Outline

1. D-STAR MAC Protocol: a Cross Layer Solution for Wireless Sensor Networks Endowed with Directive Antennas Gianfranco Manes, Romano Fantacci, Francesco Chiti, Michele Ciabatti, Giovanni Collodi, Davide Di Palma, Ilaria Nelli, and Antonio Manes Department of Electronics and Telecommunications University of Florence

2. Outline Outline • Motivations • WSNs • Directive antennas • Proposed protocol • Discovery phase • Regime phase • Performance analysis • Conclusions

3. 1. Motivations Wireless Sensor Networking • Enabling technology to the aim of intelligent environments instrumenting • Affordable solution to some challenging problems: • environmental sensing, • productive chains control, • real-time phenomena monitoring, • safety and rescue application. • WSNs represent a special case of the more general wireless Ad Hoc networking paradigm with additional constraints: • limited energetic, storage, processing and communication capabilities • low degree of mobility • presence of a small number of sinks

4. 1. Motivations Wireless Sensor Networking • This challenge might be got over through careful system design with particular regard to the communications protocols: • MAC layer: • management of both sleep and active power states • PHY layer: • introduction of directional antennas and their integration within the communications framework

5. 1. Motivations Directive Antennas • Expected benefits • Antenna gain maximization towards desired directions, concentrating energy in smaller area: • Transmitted power decreasing • Power consumption reduction • Network life time increasing • Received power increasing • Coverage range increasing • Error probability reduction

6. 1. Motivations Directive Antennas • Expected benefits • Radiation towards undesired directions minimization • Interference caused by other transmissions reduction • Co-channel interference mitigation • multiple-access problems mitigation • Collision probability reduction • Adaptability to time varying communication conditions: • Mote’s failure • Channel errors • Congestion

7. 1. Motivations Directive Antennas • Criticalities • Cost and size • On board integration • Beam switching management: • Set-up phase signaling overhead • Latency for end-to-end communications setting-up • Algorithm complexity • Switching agility • Support both to synchronous (source-initiated) and asynchronous (event-based) sensing

8. 2. Proposal Directive-Synchronous Transmission Asynchronous Reception (D-STAR) • STAR MAC concept for time synchronization [Chiti et al. in Proc. of IEEE ICC’06] • Cross-layer protocol design (MAC+PHY) • Space-time synchronization • Suitable for flat topology • Scalable

9. 2. Proposal Directive-Synchronous Transmission Asynchronous Reception (D-STAR) • Hypothesis • Beam width = θ [rad] • N possible angular sectors: • N = 2π/θ • Quasi ideal switching • Tswitch<< Tpkt • θ = π/2[rad] • N = 4 • Tswitch ≈ μs << Tpkt ≈ ms

10. 2. Proposal D-STAR: State Diagram nf < Nfd Switch on Battery < Battery_low INIT DISCOVERY OFF Emptysectors = Ns nf=Nfd REGIME Battery < Battery_low 1 < EmptySectors < Ns

11. 2. Proposal D-STAR: Discovery phase • Duty cycle δ= 100% • Listening mode for a time interval Tset-up • Beacon (HELLO) broadcasting: • 1 beacon  angular sector • Node ID and Phase (φ) transmission (time to the next awakening) • Waiting for a fixed time duration τs in search of reply messages • Switching to the following angular sector • Exit condition: • (Tset-up is expired, i.e., Nfd frame periods)

12. HELLO HELLO HELLO HELLO HELLO HELLO … 2. Proposal D-STAR: Discovery phase δ= 100% time Tset-up

13. 2. Proposal D-STAR: Regime phase • Each node sends: • HELLO messages to known neighbors belonging to different angular sectors according to the phase transmitted in previous HELLO messages, • several HELLO messages in background with the proper period to unknown neighbors in the empty angular sector. • Upon the replying of a node, a logical channel is established: • The channel access is managed by means of a CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance) approach. • Each node remains in the regime phase until there is at least one neighbor, otherwise if the number of empty angular sector is equal to Ns it re-enters the discovery phase in search of connectivity.

14. HELLO HELLO HELLO HELLO HELLO HELLO HELLO * FWD HELLO + FWD … δ Tf 2. Proposal D-STAR: Regime phase 1 HELLO broadcasting per sector … … listening sleep time Tf

15. 3. Performance Operative hypothesis • FP6-IST-1-508744-IP ‘‘GoodFood’’ reference scenario • Monitored area = 25 ∙ 25 [m2] • Number of WSN nodes: [10…50] • N = 1, 2 ,4, 6, 8 angular sectors • HELLO pkt length = 8 bytes • Bit rate = 250 kbps • Packet Error Rate = 5% • Number of channel sensing for CSMAC/CA algorithm = 6 • Frame period: 10,25,50,75,93 s • Duty-cycle: [1…5]%

16. 3. Performance Università degli Studi di Firenze Network lifetime vs nodes

17. 3. Performance Network lifetime vs duty-cycle

18. 3. Performance Network lifetime vs frame period

19. Occupied Channel Probability Collision Probability Number of nodes Number of nodes 3. Performance Hidden node analysis R' R R S' S' S S

20. Overhead [pkt/Tf] Normalized Throughput Number of nodes Number of nodes 3. Performance Efficiency and complexity Signaling Overhead Resource Utilization

21. 4. Conclusions Conclusions • Energy efficient WSN protocol design • Cross-layer approach (MAC+PHY) • Performance: • Energetic consumption • Lifetime • Latency • Collision probability • Signaling overhead • Good tradeoff • Easy to implement