Routing and Clustering

1
Routing and Clustering

• Xing Zheng
• 01/24/05

2
References

• Routing
• A. Woo, T. Tong, D. Culler, "Taming the
  Underlying Challenges of Reliable Multihop
  Routing in Sensor Networks," ACM SenSys 2003.
• LEACH
• W. R. Heinzelman, A. Chandrakasan, H.
  Balakrishnan," Energy-efficient communication
  protocol for wireless microsensor networks,"
  HICSS 2000.
• HEED
• O. Younis, S. Fahmy, "Distributed Clustering in
  Ad-hoc Sensor Networks A Hybrid,
  Energy-Efficient Approach ," IEEE Infocom 2004.

3
1

• Taming the Underlying Challenges of Reliable
  Multi-hop Routing in Sensor Networks

4
Routing Issues in WSN

• Substantially different from traditional ad-hoc
  wireless networks
• Traditional setting
• Assume 802.11 links (Abstract away the underlying
  physical layer and MAC protocol)
• Independent pair-wise connections
• Abstract the applications
• Sensor Networks
• Resource-constrained nodes
• Low-power radios
• Multi-hop aggregation
• Application-specific communication pattern

5
Underlying Factors

• Connectivity graph
• Discovered by nodes observing communication
  events and sharing the information
• Connectivity
• A statement of the likelihood of successful
  communication
• Nodes
• Nearby nodes may be in communication most of the
  time, but not always.
• Less reliable communication with distant nodes,
  but a few may have strong connectivity
• Lossy links and dynamic loss rates

6
About this study

• Routing algorithms should take into account these
  underlying factors and be evaluated in concert
  with the low level estimation mechanisms under
  realistic loads.
• Stages
• Empirical link characteristics
• Link estimation
• Neighborhood table management
• Routing protocol
• Target application
• A large collection of nodes route periodically
  sampled data over multiple hops to an individual
  sink.

7
Link Characteristics

• Set up a platform to measure loss rates between
  many different pairs of nodes at different
  distances
• Observations suggest a simple means of capturing
  probabilistic link behavior in simulations
• Create a link quality model
• For each directed node pair at a given distance
• A link probability is associated based on the
  mean and variance extracted from the empirical
  data.
• Each simulated packet transmission is filtered
  out with this probability.

8
Empirical Results
9
Link Estimation

• Individual nodes estimate link quality by
  observing packet success and loss events.
• Link quality is used in routing protocols cost
  metrics.
• Requirements
• React quickly to potentially large changes in
  link quality
• Stable
• A small memory footprint
• Simple to compute

10
WMEWMA

• Based on snoopy techniques
• Passive probing
• Loss can be inferred by tracking the sequence
  numbers.
• Window mean with EWMA
• Based on historical observations
• Compute an average success rate over a time
  period
• Can track the empirical trace fairly well

11
Neighborhood Management

• Neighborhood table
• Record information about nodes from which it
  receives packets
• Limited size
• Question How does a node determine which nodes
  it should keep in the table?
• To seek a neighborhood management algorithm that
  will keep a sufficient number of good neighbors
  in the table
• Similar to cache management

12
Management Policies

• Insertion
• Upon hearing from a non-resident source
• Adaptive down-sampling technique
• The probability of insertion the neighbor table
  size / the number of distinct neighbors
• Eviction
• RR, FIFO, Least-Recently Heard, CLOCK, etc.
• Reinforcement
• FREQUENCY algorithm
• A frequency count for each entry in the table
• Reinforce good neighbors during insertion

13
Routing Framework
14
Routing protocol

• Distance-vector based algorithms
• Parent selection
• Access the neighborhood table to select a set of
  potential parents
• MT (Minimum Transmission) cost metric
• the expected number of transmissions along the
  path
• For each link, MT cost is estimated by 1/(Forward
  link quality) 1/(Backward link quality).

15
Evaluation Remarks

• Link quality estimation and neighborhood
  management are essential to reliable routing.
• Minimum expected transmissions is an effective
  metric for cost-based routing.
• The combinations of these techniques can yield
  high end-to-end success rates.

16
2

• Energy-Efficient Communication Protocol for
  Wireless Micro-sensor Networks

17
LEACH

• Low-Energy Adaptive Clustering Hierarchy
• Designed for minimizing energy dissipation in
  sensor networks
• Model of sensor networks
• Base station fixed and far located from sensors
• Nodes homogeneous and energy-constrained

18
Conventional Approaches

• Directional vs. multi-hop
• Short system lifetime

19
Clustering

• LEACH
• Self-organized adaptive clustering protocol
• Key features
• Localized coordination and control for cluster
  set-up and operation
• Randomized rotation of the cluster heads and the
  corresponding clusters
• Local compression to reduce global communication

20
Algorithm

• Run by rounds
• Advertisement Phase
• A node becomes a cluster head if Random(0,1) lt
  T(n), which is a threshold in the system.
• Cluster heads broadcasts an advertisement message
  using a CSMA MAC protocol.
• Non-cluster-head nodes decide to join the cluster
  with the largest signal length heard from its
  head.

21
Algorithm (cont.)

• Each node reports to its cluster head using a
  CSMA protocol.
• Based on all the messages received within the
  cluster, the head node creates a TDMA schedule
  for intra-node transmission.
• During data transmission, non-cluster-nodes can
  be turned off until the nodes allocated
  transmission time.

22
Strengths

• Dynamic cluster distribution
• Extend system lifetime

23
Weaknesses

• Assumes uniform energy consumption for cluster
  heads in cluster rotation.
• Does not guarantee a good cluster head
  distribution
• Randomly selection of heads can result in faster
  death of some nodes.

24
3

• Distributed Clustering in Ad-hoc Sensor Networks
  A Hybrid, Energy-Efficient Approach

25
HEED

• Hybrid Energy-Efficient Distributed Clustering
• Design goals
• Prolonging network lifetime by distributing
  energy consumption
• Terminating in O(1) iterations
• Minimizing low control overhead
• Producing well-distributed cluster heads and
  compact clusters

26
Clustering Parameters

• For electing cluster heads
• Primary parameter residual energy (Er)
• Secondary parameter communication cost (used to
  break ties), i.e., maximize energy and minimize
  cost

27
Algorithm at node v

• Initialization

• Discover neighbors within cluster range
• Compute the initial cluster head probability
  CHprob f(Er/Emax)

• If v received some cluster head messages, choose
  one head with min cost
• If v does not have a cluster head, elect to
  become a cluster head with CHprob .
• CHprob min(CHprob 2, 1)
• Repeat until CHprob reaches 1

• Main processing

• If cluster head is found, join its cluster
• Otherwise, elect to be cluster head

• Finalization

28
Example
Discover neighbors
(0.4,3)
(0.6,2)
a10
(0.1,4)
a13
c2
a11
(0.2,2)
(0.2,5)
a7
Compute CHprob and cost
a8
(0.5,3)
a12
(0.2,3)
c3
(0.2,3)
a9
(0.8,4)
(0.1,4)
Elect to become cluster head
(0.1,2)
c1
a5
a6
(0.9,4)
a14
(0.5,4)
a2
c4
Resolve ties
a4
(0.6,4)
(0.3,2)
a3
(0.7,5)
(0.2,3)
Select your cluster head
(0.3,2)
a1
29
HEED vs. LEACH

• Longer lifetime
• Less energy consumption

30
Conclusions

• Hybrid approach
• Heads are selected based on residual energies
• Nodes join cluster to minimize communication cost
• Terminates in a constant number of iterations
• Independent of network diameter
• Location-unaware
• Prolongs system lifetime