Differentiated Surveillance for Sensor Networks

Document Sample
Differentiated Surveillance for Sensor Networks Powered By Docstoc
					Differentiated Surveillance for
Sensor Networks
  Ting Yan, Tian He, John A. Stankovic




                              CS294-1
                              Jonathan Hui
                              November 20, 2003
Idea
   Exploit node
    density/redundancy
    to maximize
    effective network
    lifetime.
   Degree of coverage
    matters!
       Sensing constraints
       Fault tolerance

                              2
Assumptions
   Static placement
   Localization
   Time Synchronization
   Uniform expected node lifetime
   For simplicity of describing protocol?
       Nodes on 2D plane
       Circular sensing radius r
       Communication range > 2r
                                             3
       Basic Protocol
            Initialization Phase
                Localization, Time Sync, Determine
                 Working Schedule (T, Ref, Tfront, Tend)
            Sensing Phase
                Nodes power on and off based on working
                 schedule
                 Round 0 (T)     Round 1 (T)               Round i (T)
                     Ref             Ref                       Ref



                 Tfront   Tend   Tfront   Tend             Tfront   Tend
                                                                           t

Init Phase                                 Sensing Phase                       4
Basic Protocol
Determining Working Schedule

   Goal: Each node determines its own
    working schedule such that all points
    within sensor coverage are covered for
    all time.
   Approach: Represent sensor coverage
    with grid of points


                                             5
      Basic Protocol
      Determining Working Schedule

         Reference Point Scheduling Algorithm
              Randomly choose Ref from [0, T) and broadcast to
               all nodes within 2r.
              For each discrete point
                   Order neighboring Ref times and calculate
                          Tfront = [Ref(i)-Ref(i-1)]/2
                          Tend = [Ref(i+1)-Ref(i)]/2
              Final schedule = union of schedules for all points
                      RefC RefA             RefB
          Point 1                                          RefC RefA          RefB
Node A:
                    RefD      RefA   RefE                 RefD         RefE
          Point 2
                                                                                     6
            Enhanced Protocol
            with Differentiation

    Working schedule for a desired                     Example (a = 1)
                                                                              RefB   RefC
                                                              RefA
     coverage of degree a.
         (T, Ref, Tfront, Tend, a)                       A
         Working period defined as:                      B
              Power On:    T  i  Ref  T front  a
                                                          C
              Power Off:   T  i  Ref  Tend  a

    Example (a = 2)                                     Example (a = 3)
          RefA                  RefB   RefC                   RefA            RefB   RefC

      A                                                   A
      B                                                   B
      C                                                   C          Uh-Oh!
                                                                                            7
Design Issues
   Possible blind spots with discrete points
       Choose points within sensing range conservatively
   Offset in time synchronization
       Power on (off) slightly earlier (later)
   Irregular sensing regions
       Okay, as long as sensing regions of neighboring nodes are
        known
       But also requires knowledge of orientation
   Fault Tolerance
       Awake nodes use heartbeat messages to detect failed nodes
       If a node fails, wakeup all nodes within 2r and reschedule.
       What if communication range < 2r?
                                                                    8
Extensions and Optimizations
   Second Pass Optimization
       After determining working schedule,
        broadcast schedule to all nodes within 2r.
       The node which has the longest schedule:
            Minimize Tfront and Tend while maintaining
             sensing guarantee
            Rebroadcasts new schedule
       Done when every node has recalculated
        schedule or when no more can be done.

                                                          9
       Extensions and Optimizations
           Multi-Round Extension for Energy
            Balance
                Calculate M schedules each with different
                 Ref values during Init Phase.
                Rotate schedules during Sensing Phase.
    RefC RefA RefB    RefB   RefC RefA RefA RefB   RefCRefA     RefC       RefB

A
B
C
                                                                              10
     Schedule 1         Schedule 2         Schedule 3         Schedule 4
Evaluation
   Simulation parameters
       Nodes distributed randomly with uniform
        distribution in 160mX160m field.
       Results taken from center 140mX140m to avoid
        edge effects
       Sensing range = 10m
       Communication range = 25m
       Ideal conditions
       Fault tolerance included?
   Compare against sponsored approach
                                                       11
Evaluation
                Total energy
                 consumption nearly
                 constant with changes
                 in density.




                Variation in total energy
                 consumed decreases
                 with greater densities.
                What’s happening with
                 the sponsored
                 approach?
                                         12
Evaluation
                Half-life increases
                 linearly as density
                 increases.




                       Coverage
                        provided for
                        longer period of
                        time.



                                       13
Evaluation
                Energy consumption
                 increases linearly with
                 different degrees.
                Energy consumption
                 constant with different
                 densities.



                Degree of coverage
                 provided >= a.
                a only guarantees a
                 lower bound.


                                       14
Comments
   Localized algorithm?
       But still requires time synchronization and doesn’t support
        mobility
   Inflexible
       mobility not supported, schedules are fixed
   No notion of the “goodness” of a node
       Nodes that have more energy should take up a larger
        portion of the working schedule
   Difficult to reliably broadcast Ref values to all 2r
    neighbors in a dense network
       Only have one chance to get it right!
       Worse in cases where communication range < 2r (i.e.
        acoustic sensors)

                                                                      15
Comments
   Working schedules determined without taking other
    schedules and protocols into account
       How does it affect other protocols (i.e. TDMA)?
   Comparison to Sponsored Coverage unfair
       Sponsored Coverage supports fault tolerance, limited
        mobility, and is more adaptable
   Ability to specify degree of coverage
       But current algorithm doesn’t correctly guarantee with a >
        2!
   Fault tolerance relies on communication range > 2r
    for heartbeat messages

                                                                     16
Conclusion
   Pros
       Improved performance in lifetime and workload
        balance
       Specify a degree of coverage
   Cons
       No upper bound on degree estimation
       Inflexible
            Static working schedule, static nodes, time
             synchronization, reliable communication


                                                           17

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:0
posted:3/24/2013
language:English
pages:17