An Energy efficient approach in Heterogeneous WSN

  An Energy efficient approach in Heterogeneous WSN.ppt (Size: 292.5 KB / Downloads: 8)
INTRODUCTION
Intrusion detection (object tracking): monitoring System for detecting the intruder
Concerns how fast an intruder can be detected by WSN
intruder detection is fast in a network with High density deployment policy
a network with small and scattered void areas will also be able to detect a moving intruder within a certain intrusion distance.
INTRUSION DETECTION SYSTEM
-system dedicated to detect unauthorized or unusual activity into a system or network.
-two types: stand-alone and cooperative
-Intrusion Detection technology can be divided into Misuse detection and Anomaly detection
Literature Survey
The works are undergoing on
-size of the intruder
-movement of the intruder
-minimizing IDS modules
-Routing the IDS data
-trust based IDS
-resource efficient IDS (for e.g. Energy)
to develop useful security mechanisms it is necessary to know and understand the constraints of Sensor security
Detection Model
Two detection models
single-sensing detection model - the intruder can be identified by using the sensing knowledge from one single sensor.
multiple-sensing detection model (K-sensing)-the intruder can only be identified by using cooperative knowledge from at least k sensors
Problem Definition
Problem Statement: Design a new algorithm for intrusion detection and finds how an Intrusion detection in Heterogeneous WSN is related to sensing range and node density in an energy efficient way.
Objective:
Finds the relationship among intrusion detection , sensing range and node density
Finds the number of nodes required for an intrusion detection
Assumptions
All nodes are static
All nodes sensor have omni antenna properties
Intruder is an object moving through WSN
Sink knows each nodes location and its neighbours.
Heterogeneous WSN
Different type of sensors
IDS has two functions
Identify the intruder
Pass this information to Sink
The minimum no. of sensors required to cover the WSN area=The minimum number of sensors required for detection.
ALGORITHM
Si- set of type i sensors in the WSN area.
S- set of all sensors
N(a)- set of neighbours of node a
Repeat
For i=0 to N
Select node a with min N(a) in set Si
If N(a)≠Ø
Put a in stack
SiN= {i/the distance between a and i (where i ε N(a))<(rs/2)}
If
S=S-(SiN U a)
Else
S= S-a
Until S is null set.
PROBABILITY ANALYSIS-Single Sensing
Theorem 1
The probability that an intruder can be immediately detected once it enters a heterogeneous WSN can be given by
where ni is the number of type i nodes activated in the area
Proof
here the area we need to consider when the intruder enters from the boundary is A1=(π rs12)/2, A2=(πrs22)/2 …AN= πrsN2/2 .
So P (0, A1) , P (0, A2)….P(0,AN) gives the probability that there is no Type 1,Type 2…type N sensors in that area.
the probability that neither type 1 nor type 2….nor type N are given P(0,A1) P (0, A2)…..P(0.AN)= e-n1e-n2…e-nN where n1,n2,…nN are the number of selected nodes from each type.
So the probability of detecting the intruder when it enters the boundary is given by complement of P (0, A1) P (0, A2)….P(0,AN) =1-e-n1e-n2….e-nN.
Theorem 2: Suppose is the maximal intrusion distance allowable for a given application, the probability P(D) that the intruder can be detected within in the given heterogeneous WSN can be derived as
where ni is the number of sensors participating in intrusion detection area
Multisensing
Theorem 3 :Let Pm (D= 0) be the probability that an intruder is detected immediately once it enters a heterogeneous WSN in multisensing detection model.
It has Pm(D=0)=
where Aj is the area covered by type j sensor and we are assuming that nj of type j sensors are activated in the area Aj.
Proof
Here the area is only one half circles with radius rs.
P(i,A) gives the probability of detecting the intruder with i sensors gives the sum of the probabilities of detecting the intruder with less than m sensors.
So the complement will give the multi sensing probability.
Implementation
Done using MatLab
500 sensors are deployed in 1000X1000x1000 cubical area.
Sensors are uniformly and independently deployed.
Maximum allowable distance for intruder=50.
Sensing range and number of sensors are varied to generate the graphs.
Conclusion
It includes an algorithm for intrusion detection (object tracking) which uses minimum resources.
It also analyses the probability of intrusion detection.
This algorithm can also used for internal intrusion detection.
Here the probability depends only on the number of sensors which activated its IDS module in that area.
Future Works
Apply security in to the intrusion detection process by selecting a set of trusted nodes and do the intrusion detection with this set of nodes.
References
[1] Lee, J.J., Krishnamachari, B., Kuo, C.C.J.: Impact of Heterogeneous Deployment on Lifetime Sensing Coverage in Sensor Networks (IEEE SECON). (2004)
[2] Hu, W., Chou, C.T., Jha, S., and Bulusu, N.: Deploying Long-Lived and Cost-effective Hybrid Sensor Networks. Elsevier Ad-Hoc Networks, Vol. 4, Issue 6. (2006) 749-767.
[3] A. P. da Silva, M. Martins, B. Rocha, A. Loureiro, L. Ruiz, and H. C. Wong, “Decentralized intrusion detection in wireless sensor networks,” in Proceedings of the 1st ACM international workshop on Quality of service & security in wireless and mobile networks.
[4]Mohamed Mubarak T,Syed Abdul Sattar,Appa rao,Sajitha M “Intrusion Detection: A 3D probability model for Intrusion Detection” International Journal of computer applications ISSN:0975-8887 Volume 6– No.12, September 2010,USA
[5] Mohamed Mubarak T,Syed Abdul Sattar,Appa rao,Sajitha M “Intrusion detection for WSN in Three dimensional space” at International journal of Emerging Technologies and applications in Engineering, Technology and Science ISSN:0974-3588.April 2010.
[6] Mohamed Mubarak T,Syed Abdul Sattar,Appa rao,Sajitha M “Energy Efficient Intrusion detection in 3D WSN” at IEEE Explore ISBN: 978-1-4244-5965-0
[7] Mohamed Mubarak T,Syed Abdul Sattar,Appa rao,Sajitha M“Energy Efficient Intrusion detection in 3D WSN IEEE International Conference on Computational Intelligence and Computing Research,TCE,Coimbatore Dec.28,29