Scribd
Upload
229 views

Path Inference in Wireless Sensor Networks

This document proposes a path inference approach called iPath to reconstruct routing paths in wireless sensor networks. iPath uses three key techniques: 1) iterative boosting to reconstruct long paths from short paths by exploiting high path similarity, 2) PSP-Hashing, a lightweight hash function to verify path similarity, and 3) fast bootstrapping to speed up inference and reconstruct more paths using packet metadata. The approach is evaluated through simulations and aims to improve on existing work that has low inference capability, execution efficiency, or cannot handle dynamic networks.

Uploaded by

arunraja98
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
229 views

Path Inference in Wireless Sensor Networks

This document proposes a path inference approach called iPath to reconstruct routing paths in wireless sensor networks. iPath uses three key techniques: 1) iterative boosting to reconstruct long paths from short paths by exploiting high path similarity, 2) PSP-Hashing, a lightweight hash function to verify path similarity, and 3) fast bootstrapping to speed up inference and reconstruct more paths using packet metadata. The approach is evaluated through simulations and aims to improve on existing work that has low inference capability, execution efficiency, or cannot handle dynamic networks.

Uploaded by

arunraja98
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 7

Path Inference in Wireless Sensor

Networks

Objective
To provide a novel path inference approach to
reconstruct the per-packet routing paths in dynamic
and large-scale networks
To exploit high path similarity to iteratively infer long
paths from short ones
To provide a fast bootstrapping algorithm to improve
inference capability and execution efficiency
To implement a novel design of a lightweight hash
function for verification of the inferred paths.
To evaluate performance of i-path from large scale
WSNs and extensive simulations.

Introduction
It is highly probable that a packet from node i
and one of the packets from i's parent will follow
the same path starting from i's parent toward
the sink. This is referred as high path similarity.
The design of iPath includes three parts:
iterative boosting,
PSP-Hashing, and fast bootstrapping.
The iterative boosting algorithm uses the short
paths to reconstruct long paths iteratively
based on the path similarity.

PSP-Hashing provides a path similarity preserving


hash function that makes the iterative boosting
algorithm be able to verify
whether two paths are similar with high accuracy.
When the global generation time and the parent
change counter are included in each packet, a
fast bootstrapping method is further
used to speed up the iterative boosting algorithm
as well as to
reconstruct more paths.

Literature survey
Title

Yea
r

Author

Description

Demerits

Predicting
201
and tracking 1
internet
path
changes

I. Cunha
et al

DTrack which is a
probe-based path
tracking
system is used to
predict and trace
Internet path changes

Low
inference
capability

Fine-grained 201
latency and 1
loss
measureme
nts in the
presence of
reordering

Lee et al

FineComb, a probebased network delay


and loss topography
approach is used to
focus on resolving
packet reordering.

Low
execution
efficiency

Passive
diagnosis for
wireless sensor
networks

201
0

Liu et
al

PAD assumes a
relatively static
network and uses
each packet to
carry one hop of a
path.

The
frequently
changing
routing path
cannot be
accurately
reconstructe
d in the
dynamic
network

How was your


journey?
Uncovering
routing
dynamics in
deployed
sensor

201
2

Keller
et al

MNT obtains a set of


reliable packets from
the received packets
at sink, then uses the
reliable packet set to
reconstruct each
received packet's path

packet loss
and wireless
dynamics

Proposed Work
Initially the length of all the paths are estimated.
The path similarity between the child and
parents path is estimated.
By comparing the recorded hash value and
the calculated hash value, the sink verifies
whether a long path and a short path share the
same path after the short path's
original node.
Iterating Boosting is used to reconstruct the
longer paths from shorter paths.

You might also like