1. Wireless Sensor Network Protocols Dr. Monir Hossen ECE, KUET Department of Electronics and Communication Engineering, KUET

2. Department of Electronics and Communication Engineering, KUET 2 Clustering Algorithm Clustering of the Large Sensor Network Classification of Clustering Algorithms Popular Clustering Algorithms MAC Protocol Different MAC protocols Agendas of This Lecture

3. Department of Electronics and Communication Engineering, KUET 3 Clustering Algorithm (1/2) Clustering is the way where the sensor nodes are grouped in order to => Achieve the network scalability => Solve the network coverage and traffic unbalance problems among the cluster heads (CHs) => Reduce the size of the routing table stored at the individual node => Distribution of nodes among the clusters evenly => Save the energy consumption by reducing data repetition  Improve network lifetime  Reduce network traffic and the contention for the channel  Data aggregation and updates take place in CHs

4. Department of Electronics and Communication Engineering, KUET Nodes divided in virtual group according to some rules Nodes belonging in a group can execute different functions from other nodes Cluster member Clusterhead Gateway node Intra-Cluster link Cross-cluster link Clustering Algorithm (2/2)

5. Department of Electronics and Communication Engineering, KUET 5 Clustering of the Large Sensor Network (1/2)  An efficient clustering algorithm is important for following reasons: => To limit the maximum hop distance => To distribute the nodes among the clusters evenly => To management of time frame properly => To solve the traffic unbalance problem among CHs => To optimize the no. of CH & hop to ensure the network coverage

6. Department of Electronics and Communication Engineering, KUET 6 Clustering of the Large Sensor Network (2/2) The curves represent the coverage ratio of total 500 nodes randomly distributed in 500x500 m 2 area, where maximum distance of a hop is 40 m.

7. Department of Electronics and Communication Engineering, KUET 7 Classification of Clustering Algorithms (1/2) Node registration algorithm Here, H lim is the maximum no. of hops for a cluster Major Approaches for clustering of WSN: 1.Independent clustering algorithm 2.Cooperative clustering algorithm

8. Department of Electronics and Communication Engineering, KUET 8 Independent Clustering Algorithm (1/2)  Independent clustering Algorithm:  Assignment of large H max causes nonsymmetrical node distribution among the clusters  Assignment of small H max reduces coverage, all the nodes are not connected  Optimum number of hop selection is required  Estimation of the Optimum hop number in advance is difficult

9. Department of Electronics and Communication Engineering, KUET 9 Independent Clustering Algorithm (2/2) Nodes under ONU1=175 Nodes under ONU2=125 Nodes under ONU3=121 Nodes under ONU4=79 Maximum hop distance=7 Independent clustering for 500 nodes in an area of 500×500 square meter.

10. Department of Electronics and Communication Engineering, KUET 10 Cooperative Clustering Algorithm (1/2)  Cooperative clustering algorithm:  Can obtain even distribution of sensor nodes among the clusters  Provides less number of hops in a cluster  Total aggregated traffic in the network is less than the independent clustering algorithm

11. Department of Electronics and Communication Engineering, KUET 11 Cooperative Clustering Algorithm (2/2) Nodes under CH1=120, CH2=127, CH3=120, CH4=120 and non-registered nodes=13 up to 4-hops CH1= 121 CH2= 132 CH3= 123 CH4= 124 Up to 5-hops

12. Department of Electronics and Communication Engineering, KUET 12 Popular Clustering Algorithms  Different clustering algorithms have been proposed to achieve energy and delay efficient WSN  Low Energy Adaptive Clustering Hierarchy (LEACH)  Hybrid Energy Efficient Distributed (HEED) Protocol  Weighted Clustering Algorithm (WCA)

13. Department of Electronics and Communication Engineering, KUET 13 Low Energy Adaptive Clustering Hierarchy (LEACH) Every node elects as CH the node that requires the least energy consumption for communication. Every CH set-up a TDMA schedule and transmitted to the nodes. Every node could transmit data in the corresponding time-slot.  Weakness of LEACH  Limited scalability  Could be complementary to clustering techniques based on the construction of a DS

14. Department of Electronics and Communication Engineering, KUET 14 Hybrid Energy Efficient Distributed (HEED) Protocol (1/3) Assumptions:  Sensor quasi-stationary  Links are symmetric  Energy consumption non-uniform for all nodes  Nodes-location unaware  Processing and communication capability-similar

15. Department of Electronics and Communication Engineering, KUET 15 Hybrid Energy Efficient Distributed (HEED) Protocol (2/3) Algorithm: Cluster head selection  Amount of residual energy (primary) and communication cost (secondary) such as node proximity Number of rounds of iterations Tentative CHs formed Final CH until CH prob =1 Same or different power levels used for intra cluster communication

16. Department of Electronics and Communication Engineering, KUET 16 Pros: Balanced clusters Low message overhead Uniform & non-uniform node distribution Inter cluster communication explained Out performs generic clustering protocols on various factors Cons: Repeated iterations and complex algorithm Decrease of residual energy smaller probability  number of iterations increased Nodes with high residual energy are increased in one region of a network Hybrid Energy Efficient Distributed (HEED) Protocol (3/3)

17. Department of Electronics and Communication Engineering, KUET 17 Weighted Clustering Algorithm (WCA) (1/4) A cluster head can ideally supports nodes to  Ensures efficient MAC functioning  Minimizes delay and maximizes throughput A cluster head uses more battery power because it:  Does extra work due to packet forwarding  Communicates with more number of nodes

18. Department of Electronics and Communication Engineering, KUET 18 Weighted Clustering Algorithm (WCA) (2/4)  A cluster head should be less mobile Helps to maintain same configuration Avoids frequent WCA invocation  A better power usage with physically closer nodes More power for distant nodes due to signal attenuation

19. Weighted Clustering Algorithm (WCA) (3/4) It is desirable to balance the loads among the clusters Load balancing factor (LBF) has defined as where, is the number of cluster heads is the number of nodes of cluster i a nd is the average number of neighbors of a cluster head Load Balancing Factor (LBF) Department of Electronics and Communication Engineering, KUET 19

20. Department of Electronics and Communication Engineering, KUET 20 Weighted Clustering Algorithm (WCA) (4/4) For clusters to communicate with each other, it is assumed that cluster heads are capable of operating in dual power mode A cluster head uses low power mode to communicate with its immediate neighbors within its transmission range and high power mode is used for communication with neighboring clusters Connectivity is defined as (for multiple component graph) Connectivity

21. Department of Electronics and Communication Engineering, KUET 21 MAC Protocol  MAC protocols play a vital role for efficient data transmission and collision avoidance in a wireless communication system  Different MAC protocols have been proposed to reduce latency as well as energy consumption in WSN  S-MAC,  Adaptive S-MAC,  LE-MAC,  T-MAC,  E 2 -MAC,  Sync-LS (Synchronised Latency Secured) MAC

22. Department of Electronics and Communication Engineering, KUET 22 Zigbee  Data transmission is followed by beacon from PANC  In each time frame 1-hop transmission is occurred  Requirement of time frames are proportional to the number of hops in the networks

23. Department of Electronics and Communication Engineering, KUET 23 Adaptive S-MAC  Sync signal is transmitted at the starting of time frame  In each time frame 2-hops transmission is occurred  End to end Latency =½*Latency in Zigbee

24. Department of Electronics and Communication Engineering, KUET 24 LE-MAC  Sync signal is transmitted at the starting of time frame  In each time frame 4-hops transmission is occurred  End to end latency= ¼*Zigbee

25. Department of Electronics and Communication Engineering, KUET 25 Sync-LS Protocol  Time frame is divided into two parts -Forward time frame -Reverse time frame

26. Department of Electronics and Communication Engineering, KUET 26 Sync-LS Protocol  In forward time frame, beacon and forward data are transmitted up to end node  To save energy, every node goes to sleep mode after forward data transmission  In reverse time frame, parent nodes transmit beacon to its children nodes and receives the reverse data  Transmission of beacon is controlled by the wakeup timer  Wakeup timer is maintained by the following equation Where, N= Maximum hop number, K= hop distance from edge node, t margin = additional sleep time before reverse time frame, ∆t r = time interval between two reverse beacon, ∆t f = time interval between two forward beacon

27. Department of Electronics and Communication Engineering, KUET 27 Comparison of Average Latency Sync-LS protocol provides lowest latency it requires a single time frame

28. Department of Electronics and Communication Engineering, KUET 28 Comparison of Energy Consumption  Energy consumption depends on the modes of a sensor node - Active mode (Transmission and Reception state) - Sleep mode  Energy consumption expressed by the following equation Where, n=number of nodes in the network, P t = power consumption per sec for data transmission, T t = data transmission time, P r = power consumption per sec for data reception or idle, T a = active time - Sleep mode energy consumption is neglected Simulation result shows that energy consumption in Sync-LS is little higher than Zigbee because Sync- LS requires two beacons.

29. Thanks for Your Kind Attention Department of Electronics and Communication Engineering, KUET