+ All Categories

Slides

Date post: 20-Nov-2014
Category:
Upload: flashdomain
View: 516 times
Download: 0 times
Share this document with a friend
Description:
 
Popular Tags:
19
U NIVERSITY OF NIVERSITY OF M ASSACHUSETTS ASSACHUSETTS , A , AMHERST MHERST Department of Computer Science Department of Computer Science Re-thinking Data Management for Storage-Centric Sensor Networks Deepak Ganesan University of Massachusetts Amherst With: Yanlei Diao, Gaurav Mathur, Prashant Shenoy
Transcript
Page 1: Slides

UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Re-thinking Data Management for Storage-Centric Sensor Networks

Deepak GanesanUniversity of Massachusetts Amherst

With: Yanlei Diao, Gaurav Mathur, Prashant Shenoy

Page 2: Slides

2UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Sensor Network Data ManagementLive Data Management: Queries on current or recent data.

Applications: Real-time feeds/queries: Weather, Fire, VolcanoDetection and Notification: Intruder, Vehicle

Techniques:Push-down Filters/Triggers: TinyDB, Cougar, Diffusion, …Acquisitional Query Processing: BBQ, PRESTO, …

Archival Data Management: Querying or Mining of past data

Applications:Scientific Analysis of past events: Weather, Seismic, …Historical trends: Traffic analysis, habitat monitoringOur focus is on designing an efficient archival data management architecture for

sensor networks

Page 3: Slides

3UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Archival Querying in Sensor Networks

Data Gathering with centralized archival query processing

Efficient for low rate, small volume sensors such as weather sensors (temp, humidity, …).

Inefficient energy-wise for “rich” sensor data (acoustic, video, high-rate vibration).

Lossless aggregation

DBMS

Internet

Gateway

Page 4: Slides

4UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Archival Querying in Sensor Networks

Acoustic stream

Store data locally at sensors and push queries into the sensor network

Flash memory energy-efficiency, cost, capacity.

Limited capabilities of sensor platforms.

Internet

Gateway

Image stream

Flash Memory

Push query to sensors

Page 5: Slides

5UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Technology Trends in Storage

Generation of Sensor Platform

CC1000

CC2420

Telos STM NOR

Atmel NOR

Communication

Storage

Micron NAND 128MB

Energy Cost

(uJ/byte)

Page 6: Slides

6UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Outline

Case for Storage-centric Sensor Networks

Challenges in a Storage-centric Sensor

Database

StonesDB Architecture

Local Database Architecture

Distributed Database Architecture

Conclusion

Page 7: Slides

7UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Optimize for Flash and RAM Constraints

Flash Memory ConstraintsData cannot be over-written, only erasedPages can often only be erased in blocks (16-64KB)Unlike magnetic disks, cannot modify in-place

Challenges:Memory: Minimize use of memory for flash database.Energy: Organize data on flash to minimize read/write/erase operationsAging: Need to efficiently delete old data items when storage is insufficient.

1. 1. Load block 2. Into Memory

3. Save block back

Eraseblock

Memory

2. Modify in-memory

~16-64 KB

~4-10 KB

Page 8: Slides

8UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

SQL-style Queries: Min, max, count, average, median, top-k,

contour, track, etc

Similarity Search: Was a bird matching signature S observed last

week?

Classification Queries: What type of vehicles (truck, car, tank, …) were observed in the

field in the last month?

Wireless Sensor Network

Support Rich Archival Querying Capability

Signal Processing: Perform an FFT to find the mode of vibration signal between time

<t1,t2>?

Page 9: Slides

9UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

StonesDB Goals

Our goal is to design a distributed sensor database for archival data management that:

Supports energy-efficient sensor data storage, indexing, and aging by optimizing for flash memories.

Supports energy-efficient processing of SQL-type queries, as well as data mining and search queries.

Is configurable to heterogeneous sensor platforms with different memory and processing constraints.

Page 10: Slides

10UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

StonesDB Architecture

Page 11: Slides

11UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Example: Indexing in StonesDB

Naïve Design:Consider a value-based index on entire streamDeletion/Aging of data triggers in-place updates involving energy-intensive block read/write/erase operations.

Page 12: Slides

12UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Indexed Storage

StonesDB Design:Split data stream into partitions and build index on each partition. Age partitions as a whole cheaply.

Flash

Block

Page 13: Slides

13UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Outline

Case for Storage-centric Sensor Networks

Challenges in a Storage-centric Sensor

Database

StonesDB Architecture

Local Database Architecture

Distributed Database Architecture

Conclusion

Page 14: Slides

14UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

StonesDB: Data Mining Queries

Similarity Search: Was a bird matching signature S observed

last week?

Proxy Cache of Image Summaries

Page 15: Slides

15UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

StonesDB: System Operation

Similarity Search: Was a bird matching signature S observed

last week?

Query Engine

Partitioned Access Methods

Page 16: Slides

16UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Research Issues

Local Database LayerImpact of RAM limitations on storage organizationEnergy-optimized indexing and aging. New cost models for self-tuning energy-efficient sensor databases.

Distributed Database LayerIntelligent split of query processing between proxy and sensor tiersAdaptively tuning quality of data cached at sensor proxy based on query needs

Page 17: Slides

UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

The End

STONES: STOrage-centric Networked Embedded Systems

http://sensors.cs.umass.edu/projects/stones

Page 18: Slides

18UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Sensor Data Management Taxonomy

Timeline vs Prior KnowledgeQ

uery

ing

Min

ing

Current Recent Past

Acquisitional Query Processing

(BBQ, …)

Pushdown Filters(TinyDB, Cougar, …)

Timeline of data being processed

Search/Mining on Archived Sensor Data

Type of data processing

Page 19: Slides

19UUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer ScienceUUNIVERSITY OF NIVERSITY OF MMASSACHUSETTSASSACHUSETTS, A, AMHERST • MHERST • Department of Computer Science Department of Computer Science

Technology Trends in Sensor Platforms

Cyclops Camera+ Mica2 Mote128 x 128 resolution images4 KB RAM, 10 MHz microcontroller

OmniVision Camera + iMote2128 x 128 resolution images64KB - 32MB RAM, 10 MHz microcontroller

Spectrum of sensing devices with different power, capability, resource constraints.


Recommended