On Minimizing TCP Traffic Congestion in Vehicular Internet of Things (VIoT)

Abstract

The performance of end-to-end wireless link congestion control algorithm in the vehicular internet of things network is plagued by the inherent limitations of spurious rate control initiation, slow convergence time, and fairness disparity. In this article, the delay assisted rate tuning (DART) approach is proposed for the vehicular network that implements two algorithms, utilization assisted reduction (UAR) and super linear convergence (SLC), to overcome the transmission control protocol (TCP) limitations. The UAR algorithm is responsible for initiating the proportionate rate control process based on the bottleneck prediction parameter, thereby regulating the needless rate control during non-congested losses. In the congestion recovery mode, the SLC algorithm executes a dynamic rate update mechanism that enhances the flow rate and minimizes bandwidth sharing disparity among TCP flows. An analytical model was developed to study the DART convergence rate and fairness performance against the existing algorithm. The vehicular simulation outcome also confirms significant enhancement in average transmission rate, average message latency, and average bandwidth sharing performances of the DART algorithms against the RFC 6582, TCP-LoRaD, and CERL + congestion avoidance algorithms under varying traffic flows and node movement scenarios.

Full text

Vol.:(0123456789)
Wireless Personal Communications
https://doi.org/10.1007/s11277-022-10024-5
1 3
On Minimizing TCP Traffic Congestion inVehicular Internet
ofThings (VIoT)
M.JosephAuxiliusJude1 · V.C.Diniesh1 · M.Shivaranjani2 ·
SureshMuthusamy3 · HiteshPanchal4 · SumaChristalMarySundararajan5 ·
KishorKumarSadasivuni6
Accepted: 29 August 2022
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022
Abstract
The performance of end-to-end wireless link congestion control algorithm in the vehicular
internet of things network is plagued by the inherent limitations of spurious rate control ini-
tiation, slow convergence time, and fairness disparity. In this article, the delay assisted rate
tuning (DART) approach is proposed for the vehicular network that implements two algo-
rithms, utilization assisted reduction (UAR) and super linear convergence (SLC), to over-
come the transmission control protocol (TCP) limitations. The UAR algorithm is responsi-
ble for initiating the proportionate rate control process based on the bottleneck prediction
parameter, thereby regulating the needless rate control during non-congested losses. In the
congestion recovery mode, the SLC algorithm executes a dynamic rate update mechanism
that enhances the flow rate and minimizes bandwidth sharing disparity among TCP flows.
An analytical model was developed to study the DART convergence rate and fairness per-
formance against the existing algorithm. The vehicular simulation outcome also confirms
significant enhancement in average transmission rate, average message latency, and aver-
age bandwidth sharing performances of the DART algorithms against the RFC 6582, TCP-
LoRaD, and CERL + congestion avoidance algorithms under varying traffic flows and node
movement scenarios.
Keywords Vehicular internet of things (VIoT)· Transmission control protocol (TCP)·
Vehicular communication network· Intelligent transportation system (ITS)· Wireless
access in vehicular environments (WAVE)
1 Introduction
An emerging connected vehicle paradigm, the VIoT [1–5], interconnects vehicles with
other vehicles, humans, and machines for optimal sharing of traffic and non-traffic-related
information for safe and comfortable commuting. VIoT was formed as an inevitable divi-
sion of the intelligent transportation system (ITS) [6–8] to improve data sharing and better
* M. Joseph Auxilius Jude
jude2193@gmail.com
Extended author information available on the last page of the article

M.J.A.Jude et al.
1 3
resource management for semi or fully autonomous vehicles. VIoT shares the vehicle infor-
mation from the intelligent embedded sensors to the cloud or a diverse group of users
through a fixed hotspot or roadside unit (RSU). The vehicle-to-everything (V2X) paradigm
[9–11] of VIoT solely relies on dedicated short-range communications (DSRC) [12–16] for
device-to-device (D2D) information exchange.
The VIoT operates in a 5.9GHz radiofrequency with 10MHz seven non-overlapping
channels operating in a 75MHz radio spectrum. VIoT also caters to support non-vehi-
cle safety services, such as file sharing, mailing service, toll payment, e-shopping, e-pay-
ment, and acts as a mobile hotspot for last-mile wireless internet connectivity. The internet
remains the backbone of the VIoT ecosystem; the boundless global web traffic and data
center traffic depends on TCP [17–19] to deliver data packets between end systems. Fig-
ure1 displays the vehicular transmission model in different modes.
TCP incorporates flow control [20], congestion control [21], and error control mecha-
nisms, ensuring the successful in-sequence transmission of each information byte between
sender-receiver processes. TCP’s bottleneck control mechanism comprises a slow start
(SS), congestion avoidance (CA), and CR mode. In the SS mode, the source device fixes
the SS threshold (SSThresh) and doubles the source window (Wnd_) rate for each successful
acknowledgment. On exceeding the SSThresh, the source device initiates the CA phase and
gradually increments the transmission rate by one data packet for each RTT. TCP starts the
transmission rate control process when the source device infers packet loss by a timeout or
three duplicate acknowledgments (3DUPACK). The source device trims down the Wnd_
rate of TCP connections during the congestion control process by a commensurable rate
reduction factor of 0.5. In the CR phase, the Wnd_ increments one packet per RTT until the
equilibrium point. Figure2 displays the Wnd_ growth or transmission pattern of additive
increase and multiplicative decrease (AIMD) traffic connection for each iteration or RTT.
The majority of web servers around the globe widely adopt RFC 6582 [22] congestion con-
trol approach due to its lesser implementation complexity and faster mechanism to detect incipi-
ent network bottleneck conditions. In addition, the proper functioning of the congestion control
algorithm in each phase results in optimal throughput performance. However, the existing RFC
6582 implementation remains inappropriate for the vehicular environment due to the following
deficiencies.
Spurious rate control initiation:The RFC 6582 algorithm invokes rate control based on the
timeout condition. This assumption results in triggering a spurious rate decrement process for
Fig. 1 VIoT communication networks

On Minimizing TCP Traffic Congestion inVehicular Internet…
1 3
non-congested RTT jitter or packet losses that arise due to frequent route breakage and radio
channel losses. The spurious rate control phenomenon severely destabilizes the throughput sta-
bility of TCP traffic flow in the wireless links.
Slow convergence rate: TCP implementation in the CR phase is plagued by the slow
Wnd_ update function that leads to additional RTTs to finish the flows, results in increased
communication latency between end systems.
Fairness disparity among TCP flows: From Fig. 1, it is observed that the existing Wnd_
growth in the CR phase maintains a fairness disparity among the concurrent TCP flows, i.e.,
flows with higher and lower transmission rates maintain a massive variation in sharing the net-
work bandwidth.
The DART implementation proposed in this article modifies the rate adjustment model and
Wnd_ growth function to overcome the limitations of TCP connections under the VIoT environ-
ment. The later part of the article is arranged into five sections. Section2 briefs contemporary
literatures on independent congestion control algorithms developed for wireless scenarios. Sec-
tion3 narrates the operation of the UAR + SLC algorithms in vehicular networks, and Sect.4
explicates the fairness and convergence model of the DART approach based on the optimization
theory. Section5 briefly outlines the vehicular simulation outcomes of the proposed and existing
algorithms under diverse scenarios. Finally, Sect.6 presents the proposed UAR + SLC imple-
mentation conclusion and discusses the future direction of the work.
2 Background andRelated Works
Over three decades, TCP’s congestion avoidance algorithm has undergone several modifica-
tions [23] to sustain a stable data rate among multiple traffic flows under both wired and wireless
conditions. This literature briefly reviews the recent developments of independent congestion
avoidance approaches developed to improve TCP’s throughput stability in lossy wireless links.
CERL + [24] method derives two parameters (queuing delay and RTT delay) for predicting the
bottleneck wireless link. The sender initiates rate adjustment when the derived delay parameter
is higher than the threshold level. Rather than of reacting to packet drop, CERL + relies on delay
Fig. 2 Transmission pattern of RFC6582 traffic connections

M.J.A.Jude et al.
1 3
parameter estimation for Wnd_ deflation. However, the Wnd_ growth function in the CR phase is
similar to the RFC 6582 approach, resulting in a slow convergence rate.
TCP Wave [25] approach implements a unique burst-based transmission of packets
instead of traditional Wnd_ based packet transmission. The WAVE algorithm tracks the
wireless network dynamics based on ∆RTT computation, derived from average RTT and
minimum RTT (RTT
min) values. The source node initiates a reduction in burst size when
the derived∆RTT is higher than the tolerated bottleneck level β.
FIT [26] algorithm is yet another derivative of RFC6582, where a newer Wnd_ incre-
ment mechanism is introduced in the CR phase based on the traffic flows. Furthermore,
FIT implements a new rate adjustment mechanism instead of the traditional rate halving
method. However, the FIT algorithm reacts to non-congested losses of lossy radio links
that cripple TCP’s throughput capability. TCP-LoRaD [27] predicts the bottleneck wireless
link based on the queuing delay parameter. However, the Wnd_ growth pattern of TCP-
LoRaD in the CR phase resembles the RFC6582 model resulting in a slower convergence
and bandwidth sharing inequality among multiple traffic flows.
ACC [28] is yet another method to improve AIMD traffic flow under lossy radio con-
ditions. The ACC approach solely relies on packet latency to initiate the transmission
rate adjustment process. First, the sender derives the queuing delay of a wireless link by
computing the RTT values of the current and old packet. The source device starts the rate
adjustment process where the threshold value lesser than the obtained RTT (ζ × qmax). In
the CR phase, the ACC algorithm implements a linear Wnd_ increase similar to RFC 6582,
resulting in a slow data throughput to attain the steady state or maximal point.
Vegas [29, 30] based end-to-end congestion avoidance implementation estimates RTT delay
parameter to invoke the rate adjustment process. However, the Vegas approach fails to update the
latest RTT
min value for wireless topology change. The discrepancy in computing RTT
min leads to
spurious initiation of a congestion control process that cripples TCP’s throughput performance.
Pegas [31], modification of Vegas algorithm implements a particle swarm approach in fixing the
new RTT
min value for dynamically changing topology. Pegas captures dropped packet count, RTT
min, and current transmission rate as the input parameter for setting the dynamic RTT
min value for
each packet interval. However, PSO implementation at the sender side requires more computa-
tion parameters for obtaining a new RTT
min value for each packet interval. Similarly, G-vegas
[32] employs grey prediction theory, and D-TCP [33] uses the cuckoo search optimization tech-
nique to estimate the dynamic RTTmin value based on the wireless dynamics.
However, the Vegas-based approaches perform worse under diverse TCP traffic condi-
tions due to RTT fluctuations [26] in wireless conditions. Table1 summarizes the behavior
of recent AIMD congestion avoidance algorithms under lossy wireless networks. However,
the AIMD approaches Wnd_ growth pattern severely compromise the steady-state conver-
gence of the TCP flows. The additive Wnd_ growth in the CR phase cripples the through-
put rate and requires additional RTTs to complete the TCP flow.
Table 1 Comparison of AIMD congestion avoidance approaches
TCP Wnd_ pattern Rate reduction factor Congestion prediction parameter
CERL + Additive increase Wnd_ *0.5 0.5*[(RTT-RTT
min)*BW]max
TCP-LoRaD Additive increase Wnd_ *0.8 (RTT
max-RTT
min)
WAVE Based on receiver wnd_ ≥ 3 segments ∆RTT > β
ACC Additive increase Wnd_ *0.5 RTT > ζ × qmax
FIT Additive increase Wnd_–(2/(3N + 1))* Wnd_ Packet loss event

On Minimizing TCP Traffic Congestion inVehicular Internet…
1 3
3 The Proposed DART Approach
The DART implements UAR and SLC algorithms which require modification in TCP’s con-
gestion control and recovery mechanisms. The UAR algorithm is a three-step process respon-
sible for the rate tuning at the sender side. Step 1 implements a bottleneck prediction (T_bottle)
parameter to determine the incipient link congestion instead of a packet loss effect. The T_bottle
parameter is computed based on the average RTT (RTT
avg) and RTT
R_min values. As mentioned
in [34], the RTT
avg values are computed from the recent five samples. The RTT
R_min is the
minimum RTT value among the recent five samples that minimize the impact of RTT fluctua-
tion in T_bottle prediction. When the estimated T_bottle rate is superior (0.7) than the threshold
(γ) rate, the sender begins the rate adjustment process. The γ value is fixed as 0.7 based on the
repeated trials using wireless internet RTT measurements. Lesser γ value results in a spurious
timeout due to frequent RTT fluctuations. A higher γ value allows the sender to respond slowly
to bottleneck conditions. The T_bottle parameter is derived as
Step 2 implements the utilization prediction (Wutility) parameter to initiate a proportion-
ate reduction mechanism during the rate adjustment process. The sender identifies low and
high rate TCP flows based on the δ threshold value (δ = 0.5), estimated based on the current
Wnd_ level (Wi) and maximum receiver capacity Wmax or initial SSThresh. The source node
implements two rate decrement factors based on the utilization levels. The Wutility param-
eter and rate decrement factor is derived as
Rate reduction procedure for high bandwidth utilized TCP flow is computed as
Rate reduction procedure for low bandwidth utilized TCP flow is computed as
Step 3, the UAR algorithm implements a modified fast retransmit mechanism during
the packet loss state, resulting in retransmission of missing data packets within the current
RTT without invoking the rate reduction process.
The SLC algorithm is responsible for the Wnd_ increment mechanism in the CR phase
with a faster recovery rate and minimum fairness disparity among traffic flows. The SLC
algorithm introduces a new dynamic Wnd_ update pattern that allows the traffic flow to
enter the steady-state transmission point at a faster rate with the least fairness disparity.
The steady-state point is an equilibrium state where the source device transmission rate
matches the link capacity. The Wnd_ increase pattern of the CR phase is given as
(1)
T
_bottle =
RTT
avg
−RTT
min
RTT
avg
(2)
W
utility =
W
i
Wmax
(3)
W
utility
<𝛿;low rate flow
Wutility
≥𝛿;
high rate flow
(4)
W
i
(k+1)
=
W
i∗
.0.5 ; W
utility ≥
𝛿
(5)
W
i(k+1)=Wi∗
{(
1−
W
i
W
max )}
;Wutility
<𝛿

M.J.A.Jude et al.
1 3
The DART approach finite state machine (FSM) model is represented in Fig.3. The
FSM model specifies interactions among different transition states and their impact on the
transmission rate.
3.1 Fairness andConvergence Efficiency Analytical Validation
The proposed DART approach’s bandwidth fairness and transmission efficiency perfor-
mance are validated using an analytical model and compared against CERL + , TCP-
LoRaD, and RFC 6582 approaches. The fairness and convergence efficiency analytical
equations are derived based on Jain’s model [35]. In wireless conditions, the fairness
performance of TCP flow is influenced by Wnd_ growth pattern and channel condi-
tions [36, 37]. The DART implementation practices a packet latency/T_bottleparameter
to initiate theWnd_ decrement process. Let γ(t) be the T_bottle threshold value; rs rep-
resents the source node dispatching rate, rm represents the highest attainable dispatch-
ing rate of the source node,WLc represents thewireless link capacity and RTarget is the
steady-state point. The source device dispatching rate matches the wireless link capacity
(6)
W
i(k+1)=Wi(k)+
(
4∗Wmax
)
W
i(k)
Fig. 3 FSM model of DART approach

On Minimizing TCP Traffic Congestion inVehicular Internet…
1 3
(RTarget = WLc) in the steady-state point. The RTarget value is taken as the maximum dis-
patching rate (rm) attained by the sender and expressed as the function of time
The source node increment rs for non-congested state γ(t) = 0 (γ < 0.7) and lessen rs
for an increase in network load γ(t) = 1 (γ ≥ 0.7).
Figure4 illustrates the communication model of the proposed DART approach.
The sender updates its dispatching rate rs(t + 1) based on γ(t) and follows an addi-
tive Wnd_ increment pattern for γ(t) = 0 and multiplicative Wnd_ decrement pattern for
γ(t) = 1. The Wnd_ increment and decrement patterns are expressed as
The ri (t) represents the current dispatching rate, rm (t) is the maximum dispatching
rate, WI and WD denotes additive Wnd_ increment pattern and multiplicative decrement
pattern. For the DART approach,WIandWDare derived from the Eq.(4), (5), and (6) as
On substituting (10) and (11) in (9), the equation can be modified as
(7)
r
s(t)=
{
ry1(t),ry2(t),ry3(t), ....rym(t)
}
(8)
𝛾(t)=0, non congested state
1, congested state
(9)
r
s
(t+1)W
I
+r
i
(t),𝛾(t)=0
W
D
∗r
i
(t),𝛾(t)=
1
(10)
W
I=
4
∗
r
m
r
i
(11)
W
D1-
r
m
ri
;ri(t)≤
𝛿
=1
2
;ri(t)>𝛿
(12)
r
s(t+1)=
4∗r
m
r
i
+ri(t),𝛾(t)=0
=1−
rm
ri
∗ri(t);ri(t)𝛿𝛾(t)=
1
=1
2
∗ri(t);ri(t)>𝛿𝛾(t)=1
Fig. 4 DART approach communication model

M.J.A.Jude et al.
1 3
The DART approach flow efficiency equation is developed based on the TCP fairness
and convergence vector diagram displayed in Fig.5. Let rs1 (x-axis) and rs2 (y-axis) are
the two source devices sharing the bottleneck wireless link. The midpoint on the graph
represents the efficiency of the two flows. The region below the efficiency midpoint is
the underutilization, and above is the congested state. The flows (rs1 and rs2) attain its
transmission efficiency when rs1 = rs2 = RTarget = WLc. The convergence rate is an essen-
tial component in attaining the efficiency of the TCP flow. The convergence rate denotes
the pace at which the source device transmission rate attains the equilibrium state. The
dispatching rates of the TCP flows get lessened when rs (t) reach the equilibrium condi-
tion and again attain the steady-state point in repeated cycles.
The condition for the convergence efficiency of TCP flow is derived as
The independent congestion avoidance algorithm’s efficiency is crippled severely by
negative feedback phenomena due to spurious rate decrement process for packet loss
conditions. However, the DART approach invokes rate adjustments solely based on the
T_bottle parameter, which considerably minimizes negative feedback phenomena and
improves the TCP flow efficiency in wireless conditions. The convergence efficiency of
the DART approach is derived by substituting Eq.(12) in (13)
(13)
𝛾
(t)= 0⇒
∑
rs(t+1)>rm(t
)
𝛾
(
t
)= 1⇒
∑rs
(
t
+1)<
rm
(
t)
(14)
𝛾
(t)=0
(
4∗r
m
r
i
+ri(t)
)
>rm(t
)
(15)
𝛾
(t)= 1⇒
(
1-
r
m
ri
)
∗ri(t)<rm(t);ri(t)𝛿
=1⇒
(
1
2∗ri(t)
)
<rm(t);ri(t)
>𝛿
Fig. 5 Efficiency and fairness
vector representation

On Minimizing TCP Traffic Congestion inVehicular Internet…
1 3
Equation15 denotes the utilization level based on the proportionate reduction for γ(t) = 1
condition. Equation14 represents the Wnd_ increment pattern for γ(t) = 0, which results in
a quicker convergence pace towards the midpoint efficiency line. Similarly, the efficiency
equation for RFC 6582 based TCP flows (CERL + and TCP-LoRaD) are derived as
Figure6 displays the convergence efficiency analysis of RFC 6582 and DART in the CR
phase. The convergence efficiency Eqs.(14, 15, 16 and 17) are validated by substituting the val-
ues rm = 36 and the initial ri value = 18. The Wnd_ growth oscillates between the minimum and
steady-state point in repeated cycles. The graph shows that DART accomplishes a faster conver-
gence pace than RFC 6582, i.e., the DART algorithm reaches efficiency midpoint within fewer
RTTs.
Fairness study measures the bandwidth distribution among two separate TCP traffic
flows (rs1 and rs2) sharing the same bottleneck wireless link ofWLc. The fairness conver-
gence is attained among the competing flows when the fairness index F(rs) moves towards
unity, i.e., rs1 = rs2. The dispatching rate of the rs1 (t) or rs1 (t) is below/above the fairness
line in graph (Fig.5) results in unfairness among sharing network bandwidth the compet-
ing flows. The F(rs) index is derived using Jain’s equation as
The Eq.18 can be modified as
(16)
𝛾
(t)=0⇒
(
1
r
i(t)+ri(t)
)
>rm(t
)
(17)
𝛾
(t)=1⇒
(1
2
∗ri(t)
)
<rm(t
)
(18)
F
(rs)=

rsi
2
n

r2
si
Fig. 6 Efficiency Convergence of DART vs. RFC6582 Wnd_ growth

M.J.A.Jude et al.
1 3
Altering Eq.(19) based on WI and WD variables
where
By substituting Eq.(14) and (15) in (20), the fairness index of the DART approach is
derived as
Table2 displays the fairness and efficiency analytical equations of proposed and exist-
ing approaches.
Figure7 displays the Wnd_ growth function of RFC 6582 and DART approaches in the CR
phase based on the fairness equations in Table2. The rm value is taken as 36; the initial ri value
for high utilized flow is taken as 18, and the low utilized flow is taken as 6. The graph shows
that DART accomplishes a minimum disparity among the competing flows within fewer RTTs.
(19)
F
(r(t+1)) =

r
si (t+1)
2
n

r
si
(t+1)2

(20)
F
(r(t+1)) =

WI+WDri(t)
2
n

W
I
+W
D
r
i
(t)

2
(21)
WI
≥
0, 0
≤
WD<1
(22)
F
(r(t+1)) =
∑4∗rm
ri
+1−rm
ri
∗ri(t)
2
n
∑
4∗rm
r
i
+1−rm
r
i
∗ri(t)
,ri(t)≤
𝛿
(23)
F
(r(t+1)) =
∑4∗rm
ri
+1
2∗ri(t)
2
n
∑
4∗rm
r
i
+1
2∗ri(t)
2,ri(t)
>𝛿
Table 2 DART, CERL + , RFC6582 and TCP-LoRaD fairness and efficiency equations
TCP variants Efficiency equation Fairness equation
DART
𝛾
(t)=0⇒

4∗rm
ri
+ri(t)

>

rm(t
)
4∗rm
ri
+1−
rm
ri
∗ri(t)
2
n
4∗rm
r
i
+1−
rm
ri
∗ri(t)

2,ri(t)≤
𝛿
𝛾
(t)=1⇒

1−
r
m
ri

∗ri(t)<rm(t);ri(t)≤
𝛿
=1⇒

1
2
∗ri(t)

<

rm(t);

ri(t)>𝛿
∑4∗rm
ri
+1
2∗ri(t)
2
n∑
4∗rm
r
i
+1
2∗ri(t)
2,ri(t)
>𝛿
CERL
+
TCP
−LoRaD
}
𝛾
(t)=0⇒

1
ri(t)+ri(t)

>

rm(t
)
∑1
ri(t)+ri(t)+ 1
2∗ri(t)
2
n∑
1
r
i
(t)+ri(t)+ 1
2∗ri(t)2
RFC 6582
𝛾
(t)=1⇒

1
2∗ri(t)

<

rm(t
)

On Minimizing TCP Traffic Congestion inVehicular Internet…
1 3
Conversely, RFC 6582 has a slower flow convergence rate and larger fairness disparity gap
between high and low-rate TCP flows.
The pseudocode for DART implementation in SS, CA and CR phases are given as.
3.2 Simulation Results andAnalysis
The DART approach performance is validated using the vehicular simulation approach
under varying node mobility and varying traffic load scenarios. The DART, RFC6582,
CERL + , and TCP-LoRaD congestion avoidance algorithms are evaluated based on aver-
age throughput, average end-to-end packet latency, fairness, and transmission efficiency
metrics. The experiments conducted using the network simulator (NS-3) [38, 39] in a

M.J.A.Jude et al.
1 3
closed road formation, as shown in Fig.8. The vehicle mobility traces are generated using
VanetMobiSim [40, 41] and ported into the simulation. The vehicular network features are
incorporated using WAVE library modules [42]. Table3 summarizes the parameters of
vehicular simulation experiments.
3.3 Mobility Analysis
The mobility analysis measures the congestion avoidance algorithm performance under
varying vehicle speed conditions. The vehicle speed varies between a minimum of 5m/s to
a maximum of 20m/s in closed road simulation experiments, and Fig.9 displays the simu-
lation outcomes. The DART approach achieves an average throughput of 9.82 Mbps in low
speed road conditions and yields an improvement of 26.57, 11.91, and 9.26% against the
RFC6582, TCP-LoRaD, and CERL + approaches. In the 20 m/s vehicle speed condition,
the DART attains an average throughput of 5.57 Mbps, and the existing methods encounter
a throughput lag of 49.55% (RFC6582), 36.98% (TCP-LoRaD), and 34.11% (CERL +). In
vehicular networks, frequent connection failure occurs due to vehicle mobility conditions,
Fig. 8 Closed road formation scenario in PyViz [43, 44] visualizer
Fig. 7 Fairness Convergence of DART vs. RFC6582 a DART Wnd_ growth and b RFC6582 Wnd_ growth

On Minimizing TCP Traffic Congestion inVehicular Internet…
1 3
triggering timeout at the sender side. Under such circumstances, the DART approach re-
transmits only the lost packet without invoking the rate adjustment that considerably stabi-
lizes the throughput rate. Furthermore, the DART rate adjustment based on T_bottle pre-
diction and faster convergence rate in the CR phase improves the throughput rate than the
existing approaches. However, the TCP-LoRaD and CERL + in high mobility conditions
attain a lesser average throughput performance due to its slow convergence rate in the CR
phase.
Similarly, the DART approach yields a lesser mean delay during packet transmis-
sion compared to the other approaches. The DART approach attains a 19.67, 14.74, and
9.26% lesser RTT packet delay than RFC6528, TCP-LoRaD, and CERL + approaches
in the low vehicle speed conditions. The DART approach attains similar mean delay
performance in the high speed conditions. During simulations, it is noted that the rate
adjustment based on T_bottle prediction allows the sender to push more data pack-
ets when the link delay is low, minimizing mean packet latency during transmission.
Conversely, the slow convergence rate and RTT fluctuation in bottleneck prediction of
CERL + and TCP-LoRaD increase the mean delay of data packets to finish the flow.
The number of successful data packets delivered to the destination determines the trans-
mission efficiency. The wireless link quality and the queuing delay are the two prime factors
Table 3 Vehicular simulation
parameters Simulation parameters
Vehicles 250
Simulation duration 900s
Simulation area 1500 × 1500m
Radiation pattern Omni Directional
Antenna elevation 1.5 Meters
Antenna gain 5dBi
Receiver sensitivity − 95dBm
Propagation type Nakagami
Channel width 10MHz
MAC IEEE 802.11p
Data rate 27Mbps
Modulation QAM
Coding rate 3/4
Transmission technique OFDM
Subcarriers 52
Sub carrier frequency 156.25kHz
Symbol interval 8µsec
Control channel 50ms
Service channel 50ms
Operating frequency 5.9GHz
System loss 1
Vehicle speed 5, 10,15 and 20m/s
Traffic 2,4,6, and 8
Wnd_ size 256
Packet size 1400 Bytes

M.J.A.Jude et al.
1 3
Fig. 9 Congestion avoidance approaches performance under mobility conditions a average throughput b
mean delay c transmission efficiency and d fairness
Fig. 10 Congestion avoidance approaches performance under traffic load conditions a average throughput b
mean delay c transmission efficiency and d fairness

On Minimizing TCP Traffic Congestion inVehicular Internet…
1 3
determining packet transmission efficiency in vehicular networks. The DART approach
attains an average efficiency between 97.31 and 99.12%, and the existing approaches reach
efficiency between 95.25 and 98.11% in diverse mobility conditions. The fairness metric
evaluates equitable bandwidth distribution among the competing TCP flows in the network.
The DART approach attains average bandwidth fairness of 95.25% in diverse mobility.
Conversely, the existing approaches achieve 92.09% bandwidth fairness among traffic
flows in mobility conditions. The DART approach proportionate rate adjustment and SLC
algorithm significantly contribute to the equitable sharing of network bandwidth among
different flows. However, the fairness capability of the existing approach is crippled pri-
marily due to its flat-rate decrement mechanism and Wnd_ growth pattern in the CR phase.
3.4 Traffic Load Analysis
The traffic load analysis measures the performance of the congestion avoidance approach
under varying TCP traffic conditions. The experiments were conducted by increasing the traf-
fic load from two to eight pairs of a source–destination process with an average vehicle speed
of 10m/s. Figure10 displays the outcomes of the proposed and existing approaches. The
DART approach attains the average throughput rate of 10.12 and 7.37Mbps under minimum
and maximum traffic load conditions, with 29.11, 12.68, and 10.23% improvement against
RFC6582, TCP-LoRaD, and CERL + approaches under varying load conditions. Further-
more, the DART approach congestion prediction assisted proportionate rate adjustment, and a
faster convergence rate considerably contributes to achieving a higher throughput rate.
During packet transmission under diverse traffic load conditions, the DART approach
attains 22.44, 17.04, and 13.09% lesser mean delay than RFC6528, TCP-LoRaD, and
CERL + approaches. The DART approach Wnd_ growth pattern allows the sender to the
swift packet transfer during the low link delay conditions that minimize mean packet
delay. Conversely, the existing approaches’ slow transfer rate in the CR phase increases
the mean delay of data packets to finish the flow.
The DART packet transmission yields a higher efficiency of 95.32–99.12% in diverse
load conditions due to its capability to infer queuing delay that minimizes packet loss
associated with buffer overflows. However, transmission efficiency is degraded due
to radio channel loss that induces considerable packet drop in the DART and existing
approaches under vehicular networks. Similarly, DART proportionate rate decrement
and SLC algorithm in the CR phase allow multiple flows to attain a 95.32% fair dis-
tribution of resources. However, the traditional Wnd_ growth pattern of similar AIMD
approaches achieves 92.56% fairness under diverse traffic conditions.
4 Conclusion
The DART congestion avoidance approach proposed in this article intends to improve TCP
traffic throughput, convergence rate, and equitable bandwidth distribution under VIoT net-
works. The DART initiates a proportionate rate adjustment mechanism based on the T_bot-
tle parameter that considerably lessens needless throughput reduction during packet losses in
wireless conditions. Furthermore, the modified Wnd_ growth pattern in the CR phase signifi-
cantly boosts the convergence rate and maintains the equitable distribution of network band-
width among low and high utilized flows. The analytical equations derived based on Jain’s
model verifies the improvement in fairness and convergence rate performance over the existing

M.J.A.Jude et al.
1 3
schemes. The experiment outcome under two scenarios proves that DART attains considerable
throughput, transmission efficiency, and equitable bandwidth distribution improvement with
the least mean delay in packet transmission. The proposed DART approach’s future variant is
customized to support the multipath TCP implementation under vehicular networks.
Acknowledgements To the Self Organised Networking Group (SONG) research members, who had spent
more than 250 person-hours to perform vehicular simulation at Vinton Network Lab, ECE department,
Kongu Engineering College.
Funding This research did not receive any specific grant from funding agencies in the public, commercial,
or not-for-profit sectors.
Data Availability The datasets analyzed during the current study are not publicly available, compromising-
our future research programs. Still, they are available from the corresponding author on reasonable request.
Declarations
Conflict of interest The authors hereby acknowledge that there is no conflict of interest.
References
1. Ji, B., Zhang, X., Mumtaz, S., Han, C., Li, C., Wen, H., & Wang, D. (2020). Survey on the internet
of vehicles: Network architectures and applications. IEEE Communications Magazine, 4(1), 34–41.
https:// doi. org/ 10. 1007/ 978- 981- 10- 3376-6_ 48
2. Manojkumar, P., Suresh, M., Ahmed, AAA., Panchal, H., Rajan, CA., Dheepanchakkravarthy, A.,
Geetha, A., Gunapriya, B., Mann, S., Sadasivuni, KK. (2021) A novel home automation distrib-
uted server management system using Internet of Things International Journal of Ambient Energy-
https:// doi. org/ 10. 1080/ 01430 750. 2021. 19535 90
3. Kaushik, S., Srinivasan, K., Sharmila, B., Devasena, D., Suresh, M., Panchal, H., Ashokkumar,
R., Sadasivuni, KK., & Srimali, N. (2021). Continuous monitoring of power consumption in urban
buildings based on Internet of Things. International Journal of Ambient Energy, 1https:// doi. org/
10. 1080/ 01430 750. 2021. 19319 61
4. Rajamoorthy, R., Arunachalam, G., Kasinathan, P., Ramkumar Devendiran, P., Ahmadi, S. P.,
Muthusamy, S., Panchal, H., Kazem, H. A., & Sharma, P. (2022). A novel intelligent transport
system charging scheduling for electric vehicles using grey wolf optimizer and sail fish optimiza-
tion algorithms. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 44(2),
3555–3575. https:// doi. org/ 10. 1080/ 15567 036. 2022. 20672 68
5. Lu, H., Liu, Q., Tian, D., Li, Y., Kim, H., & Serikawa, S. (2019). The cognitive internet of vehicles
for autonomous driving. IEEE Network, 33, 65–73. https:// doi. org/ 10. 1109/ MNET. 2019. 18003 39
6. Arena, F., Pau, G., & Severino, A. (2020). A review on IEEE 802.11 p for intelligent transportation
systems. Journal of Sensor and Actuator Networks, 9(2), 22. https:// doi. org/ 10. 3390/ jsan9 020022
7. Kaffash, S., Nguyen, A. T., & Zhu, J. (2021). Big data algorithms and applications in intelligent
transportation system: A review and bibliometric analysis. International Journal of Production
Economics, 231, 107868. https:// doi. org/ 10. 1016/j. ijpe. 2020. 107868
8. Lin, Y., Wang, P., Ma, M. (2017) Intelligent transportation system (ITS): Concept, challenge and
opportunity, In: Proc.-3rd IEEE Int. Conf. Big Data Secur. (pp. 167–172). https:// doi. org/ 10. 1109/
BigDa taSec urity. 2017. 50.
9. Thompson, A. W., & Perez, Y. (2020). Vehicle-to-Everything (V2X) energy services, value streams,
and regulatory policy implications. Energy Policy, 137, 111–136. https:// doi. org/ 10. 1016/j. enpol.
2019. 111136
10. Moubayed, A., Shami, A., Heidari, P., Larabi, A., & Brunner, R. (2020). Edge-enabled V2X service
placement for intelligent transportation systems. IEEE Transactions on Mobile Computing, 20(4),
1380–1392. https:// doi. org/ 10. 1109/ TMC. 2020. 29659 29
11. Abdel Hakeem, S. A., Hady, A. A., & Kim, H. (2020). Current and future developments to improve
5G-newradio performance in vehicle-to-everything communications. Telecommunication Systems,
75(3), 331–353. https:// doi. org/ 10. 1007/ s11235- 020- 00704-7

On Minimizing TCP Traffic Congestion inVehicular Internet…
1 3
12. Laña, I., Sanchez-Medina, J. J., Vlahogianni, E. I., & Del Ser, J. (2021). From data to actions in
intelligent transportation systems: A prescription of functional requirements for model actionabil-
ity. Sensors, 21(4), 1121. https:// doi. org/ 10. 3390/ s2104 1121
13. Mir, Z. H., Toutouh, J., Filali, F., & Ko, Y. B. (2020). Enabling DSRC and C-V2X integrated hybrid
vehicular networks: Architecture and protocol. IEEE access, 8, 180909–180927. https:// doi. org/ 10.
1109/ ACCESS. 2020. 30270 74
14. Silva, L., Magaia, N., Sousa, B., Kobusińska, A., Casimiro, A., Mavromoustakis, C. X., & De
Albuquerque, V. H. C. (2021). Computing paradigms in emerging vehicular environments: A
review. IEEE/CAA Journal of Automatica Sinica, 8(3), 491–511. https:// doi. org/ 10. 1109/ JAS. 2021.
10038 62
15. Sadovaya, Y., & Zavjalov, S. V. (2020). Dedicated short-range communications: Performance evalua-
tion over mmWave and potential adjustments. IEEE Communications Magazine, 24(12), 2733–2736.
https:// doi. org/ 10. 1109/ LCOMM. 2020. 30166 34
16. SenthamilSelvan, R., Wahidabanu, R. S. D., & Karthik, B. (2022). Intersection collision avoidance in
dedicated short-range communication using vehicle ad hoc network. Concurrency and Computation,
34(13), e5856. https:// doi. org/ 10. 1002/ cpe. 5856
17. Vega, C., Roquero, P., & Aracil, J. (2017). Multi-Gbps HTTP traffic analysis in commodity hardware
based on local knowledge of TCP streams. Computer Networks, 113, 258–268. https:// doi. org/ 10.
1016/j. comnet. 2017. 01. 001
18. Murray, D., Koziniec, T., Zander, S., Dixon, M., Koutsakis, P. (2018) An analysis of changing enter-
prise network traffic characteristics, In: 2017 23rd Asia-Pacific Conf. Commun. Bridg. Metrop.
Remote. APCC 2017. 2018-January (pp.1–6). https:// doi. org/ 10. 23919/ APCC. 2017. 83039 60.
19. Zhang, T., Wang, J., Huang, J., Chen, J., Pan, Y., & Min, G. (2017). Tuning the aggressive TCP behav-
ior for highly concurrent HTTP connections in intra-datacenter. IEEE/ACM Transactions on Network-
ing, 25, 3808–3822. https:// doi. org/ 10. 1109/ TNET. 2017. 27593 00
20. Kogias, M., & Bugnion, E. (2018). Flow control for latency-critical rpcs. In: Proceedings of the 2018
Afternoon Workshop on Kernel Bypassing Networks (pp. 15-21). https:// doi. org/ 10. 1145/ 32295 38.
32295 41
21. Kanellopoulos, D. (2019). Congestion control for MANETs: An overview. ICT Express, 5, 77–83.
https:// doi. org/ 10. 1016/j. icte. 2018. 06. 001
22. Henderson, T., Floyd, S., Gurtov, A., Nishida, Y. (2012) RFC 6582: The NewReno modification to
TCP’s fast recovery algorithm, Rfc. (pp. 1689–1699). https:// doi. org/ 10. 17487/ RFC65 82
23. Molia, H. K., & Kothari, A. D. (2018). TCP variants for mobile adhoc networks: Challenges
and solutions. Wireless Personal Communications, 100, 1791–1836. https:// doi. org/ 10. 1007/
s11277- 018- 5675-8
24. Saedi, T., & El-Ocla, H. (2021). TCP CERL+: Revisiting TCP congestion control in wireless networks
with random loss. Wireless Networks, 27, 423–440. https:// doi. org/ 10. 1007/ s11276- 020- 02459-0
25. Abdelsalam, A., Luglio, M., Roseti, C., Zampognaro, F., & Wave, T. C. P. (2017). A new reliable
transport approach for future internet. Computer Networks, 112, 122–143. https:// doi. org/ 10. 1016/j.
comnet. 2016. 11. 002
26. Wang, J., Wen, J., Zhang, J., Xiong, Z., & Han, Y. (2016). TCP-FIT: An improved TCP algorithm for
heterogeneous networks. Journal of Network and Computer Applications, 71, 167–180. https:// doi. org/
10. 1016/j. jnca. 2016. 03. 020
27. Sarkar, N. I., Ho, P. H., Gul, S., & Zabir, S. M. S. (2022). TCP-LoRaD: A loss recovery and differen-
tiation algorithm for improving TCP performance over MANETs in noisy channels. Electronics, 11(9),
1479. https:// doi. org/ 10. 3390/ elect ronic s1109 1479
28. Zhang, J., Wen, J., & Han, Y. (2017). TCP-ACC: Performance and analysis of an active congestion
control algorithm for heterogeneous networks. Frontiers of Computer Science, 11, 1061–1074. https://
doi. org/ 10. 1007/ s11704- 016- 5482-x
29. Brakmo, L. S., Peterson, L. L., & Vegas, T. C. P. (1995). End to end congestion avoidance on a global
internet. IEEE Journal on Selected Areas in Communications, 13, 1465–1480. https:// doi. org/ 10. 1109/
49. 464716
30. Guan, S., Jiang, Y., & Guan, Q. (2020). Improvement of TCP Vegas algorithm based on forward direc-
tion delay. International Journal of Web Engineering and Technology, 15(1), 81–95. https:// doi. org/ 10.
1504/ IJWET. 2020. 107690
31. Jamali, S., Alipasandi, N., Alipasandi, B., & Pegas, T. C. P. (2015). A PSO-based improvement over
TCP Vegas. Applied. Soft Computing, 32, 164–174. https:// doi. org/ 10. 1016/j. asoc. 2015. 03. 048
32. Jiang, H., Luo, Y., Zhang, Q. Y., Yin, M. Y., & Wu, C. (2017). TCP-Gvegas with prediction and
adaptation in multi-hop ad hoc networks, Wirel. Networks, 23, 1535–1548. https:// doi. org/ 10. 1007/
s11276- 016- 1242-y

M.J.A.Jude et al.
1 3
33. Sunitha, D., Nagaraju, A., & Narsimha, G. (2017). Dynamic TCP-Vegas based on cuckoo search for
efficient congestion control for MANET. International. Journal of Signal Imaging Systems Engineer-
ing, 10, 47–53. https:// doi. org/ 10. 1504/ IJSISE. 2017. 084570
34. Janowski, R., Grabowski, M., & Arabas, P. (2020). New heuristics for TCP retransmission timers. Pro-
gress in computer recognition systems (pp. 117–129). Springer International Publishing. https:// doi.
org/ 10. 1007/ 978-3- 030- 19738-4_ 13
35. Zhao, Y., Cheng, G., Zhang, W., Chen, X., & Li, J. (2020). Bctcp: A feedback-based congestion control
method. China Communications, 17(6), 13–25. https:// doi. org/ 10. 23919/ JCC. 2020. 06. 002
36. de Carvalho, J. A., da Costa, D. B., Yang, L., Alexandropoulos, G. C., Oliveira, R., & Dias, U. S. (2020).
User fairness in wireless powered communication networks with non-orthogonal multiple access. IEEE
Wireless Communication Letters, 10(1), 189–193. https:// doi. org/ 10. 1109/ LWC. 2020. 30308 18
37. Joseph Auxilius Jude, M., Diniesh, V. C., & Shivaranjani, M. (2020). Throughput stability and flow
fairness enhancement of TCP traffic in multi-hop wireless networks. Wireless Networks, 26(6), 4689–
4704. https:// doi. org/ 10. 1007/ s11276- 020- 02357-5
38. Ceballos, H. Z., Amaris, J. E. P., Jiménez, H. J., Rincón, D. A. R., Rojas, O. A., & Triviño, J. E. O.
(2021). Network Simulating Using ns-3. Wireless Network Simulation (pp. 65–95). Apress. https:// doi.
org/ 10. 1007/ 978-1- 4842- 6849-0_4
39. NS-3 Network Simulator. (2022). Discrete-event network simulator. Retrieved Sept 12, 2022, from-
https:// www. nsnam. org/
40. Tian, J., & Meng, F. (2020) Comparison survey of mobility models in vehicular ad-hoc network
(vanet). In2020 IEEE 3rd International Conference on Automation, Electronics and Electrical Engi-
neering (AUTEEE) 337–342. IEEE. https:// doi. org/ 10. 1109/ AUTEE E50969. 2020. 93155 83
41. VanetMobiSim. (2022). Vehicular Ad hoc Network mobility extension to the CanuMobiSim frame-
work. Retrieved Sept 12, 2022, from http:// vanet. eurec om. fr/
42. Drago, M., Zugno, T., Polese, M., Giordani, M., & Zorzi, M. (2020) MilliCar: An ns-3 module for
mmWave NR V2X networks. In Proceedings of the 2020 Workshop on ns-3, (pp. 9–16). https:// doi.
org/ 10. 1145/ 33894 00. 33894 02
43. Šljivo, A., Kerkhove, D., Moerman, I., De Poorter, E., & Hoebeke, J. (2018) Interactive web visualizer
for IEEE 802.11 ah ns-3 module. In Proceedings of the 10th Workshop on Ns-3 (pp.23–29). https://
doi. org/ 10. 1145/ 31999 02. 31999 04
44. NS-3 PyViz. (2022). Live simulation visualizer. Retrieved Sept 12, 2022, https:// www. nsnam. org/ wiki/
PyViz
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the
author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is
solely governed by the terms of such publishing agreement and applicable law.
M. Joseph Auxilius Jude received his B.E in Electronics and Commu-
nication Engineering, M.E with specialization in Digital Communica-
tion and Networking, and Ph.D. under Anna University, Chennai,
India, in 2005, 2007, and 2019 respectively. He is working as Associ-
ate Professor and associated with the Self Organized Networking
Group (SONG) of the ECE department at Kongu Engineering College.
He is closely related to developing advanced wireless networking lab
and organized funded industryaligned training programmes in alliance
with the Defence Research & Development Organisation (DRDO),
Department of Science & Technology (DST), and Industry Insitute
Partnership Cell. He is a peer reviewer for IEEE Access, IEEE Sensor
Letters, Wiley International Journal of Communication Systems, and
Springer Wireless Personal Communications. He also acted as a men-
tor for the collaborative project on vehicular communication with UTP
Malaysia. His research interests include design, testing, and validating
scheduling algorithms for wireless sensor networks, congestion control
algorithms (TCP and AQM) for the MANET and VANET. He is a
member of IEEE (Computer Society and Sensor Council), a life member of IETE, and institution chair of
IEEE computer society.

On Minimizing TCP Traffic Congestion inVehicular Internet…
1 3
V. C. Diniesh received his B.E in Electronics and Communication
Engineering and M.E with specialization in Embedded and Real Time
System under Anna University, Chennai, India in 2009 and 2011
respectively. He is currently pursing PhD in the area of wireless ad-hoc
and sensor networks. He is with Self Organized Networking Group
(SONG) of the ECE department at Kongu Engineering College. He is
closely connected with the routing protocol development and training
for tactical wireless networks especially for disaster and battlefield
management, organized funded programmes in collaboration with
DST and Industry Insitute Partnership. His research interests include
designing, testing and validation of routing protocols for mobile ad-
hoc network and wireless sensor network. He is a member of IEEE and
life member of IETE.
M. Shivaranjani completed her Master Degree in Communication Sys-
tems, from Kongu Engineering College affiliated to Anna University,
Chennai. She received her Bachelor Degree in Electronics and Com-
munication Engineering from Kongu Engineering College affiliated to
Anna University, Chennai. She is with Self Organized Networking
Group (SONG) and currently working as an assistant professor in
Velalar College of Engineering and Technology. Her current research
focuses on designing Transmission Control Protocol (TCP) for Wire-
less Multi-hop Networks especially vehicular ad hoc network and
Mobile ad hoc network.
Suresh Muthusamy received the bachelor’s degree in Electrical and
Electronics Engineering and the master’s degree in Power Electronics
and Drives during the year 2009 and 2011 from Anna University,
Chennai and Anna University, Coimbatore respectively. Currently, he
is working towards the Ph.D. degree in Electrical Engineering at Anna
University, Chennai in the area of Hybrid Renewable Energy Systems.
He worked as Assistant Professor in the Department of Electrical and
Electronics Engineering at Kongu Engineering College (Autonomous),
Perundurai, Erode during the period June 2011 to January 2020. From
January 2020 onwards, he has been working as Assistant Professor
Senior Grade in the Department of Electronics and Communication
Engineering at Kongu Engineering College (Autonomous), Perun-
durai, Erode. He published more than 85 research articles in well
reputed and refereed international journals like Elsevier, Springer,
Taylor & Francis, SAGE, ASME, ASTM international, MDPI, etc &
indexed in SCI, SCIE, ESCI, Scopus and Web of Science with good
impact factor. He presented several research articles in national and
international conferences and also serving as the reviewer, editor for about 52 international journals includ-
ing IET Renewable Power Generation, IET Journal of Engineering, etc. To his credit, he has filed and pub-
lished 28 Indian patents in IPR website, governed by Ministry of Commerce & Industry, Government of
India. His areas of interests include hybrid renewable energy systems, power electronic converters, hybrid
electric vehicles and battery management systems.

M.J.A.Jude et al.
1 3
Hitesh Panchal received the bachelor’s degree in Mechanical Engi-
neering from Government Engineering College, Modasa, Gujarat dur-
ing the year 2004 and the master’s degree in Internal Combustion
Engines and Automobile from L.D. College of Engineering,
Ahmedabad, Gujarat during the year 2006. He then received the Ph.D
degree in Mechanical Engineering in the area of Solar Thermal Engi-
neering from K.S.V. University, Gandhinagar, Gujarat in 2015. Dr.
Panchal has published more than 150 research articles in reputed
National and International Journal publications like Taylor and Fran-
cis, Springer, Elsevier, ASME, SAGE, etc. and also received 5000+
Google Scholar citations with h-index of 47 and i-10 index of 113.
Dr.Panchal received many prestigious awards like Best Ph.D Thesis
Award, Bharat Excellence Award, Young Scientist Award, Bright
Researchers Award and many more from various organizations.
Dr.Panchal has completed 2 Research projects from GUJCOST, DST
of 15 lakhs and currently working on one Research project from DTE.
Dr.Panchal has filed 24 full patents & 10 Industrial designs and
among them 2 patents & 7 industrial designs has been granted. Dr.Panchal has been selected in “Top 2%
Indian Scientist under the energy category” published by Stanford University for the year 2020 and 2021.
Suma Christal Mary Sundararajan received the master’s degree in
Computer Science and Engineering from Francis Xavier Engineering
College, Tirunelveli and Ph.D. degree in Computer Science and Engi-
neering from Kalasalingam University, Srivilliputhur in 2016. She is
currently working as Professor in the Department of Information Tech-
nology at Panimalar Engineering College, Poonamallee, Chennai. She
has published more than 25 research articles in Scopus, SCI indexed
journals and presented 40 papers in national and international confer-
ences. She has received Young Scientist award from Computer Society
of India and received Best Project award from Dr.Kalam Educational
Trust, Best Teacher Award from IEAE. Her areas of interest include
Network Security, Neural Networks, IoT, Virtual reality and Soft Com-
puting, etc.
Kishor Kumar Sadasivuni is working in Center for Advanced Materi-
als, Qatar University. He has published 82 Journal papers, 12 book
chapters, 7 books edited and 3 patents filed. He has about 10 years of
experience in synthesis & characterization of nanoparticles and also in
manufacturing nanocomposites for industrial applications. His areas of
interest include different types of nanocomposite fabrication, modifica-
tions, designs and their applications especially sensors, piezoelectrics,
actuators, energy storage, Dielectrics, 3D-Printing and flexible
electronics

On Minimizing TCP Traffic Congestion inVehicular Internet…
1 3
Authors and Aliations
M.JosephAuxiliusJude1 · V.C.Diniesh1 · M.Shivaranjani2 ·
SureshMuthusamy3 · HiteshPanchal4 · SumaChristalMarySundararajan5 ·
KishorKumarSadasivuni6
V. C. Diniesh
vcdiniesh@gmail.com
M. Shivaranjani
ranjani093@gmail.com
Suresh Muthusamy
infostosuresh@gmail.com
Hitesh Panchal
engineerhitesh2000@gmail.com
Suma Christal Mary Sundararajan
sumasheyalin@gmail.com
Kishor Kumar Sadasivuni
kishor_kumars@yahoo.com
1 Vinton Network Lab, Self Organised Networking Group (SONG), Department ofElectronics
andCommunication Engineering, Kongu Engineering College (Autonomous), Perundurai,
TamilNadu, India
2 Department ofElectronics andCommunication Engineering, Velalar College ofEngineering
andTechnology (Autonomous), Thindal,Erode, TamilNadu, India
3 Department ofElectronics andCommunication Engineering, Kongu Engineering College
(Autonomous), Perundurai, Erode, TamilNadu, India
4 Department ofMechanical Engineering, Government Engineering College, Patan, Gujarat, India
5 Department ofInformation Technology, Panimalar Institute ofTechnology, Poonamallee, Chennai,
TamilNadu, India
6 Centre forAdvanced Materials, Qatar University, Doha, Qatar