases, which can be used by human correspondents to contact the subscriber, while being useless to harvesting robots and spammers. The validity of an alias can be easily restricted to a specific duration in time, a specific number of received messages, a specific population of senders, and/or in other ways. The system is fully compatible with the existing E-Mail infrastructure and can be immediately accessed via any standard E-Mail client software (MUA). It can be easily deployed at any institution or organization running its private E-Mail server (MTA) with a trivial modification to that server. Our system offers a simple method to salvage the existing population of E-Mail addresses while eliminating all spam aimed at them.

The smart condo: visualizing independent living environments in a virtual world

Providing affordable, high-quality healthcare to the elderly while enabling them to live independently longer is of critical importance. In our Smart Condo project, we have deployed a wireless sensor network in an 850-square-foot condominium for assisted living. The sensor network records a variety of events and environmental parameters and feeds the related data into our web-based system. This system is responsible for inferring higher-order information about the activities of the condos occupant and visualizing the collected information in both a 2D Geographic Information System (GIS) and a 3D virtual world. The architecture is flexible in terms of supported sensor types, analyses, and visualizations through which it communicates this information to its users, including the condos occupant, their family, and their healthcare providers.

Multiple path routing in networks with inaccurate link state information

We study a collection of K-shortest path routing schemes and investigate their performance under a diverse set of network topologies and traffic conditions. We subsequently demonstrate that K-shortest path routing offers a lower blocking probability and more balanced link utilisation than other routing methods. With the proposed approach, it is possible to reduce the frequency of link state exchange, and the incurred bandwidth overhead, without sacrificing the overall performance of the network.

A high fidelity ATM traffic and network simulator

The design of an ATM traffic and network (ATM-TN) simulator which characterizes cell level network behavior is presented. The simulator incorporates three classes of ATM traffic source models: an aggregate ethernet model, an MPEG model and a World Wide Webb transactions model. Six classes of ATM switch architectures are modeled including output buffered, shared memory buffered and cross bar switch models, and then multistage switches which can be built from these three basic models. The ATM-TN simulator can be used to characterize arbitrary ATM networks with dynamic multimedia traffic loads. Call set up and tear down via ATM signaling is implemented in addition to the various types of cell traffic streams generated by voice, video and data. The simulator is built on a simple, efficient simulation language called SimKit which is capable of supporting both fast sequential and parallel execution. Parallel execution is supported using WarpKit, an optimistically synchronized kernel that is aimed at shared memory multiprocessor platforms such as the Silicon Graphics Powerchallenge and Sun Spare 1000 series machines. The paper outlines general requirements for ATM traffic and network simulation, presents an ATM-TN simulator architecture, describes its major components and discusses the major issues associated with cell level ATM modeling and simulation.

A scalable load balancer for forwarding internet traffic: exploiting flow-level burstiness

Packet scheduling in parallel forwarding systems is a hard problem. Two major goals of a scheduler that distributes incoming packets to multiple forwarding engines are to achieve high system utilization (by balancing the load evenly among the multiple engines) and to maintain packet ordering within individual flows. Additionally, from the viewpoint of the overall performance, the system should exhibit a good cache behavior by preserving temporal locality in the workload of each forwarding engine. In this paper, we show how the burstiness in Internet flows can be exploited to improve the performance of the scheduler. Specifically, TCP flows, which contribute to over 90 percent of the Internet traffic, transmit in bursts with relatively large delays in between. We propose a load balancing scheme based on this insight to achieve the scheduling goals. Our design is verified by simulations driven by real-world traces.