CS 599 Intelligent Embedded Systems1 Adaptive Protocols for Information Dissemination in Wireless Sensor Networks W.R.Heinzelman, J.kulik, H.Balakrishnan

CS 599 Intelligent Embedded

Systems 1

Adaptive Protocols for Information Dissemination in Wireless Sensor Networks

W.R.Heinzelman, J.kulik, H.Balakrishnan

CS 599 Intelligent Embedded Systems 2

Outline Introduction SPIN Other Data Dissemination Algorithms Sensor Network Simulations Conclusions Strengths and Weaknesses


Introduction Wide deployment of Wireless sensor networks Wireless sensor networks

Can aggregate sensor data to provide multi-dimensional view

Improve sensing accuracy Focus on critical events (e.g. intruder entering) Fault tolerant network Can improve remote access to sensor data – sink

nodes


Introduction contd. Limitations of Wireless sensor networks

Energy Computation Communication


Sensor Protocols for Information via Negotiation (SPIN)

Classic flooding limitations Implosion Overlap Resource blindness


Implosion Problem


Overlap problem


SPIN contd.. SPIN overcomes these deficiencies

Negotiation Resource-adaptation

Each sensor node has resource manager Keeps track of resource consumption Applications probe the manager before any activity Cut down activity to save energy

Motivated by principle of ALF Common data naming (meta-data)


SPIN Meta-Data Sensors use meta-data to describe the

sensor data briefly If x is the meta-data descriptor for data X

sizeof (x) < sizeof (X) If x==y

sensor-data-of (x) = sensor-data-of (y) If X==Y

meta-data-of (X) = meta-data-of (Y) Meta-data format is application specific


SPIN Messages

ADV – new data advertisement REQ – request for data DATA – data message

ADV and REQ messages contain only meta-data so they are smaller in size.


SPIN-1 and SPIN-2 SPIN-1

Simple 3-stage handshake protocol Data aggregation is possible Can adapt to work in lossy or mobile network Can run in a completely unconfigured

network


Node B sends a REQ listing all of the data it would like to acquire.


If node B had its own data, it could aggregate this with the data of node A and advertise.


Nodes need not respond to every message


SPIN-2

SPIN-1 with a Low-Energy Threshold When energy below energy threshold – stop

participating in the protocol Can just receive data avoiding ADV-REQ phase


Other data dissemination algos. Classic Flooding

Converges in O(d), d-diameter of the network Gossiping

Forward data to a random neighbor Avoids implosion Disseminates at a slow rate Fastest rate = 1 node/round



Ideal dissemination Every node sends sensor data along shortest path Receives each piece of distinct data only once Implementation

Network level multicast (source specific) To handle losses, reliable multicast has to be

deployed SPIN is a form of application-level multicast



Sensor Network Simulations Simulated using ns

simulator Extended ns to

create a Resource-Adaptive Node


Simulation Testbed


SPIN-1 Results Higher throughput than gossiping Same throughput as flooding Uses substantially less energy than other protocols SPIN-2 delivers more data per unit energy than

SPIN-1 SPIN-2 performs closer to Ideal dissemination Nodes with higher degree tend to dissipate more

energy than nodes with lower degree


Data Acquired Over Time


Energy Dissipated Over Time


Energy Dissipated Over Time


Unlimited Energy Simulations


Limited Energy Simulations


Limited Energy Simulations contd..


Best-Case Convergence Times For overlapping sensor data

Convergence times for ideal and flooding are the same

For non-overlapping sensor data Flooding converges faster than SPIN-1

To understand these results, we develop equations that predict convergence times of each of these protocols.


Transmission time per data packet = 8s/d

Since SPIN-1 has to process ADV, REQ, DATA so processing time = 3(d+r)

)8

)(3()8

3(

)8

(,)8

(

1b

srdlC

b

sdl

b

srdlCC

b

sdl

dSPINd

dFloodIdeald

Convergence Time – no overlap


Convergence Time – overlapping data

2

4)1(

)8

()8

3(

)8

)(3()8

3(

)8

(,)8

(

1

r

b

skd

b

skrdl

b

sdl

b

skrdlC

b

sdl

b

skrdlCC

b

sdl

lplp

lpSPINlp

lpFloodIdeallp


For the testbed network parameters

Simulation results Flooding converges in 135ms Ideal converges in 125ms SPIN-1 converges in 215ms

Convergence times of flooding and ideal are closer to their upper bound unlike SPIN-1

294.0133.0

154.0,063.0

1

SPIN

FloodIdeal

C

CC


Conclusions SPIN solves the implosion and overlap problems. SPIN-1 and SPIN-2 are simple protocols for

wireless sensor networks. SPIN outperforms gossiping. SPIN-1 consumes only 25% energy w.r.t flooding SPIN-2 distributes 60% more data per unit energy

w.r.t flooding.


Strengths and Weaknesses Implosion problem still exists in the REQ

stage The paper doesn’t consider the collisions in

the REQ stage No justification for the network parameters

chosen

i


Questions ?

Documents

CS 599 Intelligent Embedded Systems1 Adaptive Protocols for Information Dissemination in Wireless Sensor Networks W.R.Heinzelman, J.kulik, H.Balakrishnan