International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

EFFICIENT ADAPTATION OF FUZZY

CONTROLLER FOR SMOOTH SENDING RATE
TO AVOID CONGESTION IN MULTICAST
NETWORKS
Deepa V B 1 and Ushadevi M B2
1
Department of Computer Engineering, Information Science and Engineering, JawaharaLal Nehru
National College of Engineering.Shimogga, Karnataka, India
2
Department of Telecommunication and Engineering, JawaharaLal Nehru National College of
Engineering.Shimogga, Karnataka, India

ABSTRACT
This paper prefers a fuzzy-logic-based sending rate adaption scheme named FSR(Fuzzy Sending Rate)
intending to improve the evenness of TCPFriendly Multicast Congestion Control (TFMCC). To mitigate
fluctuation of sending rate for TFMCC sender, FSR intends, five actions and link utilization for tuning
sending rate and uses a fuzzy controller to determine which operation should be reaped according to the
feedback information from CLR (current limiting receiver). Asymmetrical membership functions and biased
fuzzy inference rules make FSR as friendly to TCP flows as TFMCC. Simulation results show that FSR has
exceptional smoothness and fine TCP Friendliness.

KEYWORDS
Tfmcc, Rtt, Link Utilization, Sending Rate, Fuzzy Controller

1. INTRODUCTION
Multicast Congestion Control (MCC) is one of the critical methods to weave congestion and
make network perform steadily. It is needed that MCC mechanism must not only assurance the
QoS (Quality of Service) of users, but also ensure that multicast flows could share the resource
adequately with existing flows specifically TCP flows, which is called TCP-Friendliness[1].

MCC functioning can be divided into two division with relating to the manner of sending rate[1]:
single-rate mechanism versus multi-rate mechanism. Although single-rate MCC is poor in
performance and expandability, it is easy to achieve, has fine friendliness and suits to the
conditions that is not so heterogonous. Some newly proposed multi-rate MCC which are also
called hybrid MCC[2-4] make single-rate MCC as building block that each layer applies single-
rate MCC.

One of the utilizations that IP multicast transmits most is multimedia applications (video or voice)
which have smooth transmitting rate. Repeated changes in the transmission rate may disintegrate
the quality of multimedia and import more difficulty and complicatedness to encoder/decoder.
How to sustain smooth transmitting rate is a technology obstacle associated with all MCC
mechanisms. In this paper, we prefer a sending rate adaption scheme based on Fuzzy-logic
intended to smoothen the sending rate of TFMCC 5].The rest of this paper is categorized as
follows. Section 2 summarizes related work. Section 3 interprets the excessive decrease

DOI: 10.5121/ijcsit.2018.10606 53
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, Decem
December 2018

phenomenon that make sending rate of TFMCC fluctuate and presents Fuzzy-logic logic based rate
adaption scheme. Section 4 elaborates the link Utilization of the network to decrease congestion
congestion,
Section 5 introduces Fuzzy Controller for multicast congestion control having smooth sending
Rate. Section 6 gives results and discussion
disc as well as section 7 conclusion.

2. RELATED WORK
There are different mechanisms have been adopted till now to control the congestion in the
network. We know that TFMCC [2] is a steady state equation based multi-cast multi cast technique to
calculatee the throughput of the network. But it has some problems.. First, it is slow in identify
identifying
the Congestion representative and therefore it is slow in reacting to changes in the congestion
condition. Secondly,
condly, the CLR drag down the whole TFMCC session. Therefore, Ther some
modificationsns are made to TFMCC usingusing Additive Increase Multiplicative Decrease (AIMD)[7-
(AIMD)
8]. Once congestion is detected,
ted, it is notified by using Implicit
Implicit Congestion Notific
Notification (ICN)
signaling. After receiving
ceiving this signal the intermediate
int nodes adjust its sending
ing rate so that
congestionn doesn’t occur. Further congestion
conges is implemented using
sing Fuzzy logic controller [9,10
[9,10].

2.1 Proposed Work

We have seen that many algorithms have been proposed to control the congestion in the multicast
network. These algorithms used different protocols to reduce congestion by adjusting the sending
rate of the sender and different mechanisms has been proposed to indicate the congestion
representative. The heterogeneous behavior of the network leads to the more uti utilization of
bandwidth which results in congestion in the network. We propose an algorithm to improve the
utilization by keeping the same sending rate while congestion occurs in the network. And also we
adapt efficient fuzzy controller for sending smooth rate
rate to avoid congestion using parameter RTT
and Link Utilization as input and Sending Rate as output for Fuzzy Controller.

3. ADAPTION SCHEME FOR SENDING RATE USING FUZZY (FSR)

TFMCC has high quality performance including good TCPFriendliness and efficient feedback
suppression mechanisms[3].In that idea TFMCC is extensively accepted and preferred as building
block in some multi-rate
rate MCC[2]. To be TCP-Friendly,
Friendly, each TFMCC receiver estimates its
expected sending rate by a control equation derived fromfrom a model of TCP’s long long-term
throughput[6]:

(1)

where RTT is Round-Trip Trip time, s is packet size and p is packet loss event rate. TFMCC sender
regulates sending rate by the predicted throughput of CLR(current limiting receiver), which is the
receiver who has the minimum expected throughput of the group. Once the expected throughput
of CLR T(k) is lesser thann the current sending rate S(k)) , sender will adopt the new sending rate to
the expected rate. Actually, this operation may decrease the sending rate enormously and
indirectly make throughput change, which is adverse to undergo for multimedia applications. In
the intensification following, p(vv) and RTT(v)) are the loss event rate and RTT correspondingly, at
the sending rate with v ;T( p,, RTT)) is the predictable throughput estimated by (1) with the
parameters p and RTT .We consider two consecutive steady states:states: state1 and state2.Basically, the
network is in state1 and at time t1 the backdrop traffic increases, which means that congestion is
more dangerous. In state1 e1 TFMCC has sending rate of S1 = v1 , loss rate of p1 and RTT of RTT1 ;
correspondingly state2 with S2 = v2 < v , p2 and RTT2 . During each steady state, the three
variables fulfill (1). Let p’ (v1) denote the loss rate estimated
es by the CLR after t1 but before t2 ,
during when the congestion has been more serious but the feedback has not received by the
54
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, Decem
December 2018

sender, so the sending rate is still v1.. We also further assume that the backgroundd traffic doesn’t
change after t1. It is obvious that the higher sending rate is, the larger loss rate and RTT will be if
the background
nd traffic is steady. So we get

P’ > (V1) P(V2)

RTT’(V1) > RTT’(V2) (2)

Att least one of the two inequations above is absolute because of v1 > v2 . Then the calculated
predicted throughput is,

T(p’(v1), RTT’(v1)) < T(p(v2), RTT’(v1))=v2 (3)

Once the feedback with as T(p’(v1), RTT’(v1)) referred to as T’1 for short, reaches the sender, the
sender decreases the sending rate directly to T’1,

S’=T’< v2 (4)

Afterward, the CLR will estimate new los

loss rate and RTT at the sending rate of S’2<
2< v2:

p(S’2)≤p(v2) ≤ p’(v1)
RTT(S’2) ≤ p’(v1)≤ RTT’(v1) (5)

Then the new calculated expected throughput will be higher than S’′. When the new feedback
arrive at the sender, the sender will increase the sending rate additively. Finally with several step
of adjusting, the sending rate will be close to v2 .

Figure 3.1 Excessive decrease phenomena in TFMCC and Multiplicative Decrease action in FSR.

Now, we can analyze that TFMCC will decrease the sending rate excessively and it will take a
long time to converge to the new state as depicted in Figure(3.1).
Fig Thiss conclusion will be proved
by simulations in section 5.

3.1 Rate Adjusting Actions

n order to alleviate the excessive decrease phenomenon in TFMCC, we introduce five rate
adjusting actions into FSR for different congestion degree.

• Additive Increase (AI): The additive increase component should be such that at no instant of
time should the sending rate undergo an increment greater than about 10% the current size. This
serves to distinguish an additive increase form a multiplicative increase. To ensure TCP-
Friendliness, AI action is taken when the expected rate is a little higher than sending rate.

55
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

• Additive Decrease (AD): The additive decrease component should be such that the decrement
in the sending rate should never be more 10%. This serves to differentiate it from a multiplicative
decrease. Actual, if the expected sending rate is a little lower than the current sending rate, which
is meaning congestion is not very serious, there is no need to decrease sending sharply, especially
for multimedia applications. In such case, in order to smoothen sending rate we introduce
“additive decrease (AD)” action that the sending rate will decrease by one packet per RTT.

• Multiplicative Increase (MI): The multiplicative increase component should be large enough
so that the increment in size of the sending rate is larger than 10% always. To fully utilize
resource and improve response speed, we introduce “MI” into FSR: if the expected T’1 is much
larger than the sending rate, the new sending rate will increase to S1 +T’1/2.

• Multiplicative Decrease (MD): The multiplicative decrease component should be so chosen

that the sending rate decrement is never less than 10%. From (3) we can see that the actual rate
the receiver can accept is between the two values: T’1< v2 < S1 .In FSR, the sender decreases the
sending rate to S1 + T’1/2 instead of T’1 directly, which is called “MD” action:

3.2 Link Utilization

We are proposing an algorithm to improve the utilization by keeping the same sending rate while
congestion occurs in the network. For this we have to first calculate the link utilization using old
link utilization method. The proposed algorithm entitles LUMCC is given below[11]:

Algorithm:

Link Utilization Based Multicast Congestion Control (LUMCC)

1) Initialize the total link capacity.
2) Initialize the initial sending rate.
3) Initialize the queue size.
4) Initialize the packet size.
5) Set the session time.
6) Calculate the packet loss ratio on the link.

Pls=Pd/Pd+Ps

Where Pls is the Packet loss observed on the link, Pd is the number of Packets dropped, Ps is
the number of Packets sent on the link.

7) Calculate the link utilization, αij.

Where ∑ bw f* max(Xif)f denotes the value of total traffic demand for all flows fЄF that
are transmitted through link (i, j), Cij is the link capacity.

8) Setting the threshold values:

a) If (99% of link utilization < αij)
Then, Congestion is very high and we adjust

Fnew= αij/β*2

b) If (99% ≤ αij < 90%) Then, Congestion is high and we adjust

Fnew= αij */ 2β /2
If (Fnew > 99%) Then go to step “a”.

c) If (90% ≤ αij ≤ 50%) Then,\

56
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

Fnew= αij * β*2

Congestion is medium and we adjust
If (Fnew > 90%) Then go to step “b”.

d) If (0 < αij < 49%) Then, Congestion is low and we adjust

Fnew= αij * log2α
Else increment Fnew till its value reaches to medium value.
We see the example of propose algorithm given as below:

Example: Suppose the Link Capacity is 100 Mbps, Initial Sending Rate is 80 Mbps, Packet size
is 300, RTT is 150 ms, decreasing factor, β is 0.65 (0 < β < 1), increasing factor α = S/RTT is 2
and we vary the Queue size.

Case 1: Queue size = 50 packets

Link utilization,
α = ∑ bw f* max(Xif)f /Cij fЄF
=80*50/100
=40%

the link utilization is 40% means that congestion is low. Then we use third condition and the
proposed formula is:
Fnew= αij * log2α
=40*log2(S/RTT)=40*log2(2)=40%

Again, the Fnew is 40%, then we go to step “c”.

So, our utilization comes to 65%.

Case 2: Queue size = 80 packets

Link utilization
α = ∑ bw f* max(Xif)f /Cij for all fЄF
=80*80/100=64%

The link utilization is 64% means that congestion is medium. Then we use second condition and
the proposed formula is:
Fnew= αij * β*2
=64*0.65*2
=83%

So, our utilization comes to 83%.

Case 3: Queue size = 120 packets

Link utilization,
α = ∑ bw f* max(Xif)f /Cij for all fЄF
=80*120/100=96%

As utilization is 96% which shows the high congestion in the network is according to set
thresholds. So, we have made the congestion medium. For we use first condition and the proposed
formula is:
Fnew= αij */ 2β /2
96*20.65*2=5%

We conclude that if our link utilization is high then we need more care about the congestion,
otherwise regularly needs to increase the flow speed according to low and medium include with
57
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, Decem
December 2018

medium and high factor of speed respectively. Therefore, utilization of link is very important
phenomenon for control the congestion.

4. FUZZY INFERENCE ENGINE

Fuzzy Inference System is the key unit of a fuzzy logic system having decision making as its
primary work as shown in figure (4.1).
(4.1). It uses the “IF…THEN” rules along with connectors
“OR” or “AND” for drawing essential decision rules[12].
rules

Following are
re some characteristics of FIS

• The output from FIS is always a fuzzy set irrespective of its input which can be fuzzy or
crisp.
• It is necessary to have fuzzy output when it is used as a controller.
• A defuzzification unit would be there with FIS to convert
convert fuzzy variables into crisp
variables.

The following five functional blocks describe the construction of FIS

• Rule Base − It contains fuzzy IF-THEN

IF rules.
• Database − It defines the membership functions of fuzzy sets used in fuzzy rules.
• Decision-making Unit − It performs operation on rules.
• Fuzzification Interface Unit − It converts the crisp quantities into fuzzy quantities.
• Defuzzification Interface Unit − It converts the fuzzy quantities into crisp quantities.
Following is a block diagram of fuzzy interference system.

Figure 4.1 Fuzzy Inference Engine

4.1.1 Types of Controller

● Self-organising controller – a selfish creation noticing nothing outside itself and always
observing itself only and nothing else.
● Adaptive controller – a system that is just a current situation with neither memory nor
recollections about the past and reflections about the future.
● Learning controller – an industrious student constantly developing and expanding his or her
knowledge and experience.

4.2 Mamdani Fuzzy Logic C

Controller

The most commonly used fuzzy inference technique is the so called Mamdani method (Mamdani
& Assilian, 1975) which was proposed[12], by Mamdani and Assilian, as the very first attempt to
control a steam engine and boiler combination by synthesizing a set of linguistic control rules
obtained from experienced human operators. Their work was inspired by an equally influential
58
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

publication by Zadeh (Zadeh, 1973). Interest in fuzzy control has continued ever since, and the
literature on the subject has grown rapidly. A survey of the field with fairly extensive references
may be found in (Lee, 1990) or, more recently, in (Sala et al., 2005).In Mamdani’s model the
fuzzy implication is modeled by Mamdani’s minimum operator, the conjunction operator is min,
the t-norm from compositional rule is min and for the aggregation of the rules the max operator is
used. In order to explain the working with this model of FLC will be considered the example from
(Rakic, 2010)[13] where a simple two-input one-output problem that includes three rules is
examined:
Rule1 : IF x is A3 OR y is B1 THEN z is C1
Rule2 : IF x is A2 AND y is B2 THEN z is C2
Rule3 : IF x is A1 THEN z is C3.

Step 1: Fuzzification

The first step is to take the crisp inputs, x0 and y0, and determine the degree to which these inputs
belong to each of the appropriate fuzzy sets. According to Figure (4.2) one obtains
µA1 (x0) = 0.5, µA2 (x0) = 0.2, µB1 (y0) = 0.1, µB2 (y0) = 0.7

Step 2: Rules evaluation

The fuzzified inputs are applied to the antecedents of the fuzzy rules. If a given fuzzy rule has
multiple antecedents, the fuzzy operator (AND or OR) is used to obtain a single number that
represents the result of the antecedent evaluation. To evaluate the disjunction of the rule
antecedents, one uses the OR fuzzy operation. Typically, the classical fuzzy operation union is
used :
µA∪B(x) = max{µA(x), µB(x)}.

Similarly, in order to evaluate the conjunction of the rule antecedents, the AND fuzzy operation
intersection is applied:
µA∩B(x) = min{µA(x), µB(x)}.

The result is given in the Figure (4.3).

Now the result of the antecedent evaluation can be applied to the membership function of the
consequent. The most common method is to cut the consequent membership function at the level
of the antecedent truth; this method is called clipping. Because top of the membership function is
sliced, the clipped fuzzy set loses some information. However, clipping is preferred because it
involves less complex and generates an aggregated output surface that is easier to defuzzify.
Another method, named scaling, offers a better approach for preserving the original shape of the
fuzzy set: the original membership function of the rule consequent is adjusted by multiplying all
its membership degrees by the truth value of the rule antecedent. Figure (4.4).

Step 3: Aggregation of the rule outputs

The membership functions of all rule consequents previously clipped or scaled are combined into
a single fuzzy set as shown in Figure(4.5).

Step 4: Defuzzification

The most popular defuzzification method is the centroid technique. It finds a point representing
the center of gravity (COG) of the aggregated fuzzy set A, on the interval [a, b]. A reasonable
estimate can be obtained by calculating it over a sample of points. According to Figure(3.6), in
our case results,

59
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

COG =(0 + 10 + 20) × 0.1 + (30 + 40 + 50 + 60) × 0.2 + (70 + 80 + 90 + 100) × 0.5
0.1 + 0.1 + 0.1 + 0.2 + 0.2 + 0.2 + 0.2 + 0.5 + 0.5 + 0.5 + 0.5= 67.4

4.3 Universal approximators

Using the Stone-Weierstrass theorem, Wang in (Wang, 1992) showed that fuzzy logic control
systems of the form ,, Ri : IF x is Ai AND y is Bi THEN z is Ci , i = 1, ..., n With,

Figure(4.2) Fuzzification

Figure (4.3) Rules Evaluation

Figure(3.4) Clipping and Scaling

Figure(4.5) Aggregation of the Rules output

60
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

Figure(4.6) Defuzzification

4.2.1 Mamdani fuzzy logic controller

• Gaussian membership functions

where x0 is the position of the peak relative to the universe and σ is the standard deviation
• Singleton fuzzifier
fuzzifier(x) = x
• Fuzzy product conjunction
µAi (u) AND µBi (v) = µAi (u)µBi (v)
• Larsen (fuzzy product) implication
[µAi (u) AND µBi (v)] → µCi (w) = µAi (u)µBi (v)µCi (w)
• Centroid deffuzification method

where ci is the center of Ci , are universal approximators, i.e. they can approximate any
continuous function on a compact set to an arbitrary accuracy.

4.4 Fuzzy Based Congestion Estimation

Fuzzification: The mapping from a real-valued point to a fuzzy set is known as Fuzzification
which receives other robots information in order to convert it into fuzzy linguistic variable inputs.
The fuzzy logic is chosen based upon the following two reasons: a) In between the normal and
abnormal events, clear boundaries are not present, b) Fuzzy rules should level the normality and
abnormality separation. The fuzzy set can be represented using the mathematical formation
known as membership function.

Rule Definition: Conditional statements are used to implement a membership function which
characterizes a fuzzy set A in x. When the fuzzy statement in an antecedent is true to some degree
of membership, the consequent of the same degree also proves to be true.

Rule structure: If antecedent then consequent. The rule, When both the variables have different
values high and low, then we can get a generous output otherwise a malicious output is detected.
For a fuzzy classification system, the case or an object can be classified by applying the set of
fuzzy rules which depend upon the linguistic values of its attributes. The rule is functioned at the
number given by the antecedent which has a value between 0 and 1. The input can be fuzzified by
evaluating the antecedent and then essential fuzzy operators can be applied. The consequent
obtains this result as the inference.

61
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

We will now describe our methodology for fuzzy logic approach to control congestion in the
network. In controlling congestion, the three most important variables are the RTT, Link
utilization, Sending Rate. With fuzzy logic, we assign grade values to our three variables. Our
fuzzy set therefore consists of three fuzzy variables.

Fuzzy set = {R,L,S } (R –RTT, L-LinkUtilization, S- SendingRate)

Fuzzy logic implements human experiences and preferences via membership functions and fuzzy
rules. In this work, the fuzzy if-then rules consider the parameters: R –RTT, L-Link Utilization,
S-Sending Rate.

The fuzzy logic uses two input variables and one output variable. The two input variables to be
fuzzified are RTT and Link Utilization. The inputs are fuzzified, implicated, aggregated and
defuzzified to get the output as Sending Rate. The linguistic variables associated with the input
variables are Low (L), Medium(M) and high (H). The output variables use three linguistic
variables H, M, and L where H denotes high Sending Rate, M denotes Medium Sending Rate and
L denotes Low Sending Rate. The rules for the FIS are shown below as shown in Table1. They
utilize the AND method which is based on the min function. The FIS rules of the Fuzzy Inference
System are:

Table 1 Fuzzy Rules (LU-Link Utilization, SR-Sending Rate) (L-Low, M-Medium, H-High)Fuzzy Rules

If RTT is Less and Lu is less then SR is High.

If RTT is Less and Lu is Medium then SR is Medium.

If RTT is Less and Lu is High then SR is Low.

If RTT is Medium and Lu is less then SR is Medium.

If RTT is Medium and Lu is Medium then SR is Medium.

If RTT is Medium and Lu is High then SR is low.

If RTT is High and Lu is less then SR is Medium.

If RTT is High and Lu is Medium then SR is Low.

If RTT is High and Lu is High then SR is Very Low.

Control Action to be taken after defuzzification as shown in table 2,and description is given in
section 3.1 Rate Adjusting Action. (MIAD-Multiplicative Increase and Additive Decrease,
AIAD-Additive Increase and Additive Decrease, AIMD-Additive Increase and Multiplicative
Decrease).
Table 2 (Control Action for smooth Sending Rate)

Sending Link Utilization

Rate L M H
RTT L MIAD AIAD AIMD
M AIAD AIMD AIMD
H AIAD MD MD

62
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

5. SIMULATION RESULTS AND DISCUSSION

Simulations were carried out using Network Simulation (ns- 2.35). We patch new agent TFMCC
algorithm in transport layer ns-allinone-2.35. Varying links and packet size and bandwidth
calculated throughput with comparing TFMCC Sending Rate and FSR (Fuzzy Sending Rate).
Topology is as shown in figure(5.1). The network topology for single multicast. is below of 15
receivers: Simulation parameters are as shown in Table 3.

Table 3 Simulation Parameters

Parameters Value
Link Bandwidth 15-100Mbps
Link Delay 25ms
Queue Size 60-99 packets
Sending Rate(initial) 85Mbps
No.of Groups 5
No.0f Receivers 15
RTT 150ms
Packet Size 300
Session Time 500ms
Congestion Status High, Medium, Low

Figure (5.1) Multicast Topology

5.1 Results and Comparisons

5.1.1 Packet Delivery Ratio: Number of packets sent to the recovers is more than the TFMCC
comparing FSR(Fuzzy sending Rate as shown in below. The sending rate of FSR sender which
has less monitor and notch is smoother than that of TFMCC sender. In such dynamic network,
the tight track to expected rate calculated by CLR will make multicast flow unstable. On the
contrary, FSR with fuzzy controller can adjust sending rate adaptively with the knowledge of
feedback information, so FSR has smoother sending rate than TFMCC as shown in Figure (5.2)
and (4.3).

63
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

800
Number of PacketSent
600
P
400 WithoutFSR
a
200
c WithFSR
0
10 20 30 40 50
Links
Figure (5.2) Packet Delivery Ratio

Figure (5.3) Throughtput TFMCC vs TFMCC-FSR

5.1.2 Packet loss Ratio:The variation is due to the link utilization of the network. The new
proposed sending rate (FSR) shows the better result than the existing TFMCC link utilization
strategy because of the less packet loss ratio. This is because of efficient link utilization using
Fuzzy logic Controller as shown in figure (5.4).

PacketDropped

D WithoutFSR
60
r 50 WithFSR
o 40
p 30
20
10
0
Links 20
10 30 40 50

Figure (5.4) Packet loss Ratio

5.1.3 Throuhput Comparing TFMCC vs TFMCC-FSR

Figure(5.5) shows the variation of throughput with the time. It shows the maximum time needed
for the multi-cast source till reaching a steady state throughput. It is clear that proposed approach
Outperforms existing approach.

64
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

TFMCC vs TFMCC-FSR
2.5
T
h
r2
o
1.5
u
g TFMCC
h1
t
p
0.5
u
t
0
10 20 30 40 50
Links

Figure (5.5) Throughput TFMCC vs TFMCC-FSR

5.1.4 Fuzzy Controller operations

a) RTT is low and LinkUtilization is Low then sending rate High

b) RTT is low and LinkUtilization is High then Sending Rate is Medium

65
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

c) RTT is high and LinkUtilization is Low Then Sending rate is Medium

d) RTT is High and LinkUtilization is High Sending Rate is low

Fuuzy Graph

Figure(5.6) Fuzzy Sending Rate

Control surface of the Fuzzy Congestion Controller is shown in Figure (5.6). The control surface
is shaped by the rule base and the linguistic values of the linguistic variables. By observing the
progress of simulation, and modifying the rules and definitions of the linguistic values, FSR can
be tuned to achieve better Link utilization, and smoothened Sending Rate.

6. CONCLUSION
For the requirement of multimedia application based on IP multicast, we have proposed an
improved rate adaption scheme named FSR to smoothen the sending rate of TFMCC sender.
FSR introduces four actions to adjust sending rate and uses a fuzzy controller for making decision
to choose one of the four actions adaptively. In dynamic network environment, fuzzy controller
uses the difference between expected rate and sending rate to reflect the congestion degree, as
well as the difference between two latest consecutive expected rates to predict the trend of
network. Under the fuzzy controller, MIAD,AIAD,AIMD actions eliminate the “sawtooth”
phenomenon in TFMCC, which is crucial for smoothing sending rate. When the available
66
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, December 2018

bandwidth is turning abundant, an algorithm for congestion control which based on utilization of
link and taking decision according to high, medium and low congestion.

REFERENCES
[1] J. Widmer, R. Denda, and M. Mauve, “A Survey on TCP-friendly Congestion Control,”IEEE
，
Network vol. 15, no. 3, 2001, pp.28-37.

[2] Gu-In Kwon and J. W. Byers, “Smooth Multirate Multicast Congestion Control,” Proc.of IEEE
INFOCOM 2003, San Francisco: IEEE Communications Society, 2003,pp.1022-1032.

[3] J. Li, M.Yuksel, X. Fan, and S. Kalyanaraman, “Generalized Multicast Congestion Control”.
Computer Networks. 51(2007), pp.1421-1443.

[4] M. Rodríguez Pérez, S. Herrería Alonso, M. Fernández Veiga, and C.Lόpez García. “An Adaptive
Multirate Congestion Control Protocol for Multicast Communications”.Computer Communications,
29 (2006), pp.2247–2260.

[5] J. Widmer and M. Handley, “Extending Equation-based Congestion Control to Multicast

Applications,” Proc. Of ACM SIGCOMM’01,August 2001, San Diego, California,USA.

[6] J. Padhye, V. Firoiu, D. F. Towsley, and J. F. Kurose. “Modeling TCP Reno Performance: A Simple
Model and Its Empirical Validation,” IEEE/ACM Transactions on Networking, vol. 8,no. 3, April
2000, pp.133–145.

[7] W. Kammoun and H. Youssef, “Improving the Perfor-mance of End-to-End Single Rate Multicast
congestion.

[8] G.-I. Kwon and J. W. Byers, “Leveraging Single Rate Schemes in Multiple Rate Multicast congestion
Control Design,” IEEE Journal on Selected Areas in Communica-tions, Vol. 22, No.10, 2004, pp.
1975-1 1986.

[9] R. Chakravarthi and C. Gomathy, “A Fuzzy Approach To Detect And Control Congestion In Wireless
Sensor Net-works,” Indian Journal of Computer Science and Engi-neering (IJCSE),Vol. 3, No. 3,
2011, pp. 476-483.

[10] J. Bian, Z.-Q. Wang, H.-H. Ke and G.-Z. Zhang, “Con-gestion Control Protocol Based on Delay
Parameters for Layered Multicast Communication,” ICCT International Conference on
communication Technology, 10 April 2007.

[11] Manisha Manjula,Rajesh Mishra,Joysna ,”Link Utilization Based Multicast Congestion

Control”,Published On Communication and Networks,2013,5,649-653

[12] “Fuzzy Controller”,Leonid Reznic Book, ISBN 0 7506 3429 4An imprint of Butterworth-Heinemann
Linacre House, Jordan Hill, Oxford OX2 8DP A division of Reed Educational and Professional
Publishing Ltd

[13] Rakic, A. (2010). Fuzzy Logic. Introduction 3. Fuzzy Inference, ETF Beograd. URL:
http://www.docstoc.com/docs/52570644/Fuzzy-logic-3

67
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 10, No 6, Decem
December 2018

Authors
Mrs Deepa V B working as Assistant Professor in Department of Information Science
and Engineering at JawaharaLal Nehru National College of Engineering Shimoga
Karnataka. Perusing Ph.D in Computer Network in the area of Multicast Congestion
Control.

Dr Ushadevi M B working as Prof and Head of Dept of Telecommunication

Engineering at JawaharLal Nehru National College of Engineering Shimlogga
Karnataka. She completed Ph.D Degree in 2010 at Kuvempu University Shimogga.
Her research area is Switching Networks, interested
interes areas are Wireless Sensor
Networks, Mobile Networks. She published many conference ,journal papers.

68