The Tor anonymous communication network uses self-reported bandwidth values to select routers for building tunnels. Since tunnels are allocated in proportion to this bandwidth, this allows a malicious router operator to attract tunnels for compromise. Although Tor limits the self-reported bandwidth, it uses a high maximum value, effectively choosing performance over high anonymity for all users. We propose a router selection algorithm that allows users to control the trade-off between performance and anonymity. We also propose an opportunistic bandwidth measurement algorithm to replace self-reported values that is more sensitive to load and more responsive to changing network conditions. Our mechanism effectively blends the traffic from users of different preferences, making partitioning attacks difficult. We implemented the opportunistic measurement and tunable performance extensions and examined their performance both through simulation and in the real Tor network. Our results show that users can get dramatic increases in either performance or anonymity with little to no sacrifice in the other metric, or a more modest improvement in both. Our mechanisms are also invulnerable to the previously published low-resource attacks on Tor.

As mobile systems become increasing prevalent, they are being called upon to run multimedia applications traditionally associated with the wired domain. Unfortunately, battery capacity is not keeping pace with system demand, driving the demand for energy efficiency. Unfortunately, the encapsulation of traditional layered designs limits the energy that can be saved because layers cannot cooperate. However, monolithic systems have been shown to be inefficient and unscalable. In this paper, we present the design and implementation of a system consisting of multiple adaptive applications, and adaptive MAC and transport layers. We evaluate this system by implementing an adaptive video encoding application, showing that a layered approach sharing enough information can enable energy efficiency while maintaining sufficient isolation to allow the replacement of components without requiring a whole-system redesign.

Multimedia applications have unique characteristics that can be leveraged to design energy-efficient loss recovery mechanisms. Given their loss tolerance and strict timing requirements, various frame dropping policies can be used to affect both application performance and energy consumption. However, the effects on energy consumption of such frame dropping policies depends on not only high-layer effects, but also the MAC layer underlying them. Therefore, we first present an analysis of an energy-efficient MAC-layer protocol, called Fast Transmit MAC Protocol (FTMP). We then explore three frame dropping policies: proactive, reactive, and opportunistic. We present the EELR (energy-efficient loss recovery) protocol, which uniquely combines all three frame dropping policies to provide both high energy efficiency and high application performance. EELR is implemented in a real system on top of an energy-saving MAC layer that exposes information about transmission costs to the transport layer for use in frame dropping decisions. Through analysis and real-world experiments, we show that the combination of an energy-efficient MAC layer with a transport protocol using all three frame-drop policies outperforms any other combination.

One key issue in wireless sensor networks is energy efficiency. Aggregation techniques attempt to minimize the energy expended in communication by processing the data in the network rather than forwarding it all to a central point for processing. However, the naive application of this technique can result in a net increase in total energy expenditure due to the computational costs of the processing. In this paper, we present a new model for aggregation in wireless sensor networks, incorporating both the amount of data reduction achieved by the aggregation function and its associated computation costs. We then use this model to create a powerful, lightweight, distributed aggregation tree creation protocol, called Tethys, suitable for implementation in wireless sensor networks. We show that our protocol provides significantly lower energy consumption than protocols which do not take the amount of data reduction and computational cost of aggregation into consideration.