Arne A. Nilsson

Delay-based congestion avoidance for TCP

The set of TCP congestion control algorithms associated with TCP/Reno (e.g., slow-start and congestion avoidance) have been crucial to ensuring the stability of the Internet. Algorithms such as TCP/NewReno (which has been deployed) and TCP/Vegas (which has not been deployed) represent incrementally deployable enhancements to TCP as they have been shown to improve a TCP connections throughput without degrading performance to competing flows. Our research focuses on delay-based congestion avoidance algorithms (DCA), like TCP/Vegas, which attempt to utilize the congestion information contained in packet round-trip time (RTT) samples. Through measurement and simulation, we show evidence suggesting that a single deployment of DCA (i.e., a TCP connection enhanced with a DCA algorithm) is not a viable enhancement to TCP over high-speed paths. We define several performance metrics that quantify the level of correlation between packet loss and RTT. Based on our measurement analysis we find that although there is useful congestion information contained within RTT samples, the level of correlation between an increase in RTT and packet loss is not strong enough to allow a TCP/Sender to reliably improve throughput. While DCA is able to reduce the packet loss rate experienced by a connection, in its attempts to avoid packet loss, the algorithm will react unnecessarily to RTT variation that is not associated with packet loss. The result is degraded throughput as compared to a similar flow that does not support DCA.

Architectural description of a new, easily expandable self-routing computer network topology

The topology described applies to any general-purpose computer networking environment. The scheme is named the spiral topology because the architecture is built around modules of four computer nodes each, connected by top and bottom spirals. The spiral technology features a simple internal self-routing algorithm that adapts quickly and automatically to failed nodes or links. The six most important direct consequences of the network architecture are the topologys: (1) ease of expansion; (2) fast, on-the-fly self-routing; (3) extremely high tolerance to network faults; (4) increased network security; (5) potential for the total elimination of store-and-forward transmissions due to routing decision delays; and (6) rendering the maximum path length issue moot.<<ETX>>

Identification of performance degradation in IP networks using throughput statistics

To be able to satisfy their users, interactive applications like video conferences require a certain. Quality-of-Service from heterogeneous networks. This paper proposes the use of throughput histograms as Quality-of-Service indicator. These histograms are built from local, unsynchronized, passive measurements of packet streams from the viewpoint of an application. They can easily be exchanged between sender and receiver, and their comparison provides information about severity and type of a potential bottleneck. We demonstrate the usefulness of these indicators for evaluating the transport quality perceived by a video conferencing application and its users in the presence of a bottleneck.

On Optimal Scheduling Algorithms for Time-Shared Systems

The problem of fmdlng those optimum scheduling algorithms for time-shared systems that mlmmize a cost function that depends on waiting time and required service time IS considered An optimality condmon which sometimes leads to infeasible algorithms is established The procedure is unproved upon by use of a mathematical programming technique but still does not always generate feasible algorithms. These results are used as upper bounds on the performance of known feasible algorithms so that it is possible to evaluate how close to optimal the present algorithms come.

Approximate analysis of product-form type queueing networks with blocking and deadlock

Abstract A numerical procedure for analyzing exactly closed exponential queueing networks with finite queues is presented first. Due to the finiteness of these queues, blocking and deadlock may occur. Deadlocks are assumed to be detected and resolved instantaneously. The numerical procedure is then incorporated in an approximation algorithm for analyzing closed exponential queueing networks of the product-form type, in which some of the queues are finite. These finite queues are assumed to be linked together to form a single subnetwork. The approximation algorithm is based on a variant of Nortons theorem. Comparisons between the approximate results and exact numerical results were carried out and the relative error was observed to be small.

The incremental deployability of RTT-based congestion avoidance for high speed TCP Internet connections

Our research focuses on end-to-end congestion avoidance algorithms that use round trip time (RTT) fluctuations as an indicator of the level of network congestion. The algorithms are referred to as delay-based congestion avoidance or DCA. Due to the economics associated with deploying change within an existing network, we are interested in an incrementally deployable enhancement to the TCP/Reno protocol. For instance, TCP/Vegas, a DCA algorithm, has been proposed as an incremental enhancement. Requiring relatively minor modifications to a TCP sender, TCP/Vegas has been shown to increase end-to-end TCP throughput primarily by avoiding packet loss. We study DCA in todays best effort Internet where IP switches are subject to thousands of TCP flows resulting in congestion with time scales that span orders of magnitude. Our results suggest that RTT-based congestion avoidance may not be reliably incrementally deployed in this environment. Through extensive measurement and simulation, we find that when TCP/DCA (i.e., a TCP/Reno sender that is extended with DCA) is deployed over a high speed Internet path, the flow generally experiences degraded throughput compared to an unmodified TCP/Reno flow. We show (1) that the congestion information contained in RTT samples is not sufficient to predict packet loss reliably and (2) that the congestion avoidance in response to delay increase has minimal impact on the congestion level over the path when the total DCA traffic at the bottleneck consumes less than 10% of the bottleneck bandwidth.

Tier-based analytical model for adaptive call admission control scheme in a UMTS-WCDMA system

UMTS (Universal Mobile Telecommunications System) uses WCDMA (wideband code division multiple access) as its radio interface. The main advantage of this is flexibility in resource management. UMTS systems are based on a hierarchical cellular structure that provides high data rates with extended mobile speeds. Mobility causes resource allocation complications in the concerned cell and the cellular neighborhood. In addition, the heterogeneous nature of users in the UMTS systems requires quality of service (QoS) guarantee for each type of service (ToS). This paper proposes that higher bandwidth utilization can be achieved by managing the resource allocation at a bigger neighborhood level. Higher bandwidth utilization leads to lower blocking and dropping probabilities, which are the main QoS parameters that are addressed in this paper. The prevalent call admission control schemes are used to simulate the UMTS-WCDMA system. The system is first recognized as a multirate, multi-class, priority based system. The tier-based cellular structure is then proposed and analytically modeled.

Population-based call admission control in wireless cellular networks

Wireless systems have been designed as microcellular architectures in order to provide higher capacity. This reduced coverage area of a cell has caused a higher rate of hand-off events, and providing continuous service to a wireless call after hand-off has been an important issue. A simple and efficient bandwidth reservation scheme is proposed, which dynamically adjusts the amount of bandwidth reserved for hand-off calls in a cell, based on the amount of cellular traffic in its neighboring cells. The performance of the proposed scheme is evaluated and compared with the guard channel scheme which is widely used in current cellular networks.

SOLA: lightweight security for access control in IEEE 802.11

The IEEE 802.11 wireless standard provides little support for secure access control. As a result, access control in IEEE 802.11 on a per packet basis requires a new and robust identity authentication protocol. The SOLA (Statistical One-Bit Lightweight Authentication) protocol is well suited in a wireless constrained environment because this protocols communication overhead is extremely low: only one bit. Furthermore, SOLA fulfills the requirements of being secure, useful, cheap, and robust. The synchronization algorithm performs very well. SOLA also makes it easy to develop a framework to detect and respond to, for instance, denial-of-service attacks or an adversary who tries to guess the identity authentication bit for successive packets.

Performance of error control coding techniques for wireless ATM

In order to provide a ubiquitous telecommunications network which merges the concepts of wireless mobile communications with the use of ATM as a transport mechanism, this paper focuses on the problem of reducing the effects of the high bit error rate on the data transmission. We present a simulator which can be used to study the effects of various coding techniques on the overall system bit error rate. Among these coding techniques are convolutional coding, interleaving, fragmentation, and puncturing. We show that convolutional coding, along with interleaving or fragmentation, improves transmission reliability over the bursty error channel.