Niels Möller

Making retransmission delays in wireless links friendlier to TCP

Heterogeneous communication networks with their variety of application demands, uncertain time-varying traffic load, and mixture of wired and wireless links pose several challenging problem in modeling and control. In this paper we focus on the packet delay, which is an important variable for efficient end-to-end congestion control. In particular, we study the delay effects of radio links which use power control and link-layer retransmissions. Link-layer retransmissions induce delays which do not conform to the assumptions on which the transport protocol is based. This causes undesired TCP control actions which reduce throughput. A link-layer solution based on adding carefully selected delays to certain packets is proposed to counteract this problem. All information needed for this is available locally at the link.

Using Radio Network Feedback to Improve TCP Performance over Cellular Networks

A new control structure is proposed to improve user experience of wireless Internet. Information on radio bandwidth and queue length available in the radio network controller (RNC), close to the base station, is used in a proxy that resides between the Internet and the cellular system. The control algorithm in the proxy sets the window size according to event-triggered information on radio bandwidth changes and time-triggered information on the queue length of the RNC. A stability analysis of the hybrid control system is performed. In a number of realistic use cases evaluated in ns-2, the new control structure is compared to a nominal one that employs end-to-end TCP Reno. It is shown that the proposed solution reduces the time to serve users, increases the radio link utilization, and decreases the required buffer size in the RNC.

ACK-clock Dynamics in Network Congestion Control - An Inner Feedback Loop with Implications on Inelastic Flow Impact

The focus of this paper is the window mechanism in network congestion control, whereby packet acknowledgments control when new packets are being sent. This constitutes an inner control loop that so far has received little attention. We provide a novel model of this loop that bridges between the standard integrator link model and the more recent static link model. The model is in validation experiments shown to be accurate. It is also shown that as the amount of inelastic cross-traffic increases the dynamics of this inner loop becomes slower. This may influence overall performance and stability in scenarios with heavy inelastic flows

Performance Analysis and Optimization of TCP over Adaptive Wireless Links

This paper proposes an analytical framework for performance evaluation of TCP (transport control protocol) over adaptive wireless links. Specifically, we include adaptation of power, modulation format and error recovery strategy, and incorporate some features of wireless fading channels. This framework is then used to pursue joint optimization through maximization of an objective functional, that expresses a trade-off between achievable throughput and energy costs. A set of numerical results is reported, and it is seen that hybrid ARQ schemes may provide significant benefits in the optimization framework

Influence of Power Control and Link-Level Retransmissions on Wireless TCP

A fundamental assumption of the tcp protocol is that packet losses indicate congestion on the network. This is a problem when using tcp over wireless links, because a noisy radio transmission may erroneously indicate congestion and thereby reduce the tcp sending rate.

Supporting End-to-End Applications Over HSDPA by Cross-Layer Signalling

A new control structure is proposed to improve user experience of wireless Internet. Information on radio bandwidth and queue length available in the radio network, close to the base station, is used in a proxy that resides between the Internet and the cellular system. The control algorithm in the proxy sets the window size according to event-triggered information on radio bandwidth changes and time-triggered information on the queue length at the wireless link. The mechanism is compared to TCP Reno in two simulation scenarios. The first scenario models a dedicated channel with stepwise changes in the bandwidth, while the second scenario models the High-speed Downlink Shared Channel recently introduced by 3 GPP. The proposed mechanism significantly reduces the amount of buffer space needed in the radio network, and it also gives modest improvements to user response time and link utilization. Reduced buffering is particularly beneficial for third-party end-to-end real-time services such as voice, video, and online gaming.

Inter-protocol fairness between TCP New Reno and TCP Westwood

In this paper we investigate the effect of introducing TCP Westwood+ on regular TCP New Reno. By means of analytical modeling and ns-2 simulations, we demonstrate that the two protocols get different shares of the available bandwidth in the network. Our main result is that the bandwidth sharing between the two protocols depends on one crucial parameter: the ratio between the bottleneck router buffer size and the bandwidth delay product. If the ratio is smaller than one, TCP Westwood+ takes more bandwidth. On the contrary, if the ratio is greater than one, it is TCP New Reno which gets the larger part. Inspired by our results, we propose a simple modification to the window decrease algorithm in TCP Westwood+ that solves the unfairness problem for large buffer sizes. For small buffers, the unfairness problem is still open.

MODELING AND CONTROL OF IP TRANSPORT IN CELLULAR RADIO LINKS

A fundamental assumption of the TCP protocol is that packet losses indicate congestion on the network. This is a problem when using TCP over wireless links, because a noisy radio transmission may e ...

Congestion control for small queues: Analysis and evaluation of a new protocol

A new congestion control protocol is presented, analyzed and experimentally evaluated. It consists of the standard inner-loop ACK-clock and a novel outer-loop adjusting the window size based on congestion signaling from the network. The aim of the new protocol is to maintain the efficiency and fairness properties of TCP, but with significantly smaller bottleneck queues and thereby it takes the sharing with real-time traffic into account. Stability properties of the protocol is proved using a recent fluid-flow traffic model. Experimental comparisons with New Reno and Vegas illustrate the advantages of the new protocol with respect to throughput, delay, utilization, and fairness.

Robust HGCD with no backup steps

Subquadratic divide-and-conquer algorithms for computing the greatest common divisor have been studied for a couple of decades. The integer case has been notoriously difficult, with the need for “backup steps” in various forms. This paper explains why backup steps are necessary for algorithms based directly on the quotient sequence, and proposes a robustness criterion that can be used to construct a “half-gcd” algorithm without any backup steps.