Loading in 2 Seconds...
Loading in 2 Seconds...
A Quorum-Based Energy-Saving MAC Protocol Design for Wireless Sensor Networks Chih-Min Chao, Yi-Wei Lee IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, 2010
Outline • Introduction • Preliminaries • Protocol description • Simulation results • Conclusions
Introduction • Wireless sensors are battery powered. • It is crucial for them to efficiently use their battery resources. • Most of the existing power-saving protocols achieve power savings by periodically putting sensor nodes to sleep. • Lower power efficiency • Higher latency.
Introduction • Many protocols have been proposed to extend the network lifetime of sensor networks • Deployment protocols • Power efficientmedium access control protocols • Routing protocols • Energy-hole problem • Sensor nodes that are closer to the sink deplete their power faster.
Problem Statement Host A Host B Time 8 3 4 7 5 6 0 1 2
Quorum Concept n Ra n Rb Ca Cb Host A 8 3 4 7 5 6 0 1 2 Time Host B
QMAC • To reduce power consumption and determine the sleep frequency for each sensor node based on its own traffic load. Time frame SIFS SIFS SIFS RTS DATA A Sleep B CTS ACK C Sleep D Sleep E Time
Preliminaries • Time is divided into a series of time frames. • All sensor nodes are time synchronized. • Each node has a unique ID. • Sensor nodes report their data to their common sink node. • All sensor nodes have the same transmission range. • All sensor nodes are static after deployment.
Preliminaries • Sensor nodes are randomly and uniformly distributed in the network area. Hop Count=1 Hop Count=2 Sink Hop Count=3 C1 Hop Count=4 C2 C3 C4
Quorum-Based Wake-Up Schedule • A sensor node using an n × n grid will wake up 2n − 1 out of n2 time frames. • Grid size The ratio of areas for different coronas C1:C2:C3:C4=1:3:5:7 Anode in C3 is responsible for relaying traffic for 7/5 nodes in C4
Quorum-Based Wake-Up Schedule The traffic load in C1is 1 + 3 × 5 = 16 The traffic load in C2is 1 + (5/3) × 2.4 = 5 The traffic load in C3 is 1 + (7/5) × 1 = 2.4 The ratio of areas for different coronas C1:C2:C3:C4=1:3:5:7
Latency Reduction • In allowing sensor nodes to sleep longer than one time frame to reduce energy consumption. • The price for this saved energy, though, is higher latency. • Next-hop group
Next-hop group C4 C3 C2 C1 Sink X One hop neighbor boundary Next hop group member
Simulation results • NS2 • DMAC • The shortest transmission latency • PMAC • An adaptive energy-saving protocol
Simulation results Fig. 9. Effect of different MAC protocols on the fraction of live sensor nodes at different coronas at simulation times of (a) 100 s, (b) 200 s, and (c) 300 s.
Simulation results The number of nodes for networks with 3, 4, and 5 coronas is 225, 400, and 625.
Conclusions • The sensor nodes have different loads due to their different distances to the sink • The concept of quorum to enable sensor nodes to adjust their sleep durations based on their traffic loads. • QMAC • QMAC_LR • Simulation results verify that our QMAC_LR reduces energy consumption and keeps the latency low.