SensorBench

9
SensorBench: Benchmarking Approaches to Processing Wireless Sensor Network Data Alasdair J G Gray [email protected] alasdairjggray.co.uk @gray_alasdair

description

SensorBench is a benchmark suite for wireless sensor networks. The design of wireless sensor network systems sits within a multi-dimensional design space, where it can be difficult to understand the implications of specific decisions and to identify optimal solutions. SensorBench enables the systematic analysis and comparison of different techniques and platforms, enabling both development and user communities to make well informed choices. The benchmark identifies key variables and performance metrics, and specifies experiments that explore how different types of task perform under different metrics for the controlled variables. The benchmark is demonstrated by its application on representative platforms. Full details of the benchmark are available from http://dl.acm.org/citation.cfm?id=2618252 (DOI: 10.1145/2618243.2618252)

Transcript of SensorBench

Page 1: SensorBench

SensorBench: Benchmarking Approaches to Processing Wireless Sensor Network Data

Alasdair J G [email protected]@gray_alasdair

Page 2: SensorBench

Example WSN Applications

Page 3: SensorBench

WSN Frameworks

How do we

compare different

systems?

Page 4: SensorBench

4

Resource Concerns Energy

Running off battery Computation Capabilities

Limited CPU Limited memory Limited storage

Radio Transmission Limited range Energy impact Lost transmissions

18 August 2014 SICSA CENSIS Workshop

Page 5: SensorBench

5

SensorBench WSN Platform-agnostic benchmark

Identifies Variables Characterises Tasks Measures Performance Metrics

Simulation of platforms Enables

Cross-platform comparison Accurate predictions

Available fromhttp://code.google.com/p/sensorbench

18 August 2014 SICSA CENSIS Workshop

Page 6: SensorBench

8

Benchmark Experiments

1. Impact of network size

2. Impact of node layout

3. Impact of node density

4. Impact of acquisition interval

5. Impact of proportion of source nodes

6. Impact of packet loss

7. Impact of task

18 August 2014 SICSA CENSIS Workshop

Page 7: SensorBench

9

7: Impact of Task

Variable Value

Tasks {Select, aggregation, join, LR, Outlier}

Acquisition Intervals (s)

{1, 2, 4, 8, 16, 32, 64, 128}

Network Size 25

Node Layout irregular

Node Density 3

Proportion of Sources 80

Radio Loss Rate 0

18 August 2014 SICSA CENSIS Workshop

Page 8: SensorBench

11

7: Impact of Task

18 August 2014 SICSA CENSIS Workshop

Page 9: SensorBench

12

Thank you

SensorBench: benchmarking approaches to processing wireless sensor network data. Ixent Galpin, Alan B. Stokes, George Valkanas, Alasdair J. G. Gray, Norman W. Paton, Alvaro A. A. Fernandes, Kai-Uwe Sattler, and Dimitrios Gunopulos DOI: 10.1145/2618243.2618252SSDBM 2014: 21http://code.google.com/p/sensorbench

www.alasdairjggray.co.uk [email protected] @gray_alasdair

18 August 2014 SICSA CENSIS Workshop