Fault-Tolerant Data Collection in Fault-Tolerant Data Collection in - - PowerPoint PPT Presentation

fault tolerant data collection in fault tolerant data
SMART_READER_LITE
LIVE PREVIEW

Fault-Tolerant Data Collection in Fault-Tolerant Data Collection in - - PowerPoint PPT Presentation

Fault-Tolerant Data Collection in Fault-Tolerant Data Collection in Heterogeneous Intelligent Monitoring Networks Networks Heterogeneous Intelligent Monitoring Jing Deng Department of Computer Science University of North Carolina at


slide-1
SLIDE 1

1

Fault-Tolerant Data Collection in Fault-Tolerant Data Collection in Heterogeneous Intelligent Monitoring Heterogeneous Intelligent Monitoring Networks Networks

Jing Deng

Department of Computer Science University of North Carolina at Greensboro jing.deng@uncg.edu http://www.uncg.edu/~j_deng/ Joint work with Profs. Meikang Qiu and Gang Wu

slide-2
SLIDE 2

2

Wireless Networks

  • Networks formed by wireless devices

– All communications are sent through wireless channels. – Wireless devices with limited resource

Battery energy, memory space, computation power

  • Many interesting problems:

– How to lower communication/computation cost for network activities?

Communication takes time/energy. Computation requires memory space and energy.

– How to protect systems from node failure?

Small wireless devices could easily fail or run out of battery.

slide-3
SLIDE 3

3

Failure Models

  • Fail-stop

– Device simply stops working. – No information will be sent or received. – Similarly to one with dead-battery

  • Byzantine failure

– Device can virtually do anything that it is capable of

  • Dropping packets from others
  • Sending out fabricate packets
  • Modifying packets from other nodes
  • Deviating from communication protocols

– Much more difficult to address We will use the fail-stop model

slide-4
SLIDE 4

4

Intelligent Monitoring Networks (IMNs)

  • Wireless sensor networks

– Networks with (possibly numerous) wireless micro-sensors

  • A special type of wireless sensor networks

– Likely to be deployed for building structure monitoring, forest monitoring, levee monitoring, industrial plant monitoring, etc.

  • Two key characteristics

– Node failures expected – Heterogeneous architecture

  • Mostly small devices to collect/report data
  • Some larger and more powerful devices to process/fusion data
  • These power nodes send results to observer (data sink).
slide-5
SLIDE 5

5

Illustration of Large Wireless Networks

slide-6
SLIDE 6

6

Low-power Low-Cost Devices

  • Devices usually use low-power transceivers

– Goal: to lower energy consumption and to extend lifetime

  • Forming multi-hop communication topology

– Relying on other devices’ help to deliver data

  • Interference can easily disrupt communication

– Network topology changes – Data collection paths change – Data loss

slide-7
SLIDE 7

7

BitTorrrent - P2P File Sharing Technique

  • Swarm: collection of

nodes with the file (even partially).

  • Tracker provides

swarm information

  • Client downloads

pieces from nodes in swarm.

  • At the same time,

uploading pieces to

  • ther nodes
  • Finished clients serve

as seeds (upload only) Even when some of the nodes in the swarm fail (or left), file sharing continues.

slide-8
SLIDE 8

8

BitTorrent Strategies

  • Two strategies in BitTorrent make it surprisingly efficient

– Optimistic un-choking – Rarest-first

  • Optimistic un-choking is the strategy to choose peers to

download pieces

– Suppose there are 100 peers (with ever-changing D/L speed). – Which of these 100 peers should the client choose?

  • Using all of them is impractical.
  • Choosing the top N peers w.r.t. download speed (N=5)
  • However, there might be new peers offering higher speed.
  • -> dropping one of the current N peers and randomly testing

another peer (un-choke one of the unselected peers)

– Benefits

  • Utilizing most of the peers with highest D/L speeds.
slide-9
SLIDE 9

9

BitTorrent Strategies (Cont’d)

  • Rarest-first strategy governs how to choose pieces for

download

– Suppose a peer has M of the pieces – Which of these pieces should the client choose?

  • Random selection or sequential selections?
  • -> Always choose the rarest piece among all peers (requiring piece

information from other peers).

  • So that this piece can be offered to other peers.

– Benefits

  • Increases piece redundancy
  • Maintaining torrent health
  • Improves chance of successful download
slide-10
SLIDE 10

10

IMN and BitTorrent?

  • Data collection in IMNs shares striking similarities with

P2P file sharing

Peers may go offline without warning Nodes may fail at any time. Redundancy of file pieces among peers Monitoring data redundancy among different nodes A client tries to download a full set of pieces from a swarm of nodes Observer (data sink) tries to collect data from monitoring nodes, which generate the data

P2P File Sharing IMN

slide-11
SLIDE 11

11

IMN - Connectivity Overview

  • Lines connect nodes

who can hear each

  • ther (N=100).
  • Darker squares mark

more powerful nodes (M=10).

  • Result of random

node placement

slide-12
SLIDE 12

12

Fault Tolerant Data Collection

  • Powerful nodes collect data from regular nodes

– Announcements are made from the powerful nodes. – Multiple trees are formed with data forwarding nodes.

  • Usually data forwarding nodes only need to forward data

from nodes on their own tree

– In order to provide fault tolerance, they will choose α of other

  • verheard transmissions

– α is termed support ratio

slide-13
SLIDE 13

13

Illustration of Data Collection

  • Multi-level data

collection

  • We show the [avg,

max] record on the powerful nodes

  • α=0.4
  • Some nodes fail

(marked with red x)

  • Big red dot represents

a fire burning

  • None of the powerful

nodes sees any temperature anomaly.

slide-14
SLIDE 14

14

Illustration of Data Collection (Cont’d)

  • The same topology

and data collection trees.

  • α=0.4
  • With the same failed

nodes, two powerful nodes receive the temperature anomaly (Nmax=2).

slide-15
SLIDE 15

15

Performance Results - Reading Abnormal Temp.

  • Similar simulations

were run and average Nmax computed

  • pe is node failure

probability

  • Nmax lowers as pe

increases.

  • With larger α, Nmax

increases.

slide-16
SLIDE 16

16

Data Loss due to Failed Sensors

  • Failed nodes lead to

data loss

  • Support ratio α can

dramatically reduce data loss.

slide-17
SLIDE 17

17

Conclusions

  • Wireless monitoring networks can provide robust

environment monitoring.

  • We have proposed a fault-tolerance data collection

technique for IMNs:

– Multiple multi-level data collection trees (forest) – Data forwarding nodes process overheard data. – Support ratio α

  • Benefits of our scheme have been demonstrated

– Low cost – Fault tolerant toward node failures

slide-18
SLIDE 18

18

Future Work

  • Byzantine failure model

– Instead of failed delivery, failed nodes may send wrong data!

  • Investigating our scheme under different data

processing algorithms

– Average – Maximum – Minimum – Counting

  • Analyze data loss for different support ratios