Asymmetric Links in a Mobile Environment

1
Asymmetric Links in a Mobile Environment

• By
• Yaser Khamayseh
• Kavita Jari

2
Agenda

• Introduction
• Motivation
• Simulation Methodology
• Results
• Questions

3
Questions
4
Mobile Environment

• Mobility
• High packet loss
• Lower Bandwidth
• Power constraints
• Range of cells
• Interference

5
CSMA/CA

• Carrier Sense Multiple Access/Collision Avoidance
• Collision avoidance before transmission
• Carrier sense
• Each station listens for transmission before
  transmitting
• Multiple access
• Each station is free to transmit when it has
  data to send
• Collision avoidance
• Each station listens for any media activity
  start transmission if the channel is free

6
Hidden Node Problem
S2
S1
R
S3
7
RTS/CTS

• Peer-to-Peer Protocol
• The node sends an RTS to its neighbor and waits
  for a CTS.
• Once it has received the CTS, it starts sending
  the data.
• Wait for ACK.

8
RTS/CTS
RTS
DATA S1 ? R
R
S1
CTS
S1
ACK
R
Busy
S2
Busy
S3
9
Symmetric Links
Example of a connected graph
10
Symmetric Links
Example of a disconnected graph
11
Asymmetric Links

• Each node has two radii (ranges)
• Transmitting Radius
• Receiving (sensing) Radius
• Transmission radius ? Sensing radius
• Two graphs to represent the state of the network

12
Asymmetric Links
Sensing Graph
Node 0
Node 1
Node 2
Node 1
Node 0
Transmitting Graph
Node 0
Node 1
Node 1
Node 0
13
Motivation
Does send/receive asymmetry
break our mobile protocols???
14
Simulation Methodology

• Rectangular Area
• Parameters and Factors
• Random way-point movement model
• Events
• Performance Metrics

15
Parameters Factors

• The size of the rectangular area is 1500 300
• Number of Nodes are 10, 20 or 50
• Different transmission and sensing radii
• 50 .. 300
• Different combinations will be tested
• Node pause time based on exponential distribution
• Node speed based on uniform distribution

16
Simulation Methodology
1500 m
(X,Y)
300 m
17
Simulation Methodology
1500 m
(NX,NY)
Constant Speed
(X,Y)
300 m
18
Simulation Methodology
1500 m
(NX,NY)
(tX,tY)
(tX,tY)
300 m
19
Simulation Methodology
1500 m
(NX,NY)
(tX,tY)
(tX,tY)
300 m
20
Simulation Methodology
1500 m
(NX,NY)
Pause Time
(tX,tY)
300 m
21
Events

• Start moving
• Nodes move to a new location with a specific
  speed
• Stop moving
• Nodes pause for a specific time
• Add an edge
• Connection is established between two nodes
• Remove an edge
• Connection is removed between two nodes

22
Performance Metrics

• Graph Connectivity
• Effect of different radii (Transmitting vs.
  Sensing)
• What is the best definition for graph
  connectivity in the case of asymmetric links?
• Each node in the graph can send and receive from
  every other node in the graph.

23
Expectations

• With increase in sensing radius ? Better
  collision avoidance.
• The graph connectivity maybe better !
• With increase in transmitting radius ?
• More collision

24
Previous Work

• Increasing Transmission radius
• Increased connectivity
• Increase in interference
• Loss of channel throughput
• Decreasing Transmission radius
• Increase number of hops to destination
• Transmission radius covering 6 hops will achieve
  a throughput proportional to sqrt(No. of nodes).

25
Conclusion The Goals

• Increasing connectivity
• Maximizing throughput
• Optimizing (minimizing) energy consumption
• Decreasing the number of hops (delay)

26
References

• 1 Kleinrock Silvester, Optimum transmission
  radii for packet radio networks, 1978
• 2 Sanchez, Giles, Zander, CSMA/CA with beam
  forming antennas in multi-hop packet radio
• 3 Ravi Prakash, Unidirectional links prove
  costly in Wireless Adhoc networks, 1999
• 4 Sanchez, Manzoni, Determination of critical
  transmission range in ad-hoc networks
• 5 Royer, Melliar-Smith, Moser, An analysis of
  the optimum node density for adhoc mobile
  networks

27
Questions