ELEC3500 TELECOMMUNICATIONS NETWORKS Simulation Experiment IV

Description

Experiment: Performance analysis of ALOHA and S-ALOHA medium access control protocols
Required Reading Materials:
1. Text book pages 483-487: topics: Random access, S- ALOHA and ALOHA protocols.
2. Slide set Lecture Slide 20.
3. MUST read the attached note on ALOHA and S-ALOHA on pages 4 & 5 of this document before
proceeding any further.
Objectives:
• Understand basic concepts of the random access techniques used in communication networks.
• Understand principles of ALOHA and S-ALOHA protocols.
• Analyse performance of ALOHA and S-ALOHA protocols for different traffic loads
Procedure:
This laboratory has been developed based on the ALOHA model of the OMNET++ package. Find
the model under the ELEC3500 folder of the OMNET++ directory. The model has been prebuilt so
you don’t need to develop or change any code in the model. Make sure that you use the ALOHA
project in the ELEC3500 folder. You need to modify simulation parameters in the omnetpp.ini file
to obtain necessary simulation results. Use simulation parameters listed in table-1 to obtain necessary
simulation results. Check the .ini file and change parameters appropriately.
Table: 1- Simulation parameters
Simulation Parameter Value(s)
Simulation length (sime-time-limit) 900s
No. of hosts 12
Aloha.host[*].radioDelay 0ms
Packet length 1728 bits
Packet size distribution Constant, exponential (only for second
ALOHA simulation)
Transmission rate 19200 bits/sec
S-ALOHA slot time 100 ms
Total offered load (normalised) For both ALOHA and S-ALOHA:
0.1, 0.25, 0.4, 0.55, 0.7, 0.85, 1, 1.25, 1.5,
1.75, 2
ELEC3500 Telecommunications Networks: Laboratory 4 Handout, A/Prof. J. Khan.
2
Random number seed values Use the default value and a second value only
for the second ALOHA simulation.
You need to run three sets of simulations to obtain necessary results for the simulation experiment.
Simulation 1:
In this experiment you will use the ALOHA simulation model to obtain the S-G graph (normalised
load vs throughput). In the first simulation you will use a fixed packet size. Calculate the packet
arrival rate per hosts using the normalised load values. The normalised load value is obtained by
dividing the total arrival rate by the capacity of the system.
Example: Assume an ALOHA system uses transmission rate of 100 kbits/sec and 2000 bits packet
size. Calculate the arrival rate in packets/sec at a normalised load of 0.2. Also, calculate the
corresponding packet interarrival time of each host, assuming the network has 5 hosts.
System capacity 𝐶𝐶 = 𝑅𝑅
𝐿𝐿 = 100×103
2000 = 50 packets/sec
Total arrival rate at 0.2 normalised load 𝜆𝜆 = 50 × 0.2 = 10 packets/sec
Number of packets/ host 𝜆𝜆𝐻𝐻 = 10
5� = 2 packets/sec
Packet interarrival time 𝑇𝑇𝑖𝑖𝑖𝑖𝑖𝑖 = 1
𝜆𝜆𝐻𝐻
= 0.5 𝑠𝑠𝑠𝑠𝑠𝑠
In this simulation obtain three scalar data values which will be used to generate performance
plots/table.
Scalar values to collect: Channel utilisation, number of received frames and number of collided
frames. Note that channel utilisation statistics of the simulation model represents the normalised
throughput. Normalised load vs the channel utilisation represents the S-G graph of a MAC protocol.
Collect these values for all simulation normalised loads from 0.1 to 2.
Simulation 2:
In this experiment you will use the ALOHA model to obtain same results as in simulation 1. In this
model change the packet length distribution using the exponential distribution, use the same packet
size value as the mean value of the exponential distribution function. You need to obtain two set of
results using two different random number values. Use the default random number seed value as
used in simulation 1 and two different seed values. In this simulation collect the same statistics as
simulation 1.
Simulation 3:
In this experiment you will use the S-ALOHA simulation model. For this simulation, use the same
parameters as simulation 1 (i.e. use the same fixed size packet). Obtain the same simulation results
as simulation 1. Note that you will change the inter-arrival time value under the S-ALOHA code
section in the ini file.
Report Submission Instructions:
Common section:
• Introduction to the experiment, a single paragraph (200 words maximum) explaining the
objectives of the simulation laboratory.
ELEC3500 Telecommunications Networks: Laboratory 4 Handout, A/Prof. J. Khan.
3
• Basic description of the simulation model used in the experiment (250 words maximum, 2
figures).
• Results: Provide following plots using collected simulation results:
o Figure 1: For ALOHA simulation-1: Normalised traffic load vs the channel utilisation
value.
o Figure 2: For ALOHA simulation -2: Normalised traffic load vs the average channel
utilisation value. Average channel utilisation will be generated by taking the average value
of two values obtained by using two seeds.
o Figure 3: For ALOHA simulation -2: Normalised traffic load vs channel utilisation value
for two channel utilisation values (not the average value). You will generate two separate
plots on the same graph each representing channel utilisation values for two different seeds.
o Table 1: Use four columns in the table. First column present the normalised traffic load,
second column present the number of received frames for the ALOHA simulation-1, the
third column present the number of received frames for the ALOHA simulation-2, and the
fourth column present the number of received frames for the S-ALOHA value.
o Figure 4: For S-ALOHA simulation: Normalised traffic load vs channel utilisation value
• Analysis: Compare the S-G/Utilisation graphs of three simulation scenarios. Also, compare
the collision patterns for these three scenarios.
Individual section:
1. Calculate theoretical value of normalised throughput (S) for G (normalised arrival rate) = 0.25,
0.55, 0.85, 1 and 1.25 for ALOHA and S-ALOHA protocols, use the fixed packet size value and
service rate used in the simulation model.
[10]
2. Compare the theoretical and simulation normalised throughput values for the fixed packet size.
Explain why the simulation and theoretical values differ.
[10]
3. Using the simulation results of ALOHA 1 and ALOHA 2, compare the channel utilisation and
number of received packet values. Briefly explain why results from two simulations differ. For
ALOHA 2 simulation use the average value.
[10]
4. Explain why the S-ALOHA protocol offers higher throughput than the ALOHA protocol?
[10]
5. Beside higher throughput what is the other advantage of the S-ALOHA protocol comparing load
vs channel utilisation plots.
[10]
6. Using simulation results, calculate the peak throughput of the S-ALOHA network in bits/sec.
[10]
Report submission date:
For all groups, the report submission date is 23.59pm Friday 6 November 2020 (Week 12)
ELEC3500 Telecommunications Networks: Laboratory 4 Handout, A/Prof. J. Khan.
4
Notes for the ALOHA/S-ALOHA Protocol
Medium access control (MAC) protocols are used in communication networks to allow multiple
network users to share the transmission resources of a network. ALOHA is one of the oldest random
access protocols which was developed by a research group in the University of Hawaii in the 1970s,
allowing data terminals to share the capacity of a communication network. The protocol is still widely
used. Many other modern protocols have been derived using the ALOHA concept. In the ALOHA
network data terminals can transmit their packets as soon as a terminal generates a packet, the
transmission channel is not checked to see whether the channel is currently being used or not. This
simple process may cause packet collisions on the channel when multiple terminals want to transmit
their packets at the same time. However, if a terminal is lucky (i.e. if there are no other terminal or
terminals attempting to transmit at the same time), then the lucky terminal will be able to transmit
their packets immediately. So, the packet operates in a hit and miss scenario: when the network is
lightly loaded then most attempts to transmit will likely be successful. On the other hand, at a higher
network load many of the transmitted packets will collide and be unsuccessful. Collided terminals
will backoff for a certain duration and will attempt to retransmit their packets. After a number of
attempts packets will likely to transmitted, otherwise these packets can be dropped by the respective
terminals. The efficiency of a MAC protocol determines what proportion of the offered load will be
successfully transmitted. Offered load (also referred as input load), arriving at the input of the
network.
Performance of the ALOHA protocol can be measured by using the S-G equation as shown below.
Where G represents the normalised total network input traffic, and S represents the normalised
throughput. Both the S and G values are normalised by the network capacity.
𝑆𝑆 = 𝐺𝐺𝑒𝑒−2𝐺𝐺 (1)
Example: Assume the capacity of a network (C) is 100 packets/sec. The average input arrival rate
(λ) is 44 packets/sec. Calculate the normalised input/offered load and normalised throughput. Also,
calculate the throughput in packets/sec.
𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝐺𝐺 = 𝜆𝜆
𝐶𝐶 = 44 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝/𝑠𝑠𝑠𝑠𝑠𝑠
100 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝/𝑠𝑠𝑠𝑠𝑠𝑠 = 0.44
𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑇𝑇ℎ𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟ℎ𝑝𝑝𝑝𝑝𝑝𝑝 𝑆𝑆 = 𝐺𝐺𝑒𝑒−2𝐺𝐺 = 0.44𝑒𝑒−2×0.44 = 0.44 × 0.414 = 0.182
𝑇𝑇ℎ𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟ℎ𝑝𝑝𝑝𝑝𝑝𝑝 𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝/ sec 𝑇𝑇ℎ = 𝑆𝑆 × 𝐶𝐶 = 0.182 × 100 = 18.2 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝
Above calculation shows the efficiency of the ALOHA protocol determines the throughput of an
ALOHA network. According to the theory the peak normalised throughput of the ALOHA network
is 0.184 at G=0.5. The above network is operating nearly its peak capacity.
To improve the throughput of the ALOHA network, the MAC protocol can be enhanced by changing
it into the S-ALOHA (Slotted-ALOHA) protocol, where some restrictions to the transmission
opportunity of packets are introduced. In the S-ALOHA protocol, terminals are not allowed to
transmit their packets at any time: they are only allowed to transmit at a specific time. The
transmission time is divided in time slots and terminals are only allowed to transmit their packets at
the beginning of a time slot, as shown in the following figure.
ELEC3500 Telecommunications Networks: Laboratory 4 Handout, A/Prof. J. Khan.
5
The above figure shows that terminals are only allowed to start their packet transmission at the
beginning of a time slot which are segmented into discrete time intervals. Hence, the network access
time becomes discrete. Due to the change of the transmission policy, the performance of the SALOHA protocol is represented by the following modified S-G equation:
𝑆𝑆 = 𝐺𝐺𝑒𝑒−𝐺𝐺 (2)
Let’s calculate the S-ALOHA network throughput for same network capacity and the arrival rate as
we have used in the previous example.
Using equation (2) and example parameters, we find the normalised throughput G = 0.2833 and the
throughput in packets/sec is 28.33 packets/sec. There are significant improvements obtained by
changing the access policy!
*********