Cintia B. Margi

Characterizing energy consumption in a visual sensor network testbed

In this work we characterize the energy consumption of a visual sensor network testbed. Each node in the testbed consists of a single-board computer, namely Crossbows Stargate, equipped with a wireless network card and a Webcam. We assess energy consumption of activities representative of the target application (e.g., perimeter surveillance) using a benchmark that runs (individual and combinations of) basic tasks such as processing, flash memory access, image acquisition, and communication over the network. In our characterization, we consider the various hardware states the system switches through as it executes these benchmarks, e.g., different radio modes (sleep, idle, transmission, reception), and Webcam modes (off, on, and acquiring image). We report both steady-state and transient energy consumption behavior obtained by direct measurements of current with a digital multimeter. We validate our measurements against results obtained using the Stargates on-board energy consumption measuring capabilities

Modeling energy consumption in single-hop IEEE 802.11 ad hoc networks

This paper presents an analytical model to predict energy consumption in saturated IEEE 802.11 single-hop ad hoc networks under ideal channel conditions. The model we introduce takes into account the different operational modes of the IEEE 802.11 DCF MAC, and is validated against packet-level simulations. In contrast to previous works that attempted to characterize the energy consumption of IEEE 802.11 cards in isolated, contention-free channels (i.e., single sender/receiver pair), this paper investigates the extreme opposite case, i.e., when nodes need to contend for channel access under saturation conditions. In such scenarios, our main findings include: (1) contrary to what most previous results indicate, the radios transmit mode has marginal impact on overall energy consumption, while other modes (receive, idle, etc.) are responsible for most of the energy consumed; (2) the energy cost to transmit useful data increases almost linearly with the network size; and (3) transmitting large payloads is more energy efficient under saturation conditions

Instrumenting network simulators for evaluating energy consumption in power-aware ad-hoc network protocols

In this paper, we describe our work on instrumenting network simulators to enable them to adequately and accurately account for the energy consumed by ad hoc network protocols communication-related tasks. This is accomplished by explicitly accounting for low-power radio modes and considering the different energy costs associated with each possible radio state, i.e, transmitting, receiving, overhearing, idle, sensing, and sleeping. Our energy consumption instrumentation also allows the energy accounting to be done automatically by the simulator irrespective of what layer of the stack the protocol designer is working. To validate our model, we compare: (1) simulation results using the GloMoSim/QualNet simulation platform with and without our instrumentation for the IEEE 802.11 DCF; (2) analytical results for both 802.11 and S-MAC (a power-aware MAC designed for sensor networks); and (3) simulation results reproducing testbed experiments obtained for the S-MAC protocol. Finally, by comparing S-MAC against 802.11 and AODV against DSR, we showcase the ability of a network simulation platform instrumented with our energy consumption model to evaluate energy consumption in ad-hoc network protocols.

Characterizing system level energy consumption in mobile computing platforms

This paper approaches energy consumption characterization in mobile computing platforms by assessing energy consumption of basic application-level tasks, such as processing, input/output (disk, display, etc.), communication (transmission and reception over the network), and combinations thereof. Besides providing information on the energy consumption behavior of typical tasks performed by mobile computers, task-level energy characterization enables power management decisions, such as whether, in a distributed computation, the task at hand can be executed locally or should be assigned to a different machine (given the machines current energy budget, the energy cost of executing the task locally, and the cost of sending the required information over the network to a peer). We employ a task-level energy consumption characterization benchmark that accounts for basic tasks such as processing, disk access (including reads and writes), terminal usage, and communication (transmission and reception). Using the benchmark, we perform an energy characterization case study using the Dell Latitude C600 running two versions of the Linux operating system.