Energy Consumption and Conversation in Mobile Peer-to-Peer Systems

• Observation: mobile P2P growing. how do we implement it on resource-constrained devices
• Underlying questions:
  • What are the constraints?
  • How do current P2P protocols use these resources
  • Can we make them more friendly?
• Target platform
  • smart-phone/PDA
    • Energy constraint is big challenge
    • always-on characteristic of P2P is a challenge
• Hard to find P2P protocols on embedded devices
  • Simmud, Freepastry, Tapestry
  • Simpastry: C#
• Nothing stable enough
• Chimera: A light-weight P2P protocol
  • Chimera-Chat: generic command-line chat app
• Stargate eval bench with power supply and ADC
  • CPU utilization low (<4%)
  • Wireless utilization low (70% idle)
• Not really any ground-breaking discoveries

Energy-Aware Self-organizing Density Management in Wireless Sensor Networks

• Context
  • Area monitoring (monitoring/instrusion detection)
  • sensor redundancy exists
  • human intrvention not feasible
• Goal: reduce node activity, balance load over ntwork, deploy fully decentralized **solution: Propose SAND
• GAF and SPAN do no exploit absence of traffic in active sensing state
• Area coverage problem/Clustering: high comm overhead
• Goal of SAND: be simpl, limit messages exchanges, take advantage of redundancy and collaborative nature, be simple
• Switches between Router/Sensor/Gateway/Sleep states, based on simple algorithm
• Not compared with any other algorithms except "Without SAND"