Monitoring access link capacity using TFRC probe

Abstract

Accurate estimation of network characteristics based on end-to-end measurements is an important and challenging problem. With the increasing variety and sophistication of Internet access methods, the mobility of the users and the need for seamless handoff across them, monitoring access link capacity becomes critical for efficient multimedia delivery. In fact, this allows the multimedia server to properly adjust its sending rate to the rapidly changing access speed of the mobile client. This problem is challenging because most of the existing capacity monitoring schemes are active, like Pathrate (and therefore, introduce extra overhead on the bandwidth limited access link). Moreover, the access link is often asymmetric (e.g. ADSL, 1×RTT, etc.) thus preventing the use of round-trip monitoring schemes such as Pathchar and CapProbe. In this study, we propose and evaluate a passive, one-way link capacity monitoring tool called TFRC Probe. With TFRC Probe the source can monitor the forward direction capacity of both asymmetric and symmetric access links, and can rapidly and accurately adapt its transmissions rate accordingly. We validate TFRC Probe with both simulation and testbed experiments, and show that TFRC Probe is a very flexible tool that can accurately track frequent changes in access capacity.

Introduction

Knowledge of link capacity is particularly important for network management, pricing, and QoS support, especially in emerging technologies such as in overlay, P2P, sensor and grid networks. With the emerging complexity of wireless network connection technologies, the link capacity of a network connection may vary dramatically due a variety of factors, such as vertical handoff (e.g. a handoff between 802.11 b and 1×RTT technology), dynamic channel allocation (e.g. 1×RTT and GPRS), and wireless channel quality (e.g. 802.11 b). For these sophisticated settings, knowing the link capacity will permit the source to rapidly and accurately adapt the outbound data transmissions rate. Therefore, it is of increasing interest nowadays to achieve an accurate and ‘online’ monitoring of link capacity.

The basic idea of using monitoring probes to estimate link capacity can be achieved through either active or passive measurement. Active measurement is a common approach that injects measurement (probe) packets into the network. The obvious drawbacks of active measurement have motivated the development of passive measurement techniques, which aimed to detect network properties without disturbing on-going network services (via traffic already flowing through the network). Monitoring probes are also required to reflect changes in link capacity (e.g. a handoff between 802.11 b and 1×RTT technology) in a timely manner, so a system can adjust its data transmission properties based the most ‘up-to-date’ information. Moreover, to enable seamless network protocols functionalities and other management operations, monitoring probes would have to maintain end-to-end properties. Summarizing the design goals discussed above, it is apparent that an ideal network monitoring technique should (a) provide correct information, (b) work passively without adding excess overhead to the networks, (c) promptly react to occurrences in network events and (d) maintain end-to-end semantics.

Addressing all of the above requirements for capacity estimation and monitoring, we propose TFRC Probe, an on-line capacity monitoring technique achieved through embedding the CapProbe algorithm [1] within the TFRC protocol [2]. Different from the round-trip nature of CapProbe algorithm, we specifically designed TFRC Probe to monitor the link capacity of the forward direction link only. This is based on the realization that capacity information on the forward direction link conveys critical information for any data transferring operations involving asymmetric links. Since, information traffic on asymmetric links such ADSL are usually ten times more intensive on forward direction link (download) as oppose to reverse direction link (upload) (e.g. video streaming and file downloading), the ability to appropriately establish the upper-bound of servers' sending rate can provide much assistance in regulating the quality/speed/smoothness of data delivery services. For instance, a previous study has shown that TFRC is slow in responding to a drastic capacity increase [3], and has indicated that a fast rate adaptation algorithm can significantly improve multimedia delivery (for example, by adjusting source rate, content and format) [4]. In this study, TFRC Probe is validated with both simulations and testbed experiments, and has been proven to be quite an effective tool in providing accurate capacity estimation and monitoring.

The rest of the paper is organized as follows. In Section 2, we present work related to this study. In Section 3, we briefly describe TFRC protocol and detailed the concepts of TFRC Probe. In Section 4, we evaluate the accuracy of TFRC probe in estimating link capacity through series of NS2 simulations. In Section 5, we present results from our testbed experiments to validate the monitoring capability of TFRC Probe in Internet and wireless networks. In Section 6, we present an application of using TFRC Probe in vertical handoff scenarios, where the sending rate of TFRC Probe adapts accordingly to capacity monitoring results. Section 7 concludes the paper.