Mobility Models
Guided By : Prof. Nityananda Sarmah
Presented By : Roop Basant (CSB19054)
Saif Ahmad (CSB19056)
Introduction to mobility Models
● Mobility model describe how the mobile users move over time, taking
into account their position, speed, and direction.
● Vital role in designing of mobile ad hoc networks.
● Mimic the movement pattern of the targeted real-life applications
● These mobility models contribute in designing and testing of Mobile Ad
hoc Network
● Mobility patterns may have a substantial impact on protocol
performance.
● Required in communication of MANETs
Why mobility models required in Ad Hoc Networks
● Helps in performance evaluation of protocols in MANETs
● Decide user movement in relation to a specific area at a microscopic level,
1.Cell change rate,
2.Handover traffic,
3.Blocking likelihood
● Based on peer to peer communication in MANETs, microscopic movement of individuals
required such as
1.Node location,
2.Relative velocity
● Effects of microscopic movement visible when,
1.Links formation,
2.Link severing
Classification of Realistic Mobility Models
● Stochastic - Random Way Point, Random Walk etc.
● Detailed - STRAW(Street Random Waypoint)
● Hybrid - Manhattan , Freeway Mobility Model
● Trace - CRAWDAD(Community Resource For Archiving Wireless Data at
Dartmouth)
Classification of Typical Mobility Model
Random waypoint
model
Random Models
Random Walk model
Gauss-Markov Model
Temporal
Dependency Smooth Random
Mobility Model
Mobility Model
Reference Point Group
Spatial
Dependency
Set of Correlated Models
Pathway Mobility Model
Geographical
Dependency
Obstacle Mobility model
Definition of typical Mobility Model
❖ Random Models
● Based on random movement of nodes
❖ Temporal Dependency
● Affected by movement history of node
❖ Spatial Dependency
● Tends to moves in correlated manner
❖ Geographical Dependency
● Bounded by streets, freeways and obstacles
Random Waypoint Model
● Introduced by Johnson and Maltz
● Keeping track of users location,velocity
and acceleration change over time
● Based on Pause time between change in direction
and/or speed
● Widely used in indoor mobility model
● Independent choice of random destination, speed,
and direction
● Mostly used in assessing of MANETs
Description of Random Way Model
● Each node starts by pausing for a predetermined number of seconds
● Choice of random destination and speed [0, Vmax]
● Take a fixed pause before moving to another location and speed
● Mobility behaviour depends on Vmax and Tpause
➢ Vmax → small, Tpause→ large ⇒ Stable topology
➢ Vmax → large, Tpause→ small ⇒ Dynamic topology
● Choice of optimal speed and metric to quantify that link between nodes are establishes or breaks
➢ Average speed → based on individual speed of two nodes
➢ Relative speed → based on relative speed of node → proposed by Johnson, Larsson, and Hedman → best
metric to decide
Description of Random WayPoint Model
● Relative speed] between node i and j at time t,
● Mobility metric (relative speed averaged over all node pairs and overall duration is)
● where n is the total number of nodes in the simulation field (i.e., ad hoc network),
● |i, j| is the number of different node pairings (i, j), and
● T is the simulation time.
● Average Relative Speed grows linearly and monotonically with the maximum permitted velocity.
Stochastic behaviour of Random WayPoint Model
● Understanding of theoretic characteristics such as probability distribution of transition length and
transition time for each epoch is required
● According to famous computer scientists Bettstetter, Hartenstein and Perez-Costa,
● Random waypoint model is a discrete time stochastic process
● Transition length ⇒ distance that node j move from one way point to another during ith epoch
● Expected value of transition length ⇒
● Time Average ⇒ Average of the transition length of a single Random Waypoint node j over time
⇒ li(j) is the transition length of individual node j
● Ensemble Average ⇒ Average of the transition length in a single epoch i over all the nodes
Stochastic Behaviour in Rectangular Area
● By applying the standard geometrical probability
theory, the probability density functions of
transition length and duration can be obtained as
follows,
● For rectangular simulation field area with length a
and width b, given b ≤ a
● f0(l) is the probability density function,
And l is the length between two random nodes
● The probability density function of transition
length L4,
● expected value of transition length L is,
Stochastic Behaviour in Circular Area
● Probability density function of transition Length L
● Simulation field is a circular area with Radius = a
● Expected value of transition length L
Random Walk
● Proposed to mimic some mobile nodes' movement behaviour because it is thought that
some of them move in an unexpected manner
● Special Case of Random waypoint model with Tpause = 0
● Memoryless Model -
Current Velocity is independent of Previous velocity
● In Fixed time interval-
Speed(0, V] and direction is changed from θ (t) to (0, 2π ]
● Border Effect- Node is bounced back to the simulation field when reaches to boundary
with the angle of θ(t) or π − θ(t)
Limitation of Random models
● Temporal Dependency of Velocity :- The velocity at the current epoch is independent of
the previous epoch
● Spatial Dependency of Velocity:- Movement of the mobile node is independent of
another node
● Geographic Restrictions of Movement:- Movement of a mobile node may be bounded
by obstacles, buildings, streets or freeways.
THANK YOU
