Customized Mobility Model for MANET routing

1
Customized Mobility Model for MANET routing

• By
• Anuraag Dimri
• (iwc2006001)
• Under Supervision of
• Dr Shekhar Verma.

2
Overview of Presentation

• Mobile Ad-hoc Networks
• Motivation
• Mobility Models
• Customized Mobility Model
• Simulation
• Simulation Results
• Observation
• References
• Questions

3
Why Ad Hoc Networks ?

• Ease of deployment
• Decreased dependence on infrastructure

4
Mobile Ad Hoc Networks

• Formed by wireless hosts which may be mobile and
  arbitarily located
• Without necessarily using a pre-existing
  infrastructure
• Routes between nodes may potentially contain
  multiple hops

5
Many Applications

• Enables, for example, Personal Area Networks
  (PANs)(Bluetooth, IEEE 802.15)
• For military and disaster management.
• Information distribution (meetings, seminars
  etc.)
• Internet / intranet hot spots (public
  transportation)
• New mobile devices are invented constantly and
  used various ways.

6
Two Characteristics that effect protocol
performance.

• 1. TRAFFIC CHARACTERISTCS
• 2.MOBILITY CHARACTERISTICS

7
Traffic Characteristics vary

• Traffic characteristics may differ in different
  ad hoc networks
• bit rate
• reliability requirements
• unicast / multicast / geocast
• host-based addressing / content-based addressing
  / capability-based addressing
• May co-exist (and co-operate) with an
  infrastructure-based network

8
Mobility Variations

• Mobility patterns may be different
• people sitting at an airport lounge
• Taxi cabs
• kids playing
• military movements
• personal area network
• Mobility characteristics
• speed
• predictability
• direction of movement
• pattern of movement
• uniformity (or lack thereof) of mobility
  characteristics among different nodes

9

MotivationREFMultihop Ad Hoc Networking The
Theory Conti, M. Giordano, S. Communications
Magazine, IEEE April 2007

• MANET research generally lacks realism, both
  from the socio economic and technical
  perspective. Many research areas (e.g., security
  and cooperation, energy management, and transport
  protocols) address relevant theoretical problems
  that will have a practical impact only if and
  when MANET has a large-scale deployment
• MANET research focuses on building a complex
  large-scale system with almost no attention to
  building up a user community.

10
Mistake in designing adhoc nwk protocolsvaidya,
infocom 2005Assuming Extreme Scenarioas the
Common Case
11
Extreme Ad Hoc NetworkingLarge Isolated
Networks? No infrastructure
C
E
A
B
12
Extreme Scenario

• Extreme ad hoc networks No infrastructure
• ? No certification authority
• ? No DHCP server
• ? Long-lived partitions

13
More Likely Ad Hoc NetworksAccess to
Infrastructure
internet
C
E
A
B
14
More Likely Ad Hoc NetworksSmall
15
More Realistic Multi-Hop WirelessMesh Networks
internet
Wireless backbone
B
C
A
16
More Realistic Multi-Hop WirelessHybrid Networks
internet
Access Point
Wireless channel
E
B
C
A
D
17
Practical assumptions

• Infrastructure can be accessed selectively
• Not all enumerable scenarios are relevant
• ? Design protocols and study there performance
  for the likely scenarios

18
Realistic mobility modelsREFStudy on
Environment Mobility Models for Mobile Ad Hoc
Network Hotspot Mobility Model and Route
Mobility Model. Gang Lu Belis, D. Manson, G.
Wireless Networks,Communications and Mobile
Computing, 2005 International Conference on
Volume 1,

• Mobility models are random-based
• -they dont model the situations correctly
• Some of the situations the node movement is not
  random but more or less deterministic.
• Authors propose 1.ROUTE MOBILITY MODEL-
  construct complex environment like city area.
  2.HOTSPOT MOBILITY MODEL.-attractive places
  exist in simulation area.

19
Mobility Models Designed to
describe the movement pattern of mobile users,
and how their location, velocity and acceleration
change over time.
20
Dimensions of Mobility Space

• Temporal Dependency
• Physical constraints of the mobile entity
  itself
• e.g the current velocity is more or less
  dependent on the previous velocity, according to
  certain parameter.
• Spatial Dependency
• Movement pattern is influenced by and
  correlated with nodes in its neighborhood.
• Geographic Restrictions
• Movement may be restricted along the street
  or a freeway. A geographic map may define these
  boundaries.

21
RANDOM-BASED MOBILITY MODELS

• Destination, speed and direction are all chosen
  randomly and independently of other nodes
• The Random Waypoint Model
• Mobile node randomly selects one location in
  the simulation field as the destination and
  travels constant velocity chosen uniformly and
  randomly from 0,Vmax
• The velocity and direction of a node are
  chosen independently of other nodes

22

• on reaching the destination, the node stops for a
  duration defined by the pause time parameter
  Tpause. After which it again chooses another
  random destination in the simulation field and
  moves towards it

23

• Vmax and Tpause are two key parameters that
  determine the mobility behavior of nodes.
• Small Vmax Large Tpause gt stable nwk
• Large Vmax Low Tpausegt Highly dynamic
• Random Walk Model
• Special case of Random Waypoint model with
  zero pause time.
• Limitations of Random Models
• Velocity of MN is a memoryless random process.
• MN is considered an entity that moves
  independently of other nodes and can move freely
  within simulation field without any restrictions.

24
MOBILITY MODELS WITH TEMPORAL DEPENDENCY

• the velocities of single node at different time
  slots are correlated.
• Gauss-Markov Mobility Model-
• Smooth Random Mobility Model-probabilities of
  selecting certain speed is higher in the
  range0,Vmax

25
MOBILITY MODELS WITH SPATIAL DEPENDENCY

• In some applications including disaster relief
  and battlefield, team collaboration among users
  exists and the users are likely to follow the
  team leader. Therefore, the mobility of MN could
  be influenced by other neighboring nodes.
• Reference Point Group Mobility Model -each group
  has a center
• movement of the group leader determines the
  mobility behavior of the entire group

26
MOBILITY MODELS WITH GEOGRAPHIC RESTRICTION

• In real life applications, nodes movement is
  subject to the environment. In particular, the
  motions of vehicles are bounded to the freeways
  or local streets in the urban area, and on campus
  the pedestrians may be blocked by the buildings
  and other obstacles. Therefore, the nodes may
  move in a pseudo-random way on predefined
  pathways in the simulation field.
• Obstacle Mobility Model
• Obstacles in the simulation field are present.
  
• MN is required to change its trajectory
• Obstacles also impact the way radio propagates
  

27

• Pathway Mobility Model

28

• Node movement restricted to the pathways in the
  map
• Initially, the nodes are placed randomly on the
  edges of the graph. Then for each node a
  destination is randomly chosen and the node moves
  towards this destination through the shortest
  path along the edges.
• Destination of each motion phase is randomly
  chosen, a certain level of randomness still
  exists for this model. So, in this graph based
  mobility model, the nodes are traveling in a
  pseudo-random fashion on the pathways

29
Customized Mobility Model

• Mobility is a characteristic of device Laptops
  PDAs and Cell phones exhibit different
  mobility
• Emphasis on scenario based simulation.
• Problem with predominantly used Random Waypoint
  Mobility Model
• Does not precisely mimics real world mobility
• Exhibits speed decay
• Suffers with Density wave phenomenon

30
Customized Mobility Model

• Mobility dependent on place and type of device
• Campus may have 1. Buildings
• 2. Cafeteria
• 3. Playground
• 4. Pathways
• 5. Empty spaces with no user density

31
Customized Mobility Model

• Types of devices associated can be enumerated as
  follows.
• 1.Cell phones
• 2.PDAs
• 3.Laptops
• Simulation area based on real map
• Partitioned in to zones exhibiting different
  mobility of its own
• Inter zone transition based on real world
  movements

32
Customized Mobility Model

• Decomposing simulation area in to
• different regions of different
• mobility.
• Unrealistic assumption of randomly
  distributed nodes is avoided.

Campus LAN as composed of different
entities
33
First ModelBased on domain
switching probabilities

• The simulation region is divided in to different
  sub-domains
• Based on domain switching probability
  inter-domain node transition takes place
• Studies on campus LAN have shown repetitive
  associative behavior.
• Modified Random Trip Mobility model27 so that
  it can take multiple probability transition
  matrix files for continuous node movement.
• Takes domain file and probability file as input

34
First Model

• Domain file is of the form
• r 100 100 400 400 15
• c 1000 250 200 5
• r 500 500 900 900 20
• High node density and mobility can be achieved
  by increasing number of nodes, decreasing the
  area and increasing the trips
• Domain transition matrix file probab is
• 0.2 0.6 0.2
• 0.1 0.8 0.1
• 0.2 0.7 0.1

35
First Model

• Cafeteria is depicted by the second column
• Another probability transition matrix file called
  newProbab mimics user movement to cafeteria after
  lunch time0.8 0.0 0.2
• 0.3 0.4 0.3
• 0.2 0.0 0.8
• Thus we are able to depict a situation in which
  nodes move towards the cafeteria and then
  relatively same number of nodes move back

36
First Model

• Let nodes in each domain10
• Based on probab I,II and III end up with 5,21
  and 4 nodes.
• If we model 7 nodes each to move from the II to
  I and III, the probability of files moving is
  7/21 0.3 in both the cases. And probability of
  nodes remaining in cafeteria is 0.4 (1-0.3-0.3).
• Calculate the other probabilities based on the
  node movement..

Thus real world node movement can be modeled .
37
Second ModelChoosing speed and
pause time from Gaussian Distribution

• Partition simulation region in to different
  sub-domains
• These subdomains can have different mobility
  characteristics that may be represented by,
  pursue, brownian and random waypoint mobility
  model
• The setdest program22 picks speed from uniform
  distribution and takes only fixed pause time.

38
Second Model
Uniform vs Gaussian distribution for picking
speed and pause time
39
Second Model
Different subdomains can be assigned different
minimum and maximum speeds, Pause times can also
be varied to model the mobility of different
regions differently
Snapshot of the second model
40
Simulation

• Three simulation exercises to simulateRandom
  Waypoint ModelFirst ModelSecond Model
• Simulation Parameters-simulation area
  1500x1500-number of nodes90-simulation
  time180 sec-type of datacbr-data rate 4mb\s
  per connection-speed 3m\s , pause time4m\s
  (for first simulation)-different speed and pause
  time for diff domains(third simulatin)

41
First simulationRandom
waypoint model
Throughput plot for AODV and DSDV for first
simulation at node 22
42
Second simulationBased on domain
switiching probability
Throughput plot for AODV and DSDV for second
simulation at node 22
43
Second simulationBased on domain
switiching probability
Throughput plot for AODV and DSDV for second
simulation at node 88
44
Third simulationBased on
modified speed and pause times
Throughput plot for AODV and DSDV for third
simulation at node 22
45
Third simulationBased on
modified speed and pause times
Throughput plot for AODV and DSDV for second
simulation at node 88
46
Comparative Analysis
47
Observations

• Throughput of a specific routing protocol depends
  on the underlying mobility model.
• Mobility model Node density ? Average connected
  paths ? Routing protocol performance
• The number of nodes in the simulation area have
  impact on the measured protocol performance.
• Testing a protocol according to a realistic model
  depicting the movement of nodes is more important
  rather than random movement.
• Increase in the data rate has an effect on the
  routing protocol.
• In terms of throughput AODV performs better in
  all scenarios.

48
Observations

• Throughput is affected by the simulation
  scenario. Different throughputs where obtained
  for the three simulations.
• AODV and DSDV produce different average end to
  end delay, which was calculated as
• sum of delay experienced by each packet\Total
  number of packets
• There is a drop in PDR as traffic is increased.

49
References

• 1 Multihop Ad Hoc Networking The Theory Conti,
  M. Giordano, S. Communications Magazine, IEEE
  .Publication Date April 2007 
• 2. A Survey of Mobility Models for Ad Hoc
  Network Research (2002)Tracy Camp, Jeff Boleng,
  Vanessa Davies Wireless Communications Mobile
  Computing (WCMC) Special issue on Mobile Ad Hoc
  Networking Research, Trends and Applications
• 3 Mobility Metrics to Enable Adaptive Routing
  in MANET Liang Qin Kunz. Wireless and Mobile
  Computing, Networking and Communications,2006.
  (WiMob'2006). IEEE International Conference on
  June 19-21, 2006
• 4 A Study of Recent Research Trends and
  Experimental Guidelines in Mobile
• Ad-hoc Networks. Dow, C.R. Lin, P.J. Chen,
  S.C. Lin, J.H. Hwang, S.F.

50
References

• Advanced Information Networking and Applications,
  2005. AINA 2005. 19th International Conference on
  Volume 1, Issue, 28-30 March 2005 Page(s) 72 -
  77 vol.1
• 5 Analysis of simulation environments for
  mobile ad hoc networks. Technical Report No.
  IDSIA-24-03 December 2003 Dalle Molle Institute
  for Artificial Intelligence Galleria 2, 6928
  Manno, Switzerland
• 6 A brief overview of ad hoc networks
  challenges and directions Ramanathan, R. Redi,
  J. Communications Magazine, IEEE Volume 40, Issue
  5, May 2002 Page(s)20 - 22
• 7 Mobility prediction and routing in ad hoc
  wireless networks. International Journal of
  Network Management archiveVolume 11 , Issue 1
  (JanuaryFebruary 2001) .
• 8 Impact of Mobility on Performance of Routing
  Protocols for Ad Hoc Networks, in AdHoc Networks
  Journal - Elsevier Science, Vol.1, Issue 4, pp.
  383-403, November 2003.
• 9 N. Sadagopan, F. Bai, B. Krishnamachari, and
  A. Helmy, Paths analysis of path duration
  statistics and their impact on reactive manet
  routing protocols,'' in Proceeding of ACM MobiHoc
  2003, Jun. 2003.

51
References

• 10F. Bai, N. Sadagopan, and A. Helmy, Brics A
  building-block approach for analyzing routing
  protocols in ad hoc networks -- a case study of
  reactive routing protocols, in ICC 2004, June
  2004.  
• 11 A. Helmy, Mobility-Assisted Resolution of
  Queries in Large-Scale Mobile Sensor
  Networks(MARQ), in Computer Networks Journal -
  Elsevier Science (Special Issue on Wireless
  Sensor Networks), Vol.43, Issue 4, pp. 437-458,
  Nov. 2003.
• 12 M. Grossglauser, M. Vetterli, Locating Nodes
  with EASE Mobility Diffusion of Last Encounters
  in Ad Hoc Network, in Proceedings of IEEE
  Information Communications Conference (INFOCOM
  2003), Apr.2003.
• 13M. Grossglauser, D. Tse, Mobility Increases
  the capacity of ad hoc wireless network, in
  Proceedings of IEEE Information Communications
  Conference (INFOCOM 2001), Apr. 2001.
• 14H. Dubois-Ferriere, M. Grossglauser, M.
  Vetterli, Age Matters Efficient Route Discovery
  in Mobile Ad Hoc Networks Using Encounter Ages,
  In Proceedings of ACM MobiHoc 2003, June 2003.

52
References

• 15 A Mobility Framework for Ad Hoc Wireless
  Networks. Lecture Notes in Computer Science Vol.
  1987 archive. Proceedings of the Second
  International Conference on Mobile Data
  Management. Year of Publication 2001  
• 16 Study on Environment Mobility Models for
  Mobile Ad Hoc Network Hotspot Mobility Model and
  Route Mobility Model. Gang Lu Belis, D. Manson,
  G. Wireless Networks, Communications and Mobile
  Computing, 2005 International Conference on
  Volume 1, Issue, 13-16 June 2005 Page(s) 808 -
  813 vol.1
• 17 B. Liang, Z. J. Haas, Predictive
  Distance-Based Mobility Management for PCS
  Networks, in Proceedings of IEEE Information
  Communications Conference (INFOCOM 1999), Apr.
  1999.
• 18. X. Hong, M. Gerla, G. Pei, and C.-C.
  Chiang, A group mobility model for ad hoc
  wireless networks, in Proc. ACM Intern. Workshop
  on Modeling, Analysis, and Simulation of Wireless
  and Mobile Systems (MSWiM), August 1999.
• 19 J. Tian, J. Hahner, C. Becker, I. Stepanov
  and K. Rothermel. Graph-based Mobility Model for
  Mobile Ad Hoc Network Simulation,

53
References

• 20 P. Johansson, T. Larsson, N. Hedman, B.
  Mielczarek, and M. Degermark, Scenario-based
  performance analysis of routing protocols for
  mobile ad-hoc networks, in International
  Conference on Mobile Computing and Networking
  (MobiCom'99), 1999. 
• 21. http//cone.informatik.uni-freiburg.de/teach
  ing/labcourse/Adhocnetworks w07/manet/materials/p
  apers/mobility-fanbai-chapter1.pdf
• 22 J. Chen, SH. Chan, J. He, and S. Liew,
  Mixed-Mode WLAN The Integration of Ad Hoc Mode
  with Wireless LAN Infrastructure, IEEE, Global
  Telecommunications Conference (GLOBECOM2003),
  0-7803-7974-8, 2003.
• 22 NS-2 Network simulator. http//www.isi.edu/ns
  nam/ns/
• 23 OPNET Simulator. OPNET Technologies, Inc.,
  Bethesda, MD, USA. http//www.opnet.com/.
• 24 Parallel Simulation Environment for Complex
  Systems (PARSEC). http//pcl.cs.ucla.edu/projects/
  parsec/.
• 25 QualNet Simulator, Version 3.6. Scalable
  Network Technologies, Inc., Culver City, CA, USA,
  2003. http//www.scalable-networks.com.

54
References

• 26 A Critique of Mobility Models for Wireless
  Network Simulation, Pazand, B.   McDonald, C.  
  Univ. of Western Australia, Perth. Computer and
  Information Science, 2007. ICIS 2007. 6th
  IEEE/ACIS International Conference on
• 27 The Random Trip Mobility Model ,"Perfect
  Simulation and Stationary of a Class of Mobility
  Models," J.-Y. Le Boudec and M. Vojnovic, IEEE
  Infocom 2005, Miami, FL, 2005. 
• 28 Analysis of a metropolitan-area wireless
  network, Diane Tang, Mary Baker, Stanford
  University, Stanford, CA Wireless Networks 
• 29 Characterizing Mobility and Network Usage in
  a CorporateWireless Local-Area Network, Magdalena
  Balazinska, Paul Castro.Proceedings of MobiSys
  2003 The First International Conference on
  Mobile Systems, Applications, and Services San
  Francisco, CA, USA 5-8, 2003 
• 30 Characterizing Usage of a Campus-wide
  Wireless Network, David Kotz and Kobby
  Essien,Department of Computer Science, Dartmouth
  College

55
References

• 31Building realistic mobility models from
  coarse-grained traces, Jungkeun Yoon, Brian D
  Noble, Mingyan Liu, Minkyong Kim, International
  Conference On Mobile Systems,  
• 32 Extracting mobility model from real user
  traces,Kim M, Kotz D, Kim S, 25th IEEE
  International Conference on Computer
  Communications. Proceedings. April 2006
• 33 On Modeling User Associations in Wireless
  LAN Traces on University Campuses, W.Hsu A.
  Helmy,
• 34 A mobility model based on WLAN traces and
  its validation, Tuduce C, Thomas G,ETH Zurich,
  Switzerland.
• 35 http//isis.poly.edu/qiming/scengen/index.ht
  ml
• 36 Amit Jardosh,Elizabeth M. Belding-Royer,Kevin
  C. Almeroth,Subhash Suri, Towards realistic
  mobility models for mobile ad hoc networks,
  International Conference on Mobile Computing and
  Networking archive