Transcript Ultra-Low Power Storage for Sensor Networks

Re-thinking Data Management for
Storage-Centric Sensor Networks
Deepak Ganesan
University of Massachusetts Amherst
With: Yanlei Diao, Gaurav Mathur, Prashant Shenoy
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
Sensor Network Data Management
Live Data Management: Queries on current or recent data.
Applications:
Real-time feeds/queries: Weather, Fire, Volcano
Detection 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 monitoring
Our focus is on designing an efficient archival data
management architecture for sensor networks
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
2
Archival Querying in Sensor Networks
Data Gathering with
centralized archival query
processing
Internet
DBMS
Gateway
Lossless
aggregation
Efficient for low rate,
small volume sensors
such as weather sensors
(temp, humidity, …).
Inefficient energy-wise
for “rich” sensor data
(acoustic, video, highrate vibration).
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
3
Archival Querying in Sensor Networks
Store data locally at
sensors and push queries
into the sensor network
Internet
Flash memory energyefficiency, cost, capacity.
Gateway
Push query to
sensors
Limited capabilities of
sensor platforms.
Flash Memory
Acoustic stream
Image stream
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
4
Technology Trends in Storage
Energy Cost
(uJ/byte)
CC1000
Communication
Atmel NOR
CC2420
Storage
Telos STM NOR
Micron NAND
128MB
Generation of Sensor Platform
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
5
Outline
Case for Storage-centric Sensor Networks
Challenges in a Storage-centric Sensor Database
StonesDB Architecture
Local Database Architecture
Distributed Database Architecture
Conclusion
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
6
Optimize for Flash and RAM Constraints
Memory
Flash Memory Constraints
Data cannot be over-written,
only erased
Pages 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
operations
Aging: Need to efficiently
delete old data items when
storage is insufficient.
~4-10 KB
2. Modify in-memory
1. 1. Load block
2. Into Memory
3. Save
block back
Erase
block
~16-64 KB
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
7
Support Rich Archival Querying Capability
SQL-style Queries: Min, max, count, average,
median, top-k, contour, track, etc
Wireless Sensor Network
Classification Queries: What type of
vehicles (truck, car, tank, …) were
observed in the field in the last month?
Similarity Search: Was
a bird matching
signature S observed
last week?
Signal Processing: Perform an FFT
to find the mode of vibration signal
between time <t1,t2>?
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
8
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.
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
9
StonesDB Architecture
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
10
Example: Indexing in StonesDB
Naïve Design:
Consider a value-based index on entire stream
Deletion/Aging of data triggers in-place updates involving
energy-intensive block read/write/erase operations.
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
11
Indexed Storage
StonesDB Design:
Split data stream into partitions and build index on each
partition. Age partitions as a whole cheaply.
Flash
Block
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
12
Outline
Case for Storage-centric Sensor Networks
Challenges in a Storage-centric Sensor Database
StonesDB Architecture
Local Database Architecture
Distributed Database Architecture
Conclusion
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
13
StonesDB: Data Mining Queries
Similarity Search: Was a bird matching
signature S observed last week?
Proxy Cache of Image Summaries
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
14
StonesDB: System Operation
Similarity Search: Was a bird matching
signature S observed last week?
Query Engine
Partitioned Access Methods
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
15
Research Issues
Local Database Layer
Impact of RAM limitations on storage organization
Energy-optimized indexing and aging.
New cost models for self-tuning energy-efficient
sensor databases.
Distributed Database Layer
Intelligent split of query processing between
proxy and sensor tiers
Adaptively tuning quality of data cached at sensor
proxy based on query needs
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
16
The End
STONES: STOrage-centric Networked Embedded Systems
http://sensors.cs.umass.edu/projects/stones
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
Sensor Data Management Taxonomy
Querying
Mining
Timeline vs Prior Knowledge
Search/Mining on Archived Sensor Data
Acquisitional
Query Processing
(BBQ, …)
e of data processing
Pushdown Filters
(TinyDB, Cougar, …)
Current
Recent
Past
Timeline of data being processed
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
18
Technology Trends in Sensor Platforms
Cyclops Camera+ Mica2 Mote
128 x 128 resolution images
4 KB RAM, 10 MHz
microcontroller
OmniVision Camera + iMote2
128 x 128 resolution images
64KB - 32MB RAM, 10 MHz
microcontroller
Spectrum of sensing devices with different power,
capability, resource constraints.
UNIVERSITY OF MASSACHUSETTS, AMHERST • Department of Computer Science
19