HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Performance Evaluation of
Mobile Relays in CDMA System
Presented By:
Md. Faisal Murad Hossain
faisal@cc.hut.fi
Communications Laboratory
Helsinki University of Technology
Supervisor:
Prof. Sven Gustav Häggman
Instructor:
Prof. Riku Jäntti
1
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
•
•
•
•
•
•
Outline
Introduction
Wireless Network Overview
(i) Cellular CDMA Network
(ii) Infostation Network
(iii) Multihop Cellular Network (MCN)
Relaying Design Objective in MCN
Mobility Model
Basketball Multihop Scheduling
– Overview
– Pseudocode
– Relay Selection
Performance Analysis
– Effect of User Density
– Effect of Cell Radius
Conclusion
2
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Introduction
• The study of this thesis includes proposing a
Scheduling Algorithm, for Mobile Relays in
Uplink DS-CDMA, which we name as Basketball
Multihop Scheduling Algorithm.
• We also compare its performance with other
existing technologies e.g; pure CDMA and
Infostation Systems.
3
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Wireless Network
Overview(I)
• Cellular CDMA Network
– The area that a base station covers around it, in
which communication from and to the base station is
achievable, is called its cell, thus the term comes
“cellular network”
– It uses the multiplexing scheme CDMA, Code Division
Multiple Access.
– Mobile stations communicate only with Base Stations.
4
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Wireless Network
Overview(II)
• Infostation Network
– Any mobile node communicates with the base
station only when they are close enough
(within the transmit range).
5
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Wireless Network
Overview(III)
• Multihop Cellular Network(MCN)
– In MCN the base station and mobile stations
are not always reciprocally accessible in a
single hop.
– The key advantage is that the mobile stations
can directly communicate with each other
provided that they are mutually reachable.
– MCN can perform multihop routing.
6
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Relaying Design
Objective in MCN
• Relaying Design Objective in MCN
– Communication range extension.
 Transmission through several hops.
– Higher data rates.
 Short hops-high data rate
– Better Quality.
 Cooperative Relaying
 original signal is received by several relays and forwarded to
the destination through different paths so that it does not get
stuck with a bad path and can be switched to a good path.
– Improved Capacity Again.
 simultaneous transmissions by both the BS and the relays.
7
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Mobility Model
• Mobility Model
– We used Directional Random Walk Mobility model.
– In the beginning of the simulation, each mobile node picks a random
destination and traverses to that destination in a straight line at a
uniform speed.
– When the destination is reached, each mobile node chooses a new
destination
– Each mobile node has a higher probability in moving in the same
direction as the previous move. And we assigned different probabilities
for all other directions as shown in the figure,
p0  0.7, p1  0.1, p2  0.05
so that i po  2 * p1  2 *p2  1
– The new location of the mobile node depends
on the previous location/direction and speed
of the mobile node. When a mobile node
reaches the boundary of the cell, it should
flip-over to the reverse direction of the cell.
8
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
Basketball Multihop
Scheduling(I)
Overview
–
–
–
–
–
Basic idea is to deploy Multihop transmission from mobile nodes back
to the base station.
We utilize the similarity between the basketball game and our
Multihop uplink packet scheduling problem. By regarding players, the
basket and the ball as mobiles, the base station and data packet,
respectively, we can mimic passing (Multihopping) patterns of the
basketball players.
A major difference between the two is that in the Multihopping
problem, there are many packets (balls) while in the basket ball
game, there is only one ball to shoot into the basket.
For each CDMA time slot, a number of mobile nodes are selected as
relays to let other nodes nearby to send the packets to them.
A transmitting terminal may select a relaying node that is close to it
and meanwhile the relaying node is located closer to the base station
than that transmitting one.
9
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
Basketball Multihop
Scheduling(II)
Overview (Contd.)


There are two control parameters: Transmission range r, and
relay probability p.
Scheduling rule
–
–
A mobile will act as a relay with probability p.
 If it switch to relay mode, it only receives packet during the
time slot.
A mobile who does not act as relay will try to transmit
 If the base station is in its transmission range r, it will transmit
to it directly.
 If base station is not in its range, the mobile will select the
relay node which is closest to the base station and transmit
the packet to it.
 If there are no relay nodes closer to the base station than the
mobile itself, the mobile will remain idle during the slot.
10
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Basketball Multihop
Scheduling(III)
Node that tries
to transmit
Relay node
Communication
link
Base station
r
11
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
• Transmit mode
No
Is base station
within radio range
Yes
Yes
Is there any relays
within radio range
No
Stay idle
No
Basketball Multihop
Scheduling(IV)
Is any relay closer to
The base station
than the mobile itself
Transmit to the best relay
within radio range
Yes
Transmit directly
to the base station
12
Best relay = relay closest to the base station
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
Basketball Multihop
Scheduling(V)
Pseudocode
At each slot
–
–
–
–
–
We determined packet arrivals for all users
We Checked which mobiles are relays
We went through transmit mode users and determined
active links
We also determined SINR at the receiver of active links.
Then we determined the packet transmissions in active link
 We removed transmitted packets from the transmission
buffer of the source nodes and added them to the end
of the queue at the destination nodes
 If packet were transmitted to the base station, we
recorded the packet delay.
13
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
Basketball Multihop
Scheduling(VI)
Relay Selection
–
–
–
–
Let dij(t) denote the distance between mobile i and mobile j at
time t. Let index 0 denote the base station so that di0(t) is the
distance between mobile i and the base station at time instant
t. Clearly, d00(t)=0 for all t.
Let M={1,2,…,N} denote the set of all mobiles and let
N={0,1,2,…,N} denote the set of nodes including the base
station (i=0).
Let R(t)µN denote the set of relay nodes at time slot, t. We
assumed that a node i becomes a relay at time t with
probability pi(t)¸ 0, p0(t)=1. That is the base station is always
willing to act as a relay for all the nodes. All other nodes are
wireless relays except the base station which relays the packet
to the core network (e.g. internet).
Let A(t)µM denote the set of active nodes. I.e. the nodes that
are not acting as relays
14
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
Basketball Multihop
Scheduling (VII)
Relay Selection (Contd.)
–
A node i2A(t) selects relay ki using the following rule
ki  arg max jF i (t )  dij  1    d j 0  , i A (t )
where 0··1 is a weighting parameter
of
feasible relays from mobile i.
and Fi(t) denotes the set
F i (t )   j | j R (t ), dij  r , d j 0   di 0 
–
The feasible set includes all the nodes that are within the radio
range r from the mobile i and are closer to the base station than
the node i. The parameter 0<·1 denotes a margin. A relay node
is only accepted if its distance to the base station is less than
that for node i.
15
Basketball Multihop
Scheduling (VIII)
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
• Relay Selection (Contd.)
Node transmitting a packet
Feasible relays
Must lie in the
intersection of
the two circles
Feasible (candidate) relay
di0
dj0
di0
j
dij
i
r
Unfeasible relay
F i (t )  { j}
ki  j
16
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
Performance Analysis(I)
Performance Analysis
– Simulation results are analyzed for the
comparison of Basketball Multihop system
with CDMA and Infostation systems with the
help of CDF plots.
– Analysis based on Packet Delays and
Throughputs for different systems.
17
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Performance Analysis(II)
Table 1: Simulation Parameters
Parameter
Value
No. of Nodes
100
Relay Probability
0.2
Simulation Time
120 sec
Radius of Cell
5 Km
Attenuation Factor
4
MS Transmit power
0.1 W
Noise Power
1 pW
Slot Length
10 mS
Mean Inter-Arrival Time
(80e-3)/3 sec
Std. of Inter-Arrival Time
5e-3
Transmission Range
250 m
Min. Mobile Speed
70*1000/3600 m/s
Max. Mobile Speed
100*1000/3600 m/s
Packet Size
10 bytes
Bandwidth
1.25e6 Hz
18
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Performance Analysis(III)
• Effect of User Density
Table 2: Mean and Variance of data (N=100)
Packet Delay
•
•
Normalized
Throughput
Cases
Mean
(sec.)
Variance
(sec.)
Mean
Variance
System
Packet
Delay (sec.)
System
Throughput
(pkts/sec)
Basketball
Multihop
0.6e+3
CI:0.96
7.8e+5
1.1
CI: 0.92
9.0
1.0
111.5
Pure CDMA
1.3e+3
CI:0.96
3.6e+6
6.9
CI: 1.0
119.6
6.1
697.8
One hop
Infostation
2.2e+3
CI: 1.0
8.0e+6
1.0
CI: 1.0
18.8
0.9
103.6
From Table 2, it can be said that with 100 users in the cell, the Basketball
Multihop system incurs less mean packet delay than CDMA and Infostation
systems. System Packet Delay for Basketball Multihop system is almost
similar as Infostation system but significantly less than CDMA system.
Again from Table 2, Throughput for Basketball Multihop is almost similar as
Infostation system but worse than CDMA system.
19
Performance Analysis(IV)
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
It can be seen from Figure 1 that
for Basketball Multihop case,
80% of the time the packets will
be delivered within 2000 timeslots whereas for CDMA case,
80% of the time the packets will
be delivered within around 3000
time-slots and for Infostation
case, 80% of the time the
packets will be delivered within
4400 time-slots.
Hence Basketball Multihop
system is clearly better than the
other two systems as the Service
probability is better in this system
with less delay maintaining good
Quality of Service (QoS).
Comparison of Packet Delay (CDF)
1
Packet Delay (Probability)
•
0.8
0.6
0.4
Basketball
CDMA
Infostation
0.2
0
0
2000
4000
6000
Time Slots
8000
10000
12000
Figure 1: Comparison of Packet Delays for different
systems, N=100
20
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Performance Analysis(V)
Comaprison of Throughput (CDF)
Throughput.
95
90
No. of Users
• Our results indicate that for
Basketball Multihop case 100
users can be supported with
90% probability whereas for
CDMA case 92 users can be
supported with 90% probability
and for Infostation case 98
users can be supported with
90% probability.
• Hence Basketball Multihop
system is better than the other
two systems with respect to
100
85
80
75
Basketball
CDMA
Infostation
70
65
0
0.2
0.4
0.6
0.8
User Throughput (Probability)
1
Figure 2: Comparison of Throughputs for different
systems, N=100
21
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Performance Analysis(VI)
• Effect of User Density (Contd.)
Table 3: Mean and Variance of data (N=750)
Packet Delay
•
•
Normalized Throughput
Cases
Mean
(sec.)
Variance
(sec.)
Mean
Variance
System Packet
Delay (sec.)
System
Throughput
(pkts/sec)
Basketball
Multihop
2.5e+3
CI: 1.0
6.8e+6
11.1
CI: 1.0
81.3
29.0
8.3e+3
Pure CDMA
2.1e+3
CI: 1.0
5.3e+6
2.0
CI: 1.0
39.8
2.8
1.5e+3
One hop
Infostation
1.8e+3
CI: 1.0
6.3e+6
0.8
CI: 1.0
12.9
0.7
654.2
From Table 3, it can be seen that with 750 users in the cell, the Basketball
Multihop system incurs more packet delay than CDMA and Infostation
systems.
Again from Table 3, Throughput for Basketball Multihop is greatly
increased than the other two systems.
22
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
The results indicate that for
Basketball Multihop case, 80%
of the time the packets will be
delivered within 6400 time-slots
with 0.8 service probability and
0.2 outage probability, whereas
for CDMA case, 80% of the time
the packets will be delivered
within around 4400 time-slots
and for Infostation case, 80% of
the time the packets will be
delivered within 3800 time-slots.
Hence Basketball Multihop
system is worse than other two
systems with respect to packet
delays.
Comparison of Packet Delay
1
Packet Delay (Probability)
•
Performance Analysis(VII)
0.8
0.6
0.4
Basketball
Infostation
CDMA
0.2
0
0
2000
4000
6000
Time Slots
8000
10000
12000
Figure 3: Comparison of Packet Delays for different
systems, N=750
23
• The results suggest that for
basketball Multihop case 750
users can be supported with
95% probability whereas for
CDMA case 750 users can be
supported with 95% probability
and for Infostation case 750
users can be supported with
85% probability.
• Thus, it can be said that if the
number of user is increased,
the Throughput for Basketball
Multihop system is increased
significantly, becomes stable
with good service probability
but incurs much packet delay.
Performance Analysis(VIII)
Comparison of Throughput (CDF)
750
700
650
600
550
No. of Users
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
500
450
400
350
Basketball
Infostation
CDMA
300
250
200
0
0.1
0.2
0.3
0.4
0.5
0.6
User Throughput
0.7
0.8
0.9
Figure 4: Comparison of Throughputs for different
systems, N=750
24
1
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Performance Analysis(IX)
• Effect of Cell Radius
Table 4: Mean and Variance of data (R=1 Km)
Packet Delay
Normalized Throughput
Cases
Mean (sec.)
Variance
(sec.)
Mean
Variance
System
Packet Delay
(sec.)
System
throughput
(pkts/sec)
Basketball
Multihop
211.5
CI: 0.96
1.0e+5
31.5
CI: 0.96
0.07
2.1
3.1e+3
Pure CDMA
2.4 e+3
CI: 0.96
7.3e+6
23.0
CI: 1.0
119.2
33.9
2.3e+3
One hop
Infostation
2.6e+3
CI: 0.96
6.3e+6
12.5
CI: 0.96
167.8
14.9
1.2e+3
• It can be seen from Table 4, that Basketball Multihop system can
provide less packet delay and more throughput than CDMA and
Infostation systems in a small cell with 1 km radius.
25
Performance Analysis(X)
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
From the Figure 5, it can be
seen that for Basketball
Multihop case, 80% of the time
the packets will be delivered
within 4000 time-slots with 0.8
service probability and 0.2
outage probability, whereas for
CDMA case, 80% of the time
the packets will be delivered
within around 4200 time-slots
and for Infostation case, 80% of
the time the packets will be
delivered within 5100 time-slots.
Basketball Multihop system is
better than the other two
systems with respect to packet
delays.
Comparison of Packet Delay (CDF)
1
Packet Delay (Probability)
•
0.8
0.6
0.4
Basketball
CDMA
Infostation
0.2
0
0
2000
4000
6000
8000
10000
12000
Time Slots
Figure 5: Comparison of Packet Delays for different
systems, R= 1Km
26
• The results indicate that for
basketball Multihop case 100
users can be supported with
100% probability whereas for
CDMA case 100 users can be
supported with 96% probability
and for Infostation case 100
users can be supported with
96% probability.
• Basketball Multihop system is
again better than other two
systems with respect to
Throughput.
Performance Analysis(XI)
Comparison of Throughput (CDF)
100
Bassketball
CDMA
Infostation
90
80
70
No. of User
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
60
50
40
30
20
10
0
0
0.2
0.4
0.6
0.8
User Throughput (Probability)
1
Figure 6: Comparison of Throughputs for different
systems, R= 1Km
27
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Performance Analysis(XII)
• Effect of Cell Radius (Contd.)
Table 5: Mean and Variance of data (R=5 Km)
Packet Delay
•
•
Normalized Throughput
Cases
Mean (sec.)
Variance
(sec.)
Mean
Variance
System
Packet Delay
(sec.)
System
throughput
(pkts/sec)
Basketball
Multihop
0.7e+3
CI: 0.96
6.9e+5
7.1
CI: 0.92
11.0
0.8
711.5
Pure CDMA
1.1e+3
CI: 0.96
3.3e+6
4.9
CI: 1.0
117.6
4.1
497.8
One hop Infostation
2.2e+3
CI: 1.0
8.0e+6
1.0
CI: 1.0
18.8
0.9
103.6
From Table 5, it can be said that with large cell radius (5 Km) Basketball
Multihop system provides less delay than the other two systems CDMA and
Infostation systems.
Again from Table 5, Throughput for Basketball Multihop is better than Infostation
and CDMA systems when the cell radius is increased to 5Km.
28
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Comparison of cdf-packetdelay
1
0.9
0.8
Packet Delay (probability)
• The results suggest that for
Basketball Multihop case,
80% of the time the packets
will be delivered within 800
time-slots, whereas for
CDMA case, 80% of the time
the packets will be delivered
within around 3000 timeslots and for Infostation case,
80% of the time the packets
will be delivered within 3500
time-slots.
• Hence Basketball Multihop
system is clearly better than
the other two systems with
respect to packet delays.
Performance Analysis(XIII)
0.7
0.6
0.5
Basketball
CDMA
Infostation
0.4
0.3
0.2
0.1
0
0
2000
4000
6000
8000
10000
12000
Time-slots
Figure 7: Comparison of Packet Delays for different
systems, R= 5Km
29
•
•
Performance Analysis(XIV)
From the Figure 8, it can be seen
that for Basketball Multihop case,
from 28 to 100 users can be
supported with 100% probability
whereas for CDMA case from 32
to 100 users can be supported
with 100% probability and for
Infostation case from 28 to 100
users can be supported with 100%
probability. Yet Basketball
Multihop system is more stable
than CDMA because it can
provide any user with more than
80% probability which CDMA
cannot provide.
Thus we can see that the
Basketball Multihop system works
better in a larger cell with respect
to both Packet Delay and
Throughput.
Comparison of cdf-throughput
1.1
1
User throughput (probability)
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
0.9
0.8
0.7
Basketball
Infostation
CDMA
0.6
0.5
0.4
0
10
20
30
40
50
60
No. of Users
70
80
90
100
Figure 8: Comparison of Throughputs for different
systems, R= 5Km
30
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
•
Conclusion
Conclusion


The simulations have shown that there certainly is a need for relaying
in cellular networks due to improved capacity and shorter delays.
However the Basketball Multihop system works better for the
following conditions:
– More users for increased throughput and stability
– Larger cell for both shorter delay and improved throughput
– Highway-vehicle users for both increased throughput and
improved delay
– High traffic volume for both shorter delays and higher throughput
– Short Transmission range for both improved delays and increased
throughput.
31
HELSINKI UNIVERSITY OF TECHNOLOGY
Performance Evaluation of Mobile Relays in CDMA System
Md. Faisal Murad Hossain
Thank You
Questions?
32