31
Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks Mohamed Aly In collaboration with Panos K. Chrysanthis and Kirk Pruhs Advanced Data Management Technologies Lab Dept. of Computer Science University of Pittsburgh

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

  • Upload
    drake

  • View
    40

  • Download
    0

Embed Size (px)

DESCRIPTION

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks. Mohamed Aly In collaboration with Panos K. Chrysanthis and Kirk Pruhs Advanced Data Management Technologies Lab Dept. of Computer Science University of Pittsburgh. Motivating Application: Disaster Management. - PowerPoint PPT Presentation

Citation preview

Page 1: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in

Sensor Networks

Mohamed Aly In collaboration with

Panos K. Chrysanthis and Kirk Pruhs

Advanced Data Management Technologies LabDept. of Computer Science

University of Pittsburgh

Page 2: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 2 Mohamed Aly

Motivating Application: Disaster Management

Page 3: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 3 Mohamed Aly

Disaster Management Sensor Networks

Sensors are deployed to monitor the disaster area. First responders moving in the area issue ad-hoc queries

to nearby sensors. The sensor network is responsible of answering these

queries. First responders use query results to improve the

decision making process in the management process of the disaster.

Page 4: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 4 Mohamed Aly

Data Storage Options in Sensor Networks

Base Station Storage: Events are sent to base stations where queries are

issued and evaluated. Best suited for continuous queries.

In-Network Storage (INS): Events are stored in the sensor nodes. Best suited for ad-hoc queries. All previous INS schemes were Data-Centric Storage

(DCS) schemes.

Page 5: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 5 Mohamed Aly

Data-Centric Storage (DCS)

Quality of Data (QoD) of ad-hoc queries. Define an event owner based on the event value. Examples:

Distributed Hash Tables (DHT) [Shenker et. al., HotNets’02]

Geographic Hash Tables (GHT) [Ratnasamy et. al., WSNA’02]

Distributed Index for Multi-dimensional data (DIM)[Li et. al., SenSys’03] Greedy Perimeter Stateless Routing algorithm

(GPSR)[Karp & Kung, Mobicom’00]

Among the above schemes, DIM has been shown to exhibit the best performance.

Page 6: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 6 Mohamed Aly

The DIM DCS Scheme

Page 7: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 7 Mohamed Aly

Problems of Current DCS Schemes Storage Hot-Spots:

A large percentage of events is mapped to few sensor nodes.

Our Solutions The Zone Sharing (ZS) algorithm on top of DIM

[DMSN’05] The K-D Tree based DCS scheme (KDDCS) [submitted]

Query Hot-Spots: A large percentage of queries target events stored in

few sensor nodes. Our Solutions [MOBIQUITOUS’06]

The Zone Partitioning (ZP) algorithm The Zone Partial Replication (ZPR) algorithm

Page 8: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 8 Mohamed Aly

Query Hot-Spots in DIM

Definition: A high percentage of queries accessing a “hot zone” stored by a small number of nodes.

Existence of query hot-spots leads to: Increased node deaths Network Partitioning Reduced network lifetime Decreased Quality of Data (QoD)

Page 9: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 9 Mohamed Aly

Query Hot-Spots Decomposition Algorithms

Uniform vs. skewed distribution of the number of accesses among the hot-zone events: The Zone Partitioning (ZP) algorithm The Zone Partial Replication (ZPR) algorithm

Basic Idea: Each sensor keeps track of the Average Querying

Frequency (AQF) of its stored events Periodically compares its AQF to its neighbors’ AQFs In case a large difference is detected, the node

(donor) selects the Best neighbor (receiver) that can receive part of its responsibility range

Donor locally determines receiver Partitioning Criterion (PC)

Page 10: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 10 Mohamed Aly

The Zone Partitioning (ZP) Algorithm

Page 11: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 11 Mohamed Aly

The Zone Partial Replication (ZPR) Algorithm

Page 12: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 12 Mohamed Aly

Query Hot-Spots Decomposition Algorithms

Uniform vs. skewed distribution of the number of accesses among the hot-zone events: The Zone Partitioning (ZP) algorithm The Zone Partial Replication (ZPR) algorithm

Basic Idea: Each sensor keeps track of the Average Querying

Frequency (AQF) of its stored events Periodically compares its AQF to its neighbors’ AQFs In case a large difference is detected, the node

(donor) selects the Best neighbor (receiver) that can receive part of its responsibility range

Donor locally determines receiver Partitioning Criterion (PC)

Page 13: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 13 Mohamed Aly

Query Hot-Spots Decomposition Algorithms

Uniform vs. skewed distribution of the number of accesses among the hot-zone events: The Zone Partitioning (ZP) algorithm The Zone Partial Replication (ZPR) algorithm

Basic Idea: Each sensor keeps track of the Average Querying

Frequency (AQF) of its stored events Periodically compares its AQF to its neighbors’ AQFs In case a large difference is detected, the node

(donor) selects the Best neighbor (receiver) that can receive part of its responsibility range

Donor locally determines receiver Partitioning Criterion (PC)

Page 14: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 14 Mohamed Aly

PC: Storage Safety Requirement

The sum of the pre-partitioning load of the receiver and the traded zone should be less than the receiver’s storage capacity

T + lreceiver ≤ S

Page 15: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 15 Mohamed Aly

PC: Energy Safety Requirement (1)

The energy consumed by the donor in the partitioning process should be much less than the total energy of the donor

T / edonor ≤ E1

E1 ≤ 0.5

Page 16: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 16 Mohamed Aly

PC: Energy Safety Requirement (2)

The energy consumed by the receiver in the partitioning process should be much less than the total energy of the receiver

(T * re) / ereceiver ≤ E2

E2 ≤ 0.5

Page 17: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 17 Mohamed Aly

PC: Access Frequency Requirement

The average access frequency of the donor is much larger than that of the receiver

AQF(donor) / AQF(receiver) ≥ Q1 Q1 should be greater than 2 to avoid cyclic

migrations

Page 18: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 18 Mohamed Aly

ZPR Initiation Requirements

In case all previous requirements are satisfied: ZP initiated

If a hot sub range of small size exists within the hot range ZPR initiated instead of ZP

AQF(hot sub range) / AQF(total range) ≥ Q2

Q2 should be close to 1, for e.g. 0.9 size(hot sub range) / size(total range) ≤ Q3

Q3 should be close to 0, for e.g. 0.2

Page 19: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 19 Mohamed Aly

Partitioning Criterion (PC)

1. T + lreceiver ≤ S

2. T / edonor ≤ E1

3. (T * re) / ereceiver ≤ E2

4. AQF(donor) / AQF(receiver) ≥ Q1

5. AQF(hot sub range) / AQF(total range) ≥ Q2

6. size(hot sub range) / size(total range) ≤ Q3

1:4 satisfied ZP initiated

1:6 satisfied ZPR initiated

Page 20: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 20 Mohamed Aly

More about the Algorithms

Mechanism to lower messaging overhead GPSR Modifications

Traded Zone List (TZL) Coalescing Process Insertion process in ZPR Bound on the replication hops of ZPR

Page 21: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 21 Mohamed Aly

Roadmap

Background Problem Statement: Query Hot-spots Algorithms: ZP, ZPR Experimental Results Conclusions

Page 22: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 22 Mohamed Aly

Simulation Description

Compare: DIM, ZP/ZPR. Simulator similar to the DIM’s [Li et. al., SenSys’03] Two phases: insertion & query. Insertion phase (to achieve a steady state of network

storage) Each sensor initiates 5 events Events forwarded to owners

Query phase Each sensor generates 20 single-event queries

(worst case scenario)

Page 23: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 23 Mohamed Aly

Experimental Setup

Parameter Value

Network size 50 to 300 sensors

Initial energy 100 units

Energy unit energy needed to send one event

E1 & E2 2

Q1 , Q2 , and Q3 3, 0.8, and 0.2

Number of hot-spots 1

Hot-spot sizes 0.05% - 10% of attribute ranges

Sensor node storage capacity 10 units (events)

Page 24: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 24 Mohamed Aly

Experimental Results: Quality of Data (QoD)

5% hot-spot

0

1

2

3

4

5

6

50 100 150 200 250 300

Th

ou

san

ds

Network Size

No

. U

nan

swer

ed Q

uer

ies

DIM

ZP-ZPR

Page 25: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 25 Mohamed Aly

Experimental Results: Balancing Energy Consumption

200 nodes, 0.33% hot-spot

020

4060

80100

120140

160180

200

10 20 30 40 50 60 70 80 90 100

Node Energy Level

Nu

mb

er o

f N

od

es

DIM

ZP-ZPR

Page 26: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 26 Mohamed Aly

Experimental Results: ZP/ZPR Strengths

Increasing the QoD by partitioning the hot range among a large number of sensors, thus, balancing the query load among sensors and keep them alive longer to answer more queries.

Increasing energy savings by balancing energy consumption among sensors.

Increasing the network lifetime by reducing node deaths.

Page 27: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 27 Mohamed Aly

Acknowledgment

This work is part of the “Secure CITI: A Secure Critical Information Technology Infrastructure for Disaster Management (S-CITI)” project funded through the ITR Medium Award ANI-0325353 from the National Science Foundation (NSF).

For more information, please visit: http://www.cs.pitt.edu/s-citi/

Page 28: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 28 Mohamed Aly

Conclusions and Extensions

Query Hot-Spots: An important problem in current DCS schemes.

Contribution: A query hot-spots decomposition scheme for DCS

sensor nets, ZP/ZPR, working on top of the DIM DCS scheme.

Experimental validation of the ZP/ZPR practicality Work under submission:

KDDCS: A unified DCS scheme for load balancing storage and query loads.

Page 29: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 29 Mohamed Aly

Thank You

Questions ?

Advanced Data Management Technologies Labhttp://db.cs.pitt.edu

Page 30: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 30 Mohamed Aly

Experimental Results: Load Balancing

0.05% hotspot

1.5

2

2.5

3

3.5

4

4.5

50 100 150 200 250 300

Network Size

Avg

. N

od

e S

tora

ge

DIM

ZP-ZPR

Page 31: Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks

Decomposing Data-Centric Storage Query Hot-Spots in Sensor Networks 31 Mohamed Aly

Experimental Results: Load Balancing

0.05% hot-spot

5

10

15

20

25

30

35

50 100 150 200 250 300

Network Size

No

. o

f F

ull

No

des

DIM

ZP-ZPR