Leader Election

1
Leader Election

• Bernard Chen

2
Outlines

• Introduction to leader election
• Complete Topology
• Logical Ring Topology
• Tree Topology
• Secure Extrema Finding Algorithm (SEFA)
• References

3
What is leader selection?

• The existence of a centralized controller greatly
  simplifies process synchronization
• However, if the central controller breaks down,
  the service availability can be limited
• The problem can be alleviated if a new controller
  (leader) can be chosen

4
When do we need leader election?

• 1. The initiation of a system of processes
• 2. When an existing leader fails
• (The detection of failures is normally based on a
  time-out mechanism)

5
Two Types of Election Criteria

• 1. a leader election can be based on a global
  priority. This type of election is called
  extrema finding. (Every node has a fixed
  evaluation value)
• 2. processes in the group can vote for a leader
  based on a nodes preference, its called a
  preference-based leader election algorithm.

6
Leader Election vs. Mutual Exclusion

• In many aspects, leader election and mutual
  exclusion are the same they both try to reach an
  agreement for identifying a unique process
• However, there are some differences

7
Leader Election vs. Mutual Exclusion

• The two major differences are
• A mutual exclusion must ensure that no process is
  starved, while a leader election is more
  concerned with the fast and successful
  termination of the election process
• Another distinction is that the result for leader
  election need to be announced to all processes

8
Outlines

• Introduction to leader election
• Complete Topology
• Logical Ring Topology
• Tree Topology
• Secure Extrema Finding Algorithm (SEFA)
• References

9
Complete Topology

• In a complete topology, each process in the group
  can reach any other process in the same group in
  one message hop.
• Each process as a global priority, and the
  highest-priority is elected leader

10
The Bully Algorithm

• Its a extrema-finding algorithm
• A process starts a leader election if it suspects
  that the coordinator has failed.
• If a process Pheavy receives an election message
  from a lighter process Plight, it sends a
  take-over message to Plight. Plight is out of the
  race.
• If a process doesnt get a take-over message
  back, it wins, and sends a victory message to all
  other processes.

11
The Bully Algorithm
The bully election algorithm Process 4 holds an
election Process 5 and 6 respond, telling 4 to
stop Now 5 and 6 each hold an election
12
The Bully Algorithm

• Process 6 tells 5 to stop
• Process 6 wins and tells everyone

13
Outlines

• Introduction to leader election
• Complete Topology
• Logical Ring Topology
• Tree Topology
• Secure Extrema Finding Algorithm (SEFA)
• References

14
Logical Ring Topology

• A logical ring is easy to construct and offers
  the unique property that a message initiated by
  any node will return to the node, indicating
  completion of a round of broadcast

15
Leader Election

• Initiation
• A process sends an ELECTION message to its
  successor (or next alive process) with its ID
• Each process adds its own ID and forwards the
  ELECTION message
• Leader Election
• Message comes back to initiator
• Initiator announces the winner by sending
  another message around the ring

16
Ring Algorithm Initiation

17
Ring Algorithm - Election

18
Outlines

• Introduction to leader election
• Complete Topology
• Logical Ring Topology
• Tree Topology
• References

19
Tree Topology

• To construct a logical ring structure is easy if
  the underlying network supports broadcast
  facilities
• However, in Ad Hoc networks, user mobility may
  results in frequent leader election, making the
  process a critical component of system operation

20
Leader election in Ad Hoc networks

• Most of the leader election algorithms for mobile
  ad hoc networks elect a random leader and hence
  are not extrema-finding
• SEFA (Vasudevan S., 2003) is a round-based
  hierarchy-building approach which also considered
  secure problem
• Good reviews can be found in 3(G. Tel. 2000)
  and 4(N.Lynch 1996)

21
Secure Extrema Finding Algorithm (SEFA)

• SEFA is an extrema-finding algorithm in which all
  nodes use a Common Election Algorithm (CEA) to
  determine their rank
• Decision factors might include node identity,
  battery lifes, or level of trustetc
• Larger the CEA value, the more desirable the
  node is a leader

22
Algorithm Example

23
Algorithm Example
24
(No Transcript)
25
Outlines

• Introduction to leader election
• Complete Topology
• Logical Ring Topology
• Tree Topology
• Secure Extrema Finding Algorithm (SEFA)
• References

26
References

1. Chow, Randy, et. al.,Distributed Operating
   Systems Algorithms, Addison Wesley, March 18,
   1997
2. Vasudevan, S. DeCleene, B. Immerman, N.
   Kurose, J. Towsley, D. Leader election
   algorithms for wireless ad hoc networks, DARPA
   Information Survivability Conference and
   Exposition, 2003. Proceedings
3. G. Tel. Introduction to Distributed Algorithms.
   Second Edition, 2000. Cambridge University Press
4. N.Lynch Distributed Algorithms. 1996, Morgan
   Kaufmann Publications
5. V. Park and M. Corson. A Highly adaptive
   Distributed Routing Algorithm for mobile Wireless
   Networks. In proceediing of IEEE INFOCOM, April
   1997

27
Thank You

• Questions??