1 / 21

Performance Tradeoffs Among Percolation-Based Broadcast Protocols in Wireless Sensor Networks

Performance Tradeoffs Among Percolation-Based Broadcast Protocols in Wireless Sensor Networks. Vijay Raman and Indranil Gupta University of Illinois at Urbana-Champaign. Introduction. Broadcast in wireless sensor networks – a fundamental operation Used for:

keitha
Download Presentation

Performance Tradeoffs Among Percolation-Based Broadcast Protocols in Wireless Sensor Networks

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Performance Tradeoffs Among Percolation-Based Broadcast Protocols in Wireless Sensor Networks Vijay Ramanand Indranil Gupta University of Illinois at Urbana-Champaign WWASN 2009

  2. Introduction • Broadcast in wireless sensor networks – a fundamental operation • Used for: • Sending code updates, e.g. Trickle • Initiating aggregation, e.g. TAG, directed diffusion • Sensing route updates • Membership/multicast group management • Several specific algorithms to address the broadcast problem • Pure gossiping • Trickle • PBBF (probability-based broadcast forwarding)

  3. Motivation • Broadcast applications have varied requirements • Code updates – reliability • Information query propagation – time sensitivity • Membership management – energy minimization • Which protocol to use for what application? • How do the broadcast protocols compare with each other?

  4. Flooding - Simplest Broadcast Protocol • Packets simply transmitted over the air • Any node that receives the packet repeats the process S Not all transmissions shown

  5. Flooding • In a N node network, k neighbors per node (average) • 1 message broadcast per node • k messages received per node on average • (3 + k).Pow power consumed per node • Pow - power consumed per message reception • Power transmit = 3.power received

  6. Percolation-Based Broadcast Protocols • A probabilistic model for broadcast • Not all nodes involved in broadcast • Exact mechanism depends on the actual scheme used • Three categories discussed in this presentation • Site Percolation • Bond Percolation • Modified-bond Percolation

  7. Site Percolation • A gossip mechanism • Every node decides to broadcast based on some probability ? ? ? ? ? ? S ? ? ? ? ? ? ? ? ? Not all transmissions shown

  8. Site Percolation • If p is the probability that a node broadcasts • p messages broadcast per node on average • kp messages received per node on average • (3 + k).p.Pow powerconsumed per node • Pow - power consumed per message reception • Power transmit = 3.power received

  9. Bond Percolation • An alternative gossip mechanism • Nodes pick a subset of neighbors and unicast packets only to the chosen nodes S Not all transmissions shown

  10. Bond Percolation • If m out of k neighborsare chosen • m unicastmessages are sent per node on average • m messages received per node on average • (3m + m).Pow = 4m.Pow power consumed per node • Pow - power consumed per message reception • Power transmit = 3.power received

  11. Modified Bond Percolation • Same principle as bond percolation • Packets are broadcast (not unicast) • # neighbors receive broadcast depending on their sleep cycles • If m out of k neighborsare chosen • 1 message sent per node on average • m messages received per node on average • (3 + m).Pow power consumed per node • Pow - power consumed per message reception • Power transmit = 3.power received

  12. Power Consumption Summary (3+k).pow (3+k).p.pow Flooding Site Percolation >= >= >= <= => Bond Percolation Modified Bond Percolation >= 4m.pow (3+m).pow

  13. Power Consumption Summary • Modified bond percolation consumes the least power per broadcast • Bond percolation consumes the highest power followed by flooding • Site percolation consumes least power for sparse networks

  14. Simulation Model • Three topologies – random, clustered, and grid • Number of nodes varied from 1000 to 2500 • Grid side (for grid topology) varied from 30 to 50 • MAC independent link layer based on actual traces • For site percolation, p = 0.7 • For bond percolation, m/k = 0.5

  15. Power Consumed – Random Topology Average power # Nodes Modified bond percolation Bond percolation Site percolation Flooding

  16. Broadcast Latency – Random Topology Average latency # Nodes Modified bond percolation Bond percolation Site percolation Flooding

  17. Latency-Power Tradeoff – Random Topology Average power Average latency Modified bond percolation Bond percolation Site percolation Flooding

  18. Tradeoff Summary • Flooding – More messages, more power, less latency • Site Percolation – Fewer messages, low power (in sparse networks), less latency • Bond percolation – Unicast messages, more power, high latency • Modified bond percolation – Few messages, least power, high latency

  19. Where do they fit? • Code updates – reliability, low latency • Any of the percolation-based schemes (right parameters!) • Information query propagation – time sensitivity • Flooding or site percolation • Membership management – energy minimization • Modified bond percolation (dense) • Site percolation (sparse)

  20. Conclusion • Broadcast protocols may have varied performance • Latency, power, number of messages propagated • Applications have varied requirements • Code updates, query propagation, membership management • Choosing the right broadcast protocol for the right application important • Code update – percolation • Query propagation – Flooding, site percolation • Membership management – Mod. bond or site percolation • Preferred match of protocols and applications shown through simulations and analyses

  21. Thank You! WWASN 2009

More Related