Multimedia & Networking Lab Network Aware Energy-Efficient Communication Protocol (NAEEC) for Heterogeneous Wireless Sensor Networks By : Manoj K. Garg Advisor : Dr. Balakrishnan Prabhakaran A Clustering Protocol Multimedia & Networking Lab
Overview Introduction Problem Statement Related Work System Model Propose NAEEC Protocol Simulation Results Our Contribution Future Work Multimedia & Networking Lab Multimedia & Networking Lab
Homogeneous Vs Heterogeneous WSN Wireless sensor networks (WSN) Large number of distributed sensor Nodes works in cooperation Application: Surveillance, machine failure diagnosis, and chemical/ biological detection Homogeneous sensor networks Sensor nodes with identical characteristics Most common characteristics: Power, Processing, Storage and Radio capabilities Heterogeneous sensor networks Sensor nodes with different level of common characteristics Multimedia & Networking Lab
Motivation Applications for Heterogeneous WSN Re-energization of sensor networks Practical constraints such as Cost Different energy or resources consumption Sensor nodes are energy critical Small battery size Embedded batteries, difficult recharging or changing procedures Wireless transmission costlier than local computation Multimedia & Networking Lab
Network Architecture Sink Sensor Nodes Sensor Field Task Manager Node Sink Internet & Satellite Sensor Field Sensor Nodes Multimedia & Networking Lab
Problem Statement Direct transmission Muti-hop transmission Nodes that are far away from the sink would die first Muti-hop transmission Nodes near the sink acts as relay with very high probability Nodes near to the sink die first TDMA technique Transmission is random Large number of sensor nodes The clustering schemes Homogeneous clustering scheme doe snot result an optimize solution We need to design an new clustering protocol with the MAC layer support to save the sensor’s energy in heterogeneous wireless sensor networks Multimedia & Networking Lab
Related Work – LEACH LEACH (Low Energy Adaptive clustering Hierarchy) A clustering protocol for homogeneous/heterogeneous SN A probabilistic method Based on residual energy relative to the current system energy Advantage Distribute the cluster-head load Avoid the need for global network knowledge. Disadvantage Doesn't guarantee that high energy/resources nodes always be chosen as cluster-head Doesn’t control the number and even placement of cluster-heads Multimedia & Networking Lab
Related Work – SEP, DEEC, SWEET Variation of LEACH Work effectively for heterogeneous networks None of them targets the efficient cluster-head placement problem of LEACH Do not reduce the initial high energy requirement and the number of dead nodes formed over time Multimedia & Networking Lab
Related Work – other algorithms Based on two types of sensor nodes type-I and type-II. Type-II sensor nodes More powerful . Fewer in number . Called overlay sensors. Type-I sensor nodes Normal sensor nodes. Report to the overlay nodes. The system model Assumes pre-estimated fraction, and position of overlay sensor nodes. No applicable when node heterogeneity is a result of the operation of the sensor network and not a choice of optimal network setting. Multimedia & Networking Lab
System Model Large number of small-sized nodes. Fraction of nodes (α) are equipped with the additional energy . The nodes are uniquely numbered, randomly distributed and not mobile. Nodes can control their transmission power levels and have the capability to directly reach the sink. Sink is not energy-limited and is located at the center of sensor networks. The coordinates of the sink is known to every node in the network. Multimedia & Networking Lab Multimedia & Networking Lab
Energy (M, d) = Et x M + Etckt x M x d2 System Model The energy consumption model can be given by equation To transmit M bits across a distance d Energy (M, d) = Et x M + Etckt x M x d2 To receive a message from a distance d Energyrcvd = Etckt x M Et : Transmission energy. Etckt : Energy to run transceiver circuit Multimedia & Networking Lab
NAEEC - Introduction Network operates in rounds. Length application specific Distribute the node role, and further the cluster-head role Reflect the network changes on the protocol execution Executed in the beginning of a round to build the initial intelligent infrastructure. Consist of phases namely setup phase, configuration phase and execution phase. Each phase consists of a number of sub-phases Fixed length Number of sub-phases depends on the termination condition achieved in a phase. Multimedia & Networking Lab
Time Line of NAEEC Protocol Round 3-Phases Setup Phase Configuration Execution STA CTA ETA Time Line of NAEEC Protocol Multimedia & Networking Lab
Types of messages Introduction message Reply message Beacon message Energy Level (EL) value Node Id Reply message Beacon message Time slot information Multimedia & Networking Lab
NAEEC - Setup Phase Use threshold value to divide the nodes into Candidate cluster-head Non-cluster head Candidate cluster-head: Broadcast introduction message. Receive introduction message from other nodes. Non cluster-head: Receive introduction message Node stores EL value. Look for the node with highest EL value. Multimedia & Networking Lab
NAEEC - Setup Phase continue… At End of Setup Phase: Division of Candidate cluster-head nodes Cluster Heads (A): Sensor nodes directly communicating with the sink or base station. Cluster Heads (B): Sensor nodes communicate with level (A) Cluster Heads. Each node knows its next hop communicating node. Multimedia & Networking Lab
NAEEC - Configuration Phase Cluster Heads (A): Receives the reply messages. Cluster Heads (B): Send the reply message. Non Cluster Heads: At the end of Configuration phase Cluster-head knows the size of the cluster head, and its cluster members. Multimedia & Networking Lab
NAEEC - Execution Phase Cluster Heads (A): Broadcast the beacon messages. Cluster Heads (B): Listen to receive the beacon messages. Non Cluster Heads: At the end of Configuration phase Every cluster-head and non-cluster head node knows its contention free time slot. Multimedia & Networking Lab
Simulation Setup N = 300 to 800. Area of A = 400 meter square. Etckt = 50nJ/bit Et = 100pJ/byte Message size, M =400 bytes Communication range is 30 meters Threshold energy is 6J Fraction of Nodes Energy Level α1 = 90% EL1 = 6J α2 = 5% EL2 = 9J α3 = 3% EL3 = 8J α4 = 2% EL4 = 7J Multimedia & Networking Lab
Simulation Results Energy Requirement (E) Vs Field Area (A). Multimedia & Networking Lab
Simulation - Continue Energy Requirement (E) Vs Number of Sensor Nodes (N). Multimedia & Networking Lab
Simulation -Continue Number of Dead Node Formed in a Round. Multimedia & Networking Lab
Conclusion Proposed an energy-efficient deterministic distributed clustering algorithm for heterogeneous wireless sensor networks. Unlike LEACH and other heterogeneous clustering protocols, NAEEC provides: Efficient distribution of cluster-heads Reduce the energy requirement Number of dead nodes formed over time. Multimedia & Networking Lab
Our Contribution A distributed deterministic clustering protocol that Evaluates and compares the node characteristics with its neighbors to decide its role in the network. Always makes sure that the high energy nodes are chosen as cluster-heads . Provide efficient distribution of sensor nodes. Reduces the number of dead nodes formed over time. Multimedia & Networking Lab
Future Work Designing of the MAC layer We are working on the MAC layer design that support the execution of the NAEEC protocol. Multimedia & Networking Lab
Appendix Multimedia & Networking Lab
Start IS EL > TH ? No Listen Channel Yes EE or OO ? RTS/CTS ? Broadcast IM No Listen Channel Try for RTS-CTS No + Sub-Phase Ends Multimedia & Networking Lab
Yes Start Is CH- A ? No Try Again RTS-CTS Send Reply CH- B IS EE or OO ? RTS/CTS ? Listen Channel No + Sub-Phase Ends
No Yes Wait Start Is CH- A ? CH- B IS EE or OO ? RTS/CTS ? Try Again RTS-CTS Broadcast BM No + Sub-Phase Ends Listen Channel
Thank you!! Questions Multimedia & Networking Lab