Per Gunningberg

A new end-to-end probing and analysis method for estimating bandwidth bottlenecks

We present a network friendly bandwidth measurement method, TOPP, that is based on active probing and includes analysis by segmented regression. This method can estimate two complementing available bandwidth metrics in addition to the link bandwidth of the congested link. Contrary to traditional packet pair estimates of the bottleneck link bandwidth, our estimate is not limited by the rate at which we can inject probe packets into the network. We also show that our method is able to detect bottlenecks that are invisible to methods such as the C-probe. Further more, we describe scenarios where our analysis method is able to calculate bandwidth estimates for several congested hops based on a single end-to-end probe session.

A testbed and methodology for experimental evaluation of wireless mobile ad hoc networks

Wireless mobile ad hoc network experimentation is subjected to stochastic factors from the radio environment and node mobility. To achieve test repeatability and result re-producibility such stochastic factors need to be controlled or assessed in order to obtain conclusive results. This has implications on the design of testbeds. We present a methodology that addresses repeatability and describe how it has guided us in the design of our Ad hoc Protocol Evaluation (APE) testbed. Finally, by using APE, we present side-by-side routing protocol comparison results and show a radio phenomena that is not visible in simulations.

SoNIC: classifying interference in 802.15.4 sensor networks

Sensor networks that operate in the unlicensed 2.4 GHz frequency band suffer cross-technology radio interference from a variety of devices, e.g., Bluetooth headsets, laptops using WiFi, or microwave ovens. Such interference has been shown to significantly degrade network performance. We present SoNIC, a system that enables resource-limited sensor nodes to detect the type of interference they are exposed to and select an appropriate mitigation strategy. The key insight underlying SoNIC is that different interferers disrupt individual 802.15.4 packets in characteristic ways that can be detected by sensor nodes. In contrast to existing approaches to interference detection, SoNIC does not rely on active spectrum sampling or additional hardware, making it lightweight and energy-efficient. In an office environment with multiple interferers, a sensor node running SoNIC correctly detects the predominant interferer 87% of the time. To show how sensor networks can benefit from SoNIC, we add it to a mobile sink application to improve the applications packet reception ratio under interference.

Global source mobility in the content-centric networking architecture

The Content-Centric Networking (CCN) architecture, a clean-slate network design, borrows its routing concepts from IP. If content is located on mobile sources, CCN also inherits some of the mobility problems known from IP. In this paper, we explore the design space of CCN mobility solutions by revisiting well-known IP approaches that aim to solve a remarkably similar problem. While mobility solutions may be quite similar in both architectures, we find that a locator/identifier split should be implemented at the network layer in CCN to prevent temporary, topology-dependent information to leak into content that ought to be permanent. Mobility handling further benefits from CCNs security model and multipath forwarding. To provide a starting point for further research, we present a simple mobility approach based on an explicit locator/identifier split.

MARCH: a distributed content adaptation architecture

A novel, server-centric architecture for adapting media content to suit the operational environment for heterogeneous devices and networks is presented. The given architecture, so called MARCH, exhibits several advantages over traditional static proxy solutions. The viability of the MARCH framework has been demonstrated through a prototype implementation.

Adaptive resource-based Web server admission control

Web servers must be protected from overload since server overload leads to low server throughput and increased response times experienced by the clients. Server overload occurs when one or more server resources are overutilized. We present an adaptive architecture that performs admission control based on the expected resource consumption of requests. By dynamically setting the maximum rate of accepted requests, we avoid overutilization of the critical resources. We also provide mechanisms for service differentiation. We present our admission control architecture and experiments that show that it can sustain low response times and high throughput for premium clients even during high load.

Distributed multimedia applications on gigabit networks

MultiG, an open research program addressing issues that range from end-user requirements on distributed multimedia applications supporting collaborative work to medium-access protocols for multi-gigabit networks on optical fibers, is discussed. The projects in MultiG are described from the top down, beginning with computer-supported cooperative work. Distributed multimedia applications and application generators are then considered, followed by networking issues, including interprocess communication, transport services, network protocols, and high-speed protocol processing. Testbed activities and future plans are described.<<ETX>>

Application protocols and performance benchmarks

The authors present results from performance measurements of application services. They have chosen widely available implementations of the file transfer protocols File Transfer Access and Management (FTAM) and FTP, and the transaction protocols Remote Operations Service Element (ROSE) and Suns Remote Procedure Call (SunRPC). The measurements have all been run on the same Sun 3/60 workstations over Ethernet, using protocol-independent benchmarks. A tool called SICS Protocol Implementation Measurement System (SPIMS) is used for the measurements. It is structured as a protocol, with two peer entities communicating with each other using the service of the measured protocol. The authors describe how SPIMS can be used to measure response time, throughput, and the time to open and close connections. They describe the protocols and their implementations, and draw some conclusions, and present and comment on the results.<<ETX>>

Repeatable Experiments with Mobile Nodes in a Relocatable WSN Testbed

We present Sensei-UU, a testbed that supports mobile sensor nodes. The design objectives are to provide wireless sensor network (WSN) experiments with repeatable mobility and to be able to use the same testbed at different locations, including the target location. The testbed is inexpensive, expandable, relocatable and it is possible to reproduce it by other researchers. Mobile sensor nodes are carried by robots that use floor markings for navigation and localization. The testbed is typically used to evaluate WSN applications when sensor nodes move in meters rather than millimeters, eg. when human carries a mobile data sink (mobile phone) collecting data while passing fixed sensor nodes. To investigate the repeatability of robot movements, we have measured the achieved precision and timing of the robots. This precision is of importance to ensure the same radio link characteristics from one protocol experiment to another. We find that our robot localization is accurate to ±1 cm and variations in link characteristics are acceptably low to capture fading phenomena in IEEE 802.15.4. In the paper we show repeatable experiment results from three environments, two university corridors and from an anechoic chamber. We conclude that the testbed is relocatable between different environments and that the precision is good enough to capture fading effects in a repeatable way.

Experiences from measuring human mobility using Bluetooth inquiring devices

We present an analysis of human mobility measurements using Bluetooth devices. A number of data traces from such measurements have been made publicly available for the benefit of the research community. However, the limitations of the measurement approaches are in general not well known. We have been given access to the Intel Motes and to the software that was used to perform some of the previous measurements. We present a detailed description of this platform, an analysis of the software used, and provide information on the improvements we have made to remove a number of limitations in the software. To illustrate the impact of our changes, we provide results from two new data collection efforts and compare them to the previous experiments. One of our major results is that our new software makes significant improvements in the number of contacts that can be sampled during each Bluetooth scan. We discuss the impact of these findings.