Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Available online at www.worldscientificnews.com
( Received 17 January 2019; Accepted 05 February 2019; Date of Publication 06 February 2019 )
WSN 120(2) (2019) 154-167 EISSN 2392-2192
Cross Layer Enhanced Adaptive Handoff Initiation Technique
Jeremiah O. Abolade1,a, Augustus E. Ibhaze2,b, Agbotiname L. Imoize2,c,
Aderemi A. Atayero1,d
1Department of Electrical and Information Engineering, Covenant University, Ota, Nigeria
2Department of Electrical and Electronics Engineering, University of Lagos, Akoka-Lagos, Nigeria
a-dE-mail address: [email protected] , [email protected] , [email protected] , [email protected]
ABSTRACT
The population of mobile communication subscribers keeps increasing exponentially with a
growing industrial interest in the provision of improved quality of service. To maintain quality of service
in a heterogeneous network built for high capacity utilization, the communication network must be
stable as mobile users interact within the implemented coverage distribution. One of the basic
mechanisms behind this communication metric is its ability to communicate wirelessly with respect to
changing locations. Mobility is one critical benefit of mobile communication, which is made possible
through the process of Handoff. One of the key performances in mobile communication is Handoff
failure (call termination) rate, which is dependent critically on the handoff initiation mechanism. In this
paper, the effect of traditional single parameter of either signal-to-noise ratio (SNR) or throughput on
handover threshold and hysteresis was investigated. Consequently, a cross layer enhanced adaptive
handoff initiation algorithm using fuzzy logic was proposed. The proposed method exploits the
capability of soft computing to handle the unpredictable nature of the radio propagation environment.
Four critical handoff Network-link parameters were employed to vary the Handoff threshold and
Hysteresis in test the adaptive functionality of the algorithm. The proposed algorithm was seen to be
adaptive to throughput, SNR, mobile transmitting power (MTP) and mobile velocity, thereby resulting
in improved quality of service, marked with a decrease in handoff failure probability rate.
World Scientific News 120(2) (2019) 154-167
-155-
Keywords: Base station, Fuzzy Inference, Handoff, Mobile Transmitting Power, signal-to-noise ratio,
Throughput, Velocity
1. INTRODUCTION
Mobility is one the characteristics of a cellular system. Even though the mobile phone
users are moving in and out of the cells, communication among users is still sustained. The
active call should be transferred from one cell to another in order to avoid call termination
during boundary crossing levels. This is accomplished by handoff process. As technology
integration evolves from single infrastructure technology to heterogeneous architecture of
evolving technologies, Oboyerulu et al. [1] establishes the need for efficient handoff algorithms
in terms of vertical and horizontal handoff techniques within the heterogeneous network. The
horizontal handover is an Intra-technology based handoff that is, handoff between Node B’s of
the same network technology as shown in Fig. 1, when the UE is transferred from one WLAN
access point to another. Vertical handover is an inter-technology based handover as shown in
Fig. 1, when WLAN transfer UE to LTE network.
Handoff is a mechanism of transferring ongoing calls from one cell to another or from
one MSC to another or one channel to another as users move through the coverage area of
diverse serving Node B’s [2]. Handover is the key process in wireless networks to determine
users QoS experience during mobility [3] [4]. Handoff is a growing research area in the field
of mobile WiMAX [5], HSPA+ and LTE [6]. According to Hamad [7], handover (HO) is a
concept, which aims to grant stability of connection while crossing diverse networks. With the
exponential growth rate in wireless mobile communication utilization, Beach [8] supposes that
the system requirements for higher data rate also increases exponentially due to the invention
of smart device technologies. The number of users and their demand for better resources
increases daily so today’s 3G is insufficient meeting these demands.
Figure 1. Horizontal and Vertical Handover Mechanism
Vertical Horizontal
World Scientific News 120(2) (2019) 154-167
-156-
Conventional Handoff decision are generally single metric based which is not sufficient
for handoff initiation in the present sophisticated networks and environment (especially in an
urban area) [9]. The commonly used parameter in conventional handoff is Received Signal
Strength Indicator (RSSI), which measures the power present in a received radio signal [10].
Pollini [11] developed an algorithm that could initiate the handoff process based on the received
signal strength although limited by the ping-pong effect experienced in overlapping cell
coverage regions; resulting in unnecessary handoffs. Zhang and Holtzman [12] integrated a
margin between a couple of thresholds to eliminate the ping-pong effect while Marichamy et
al. [13] improved on the system by considering the application of hysteresis. In an attempt to
improve on the received signal strength-based handoff algorithm, Pahlavan et al. [14]
developed an algorithm based on artificial neural network to determine when to initiate handoff.
For better quality of service, an algorithm which is network-environment sensitive is needed.
Soft computing techniques deal with the uncertainty, partial truth, and approximation, and
depend on expert advice to reduce computational cost [15]. The aim of this work is to develop
a cross-layer sensitive handoff initiation algorithm to solve the real world problem of handoff.
This work only covers the handoff initiation phase as it serves as the basis for any handoff
optimization algorithm. The focus of this work is on how to effectively determine when handoff
should be requested rather than how it should be requested.
2. RELATED WORKS
Lal and Panwar [16], propose a dynamic hysteresis margin based handoff algorithm,
which uses distance between the mobile station, and base station and simulated in micro-cellular
environment. The authors used analytical method to test the algorithm performance using
handoff probability as the performance metric. Chang and Chen [2] proposed an adaptive
Hysteresis based horizontal handoff using distance between the BS and the MT to vary the
hysteresis value. The authors applied Okumura-Hata model, but several things apart from
distance can lead to degradation of signal such as co-channel interference, channel fading, etc.
Therefore, the algorithm can only be effective for a noiseless channel, which is an ideal
situation. Çeken et al. [17] proposed a fuzzy-based decision algorithm, which uses: user
preference, bandwidth, RSS, quality of service and network coverage. The algorithm, due to
the reduced Ping-Pong effect, reduced the number of unnecessary handoffs. Note that this
scheme only reduces the Ping-Pong effect and does not remove it. Mir and Bhatti [18] proposed
a Fuzzy-based handoff decision approach based on time triggering. Handoff design metrics
such as delay, throughput and number of handoffs were articulated among others. However,
handoff decision has to be performed by continuously sensing the arrival and departure patterns
of registered users. In Stemm and Katz [19] and Min-Hua et al. [20], more sophisticated
handover procedures, which considered more parameters apart from RSS, were investigated.
Chan et al. [21] established that fuzzy logic is advantageous in dealing with complexity of the
information in heterogeneous mobile access network.
Val et al. [22] proposed a spectrum handoff mechanism built with the capability of
interference detection and node switching within a predefined time threshold. The algorithm
was implemented using a Field Programmable Gate Array (FPGA) and OPNET simulator in
order to characterize the performance metric. The authors suggested more parameters to be
added in order to provide better quality of service. It is also noteworthy that the work did not
World Scientific News 120(2) (2019) 154-167
-157-
indicate at elimination or minimization of Ping-Pong effect. In Aibinu et al. [23], the authors
used several input parameters to determine the handoff possibility in wireless mobile
communication by cascading Artificial Neural Network (ANN) and Fuzzy Logic for handover
decision. The fuzzy rule used in the work was adjudged able to make necessary handoff
decisions while avoiding the Ping-Pong effect when evaluated. Çalhan and Çeken [24]
proposed a fuzzy based vertical handoff decision controller by taking best network parameters
among available networks and compared those parameters with present network parameters as
set by vertical handoff decision controller for vertical handoff decision. In the vertical handoff
proposed by Jain and Tokekar [25], handover decision is dependent on coverage area and
velocity. The algorithm is applicable where speed is critical and it was simulated for particular
coverage range of network during which handoff is beneficial. The work reduces the
unnecessary handoff, which results in an improved QoS but only limited by the number of
factors considered; which are not enough in making handoff decision. For this reason,
researchers should consider more parameters for proper vertical handoff decision. In [7], the
author proposed a handover scheme between GSM and WiMAX mobile system in physical
layer mode. This solution provided a collection of high data rates, high mobility, and solving
the problem of traffic congestion in cellular mobile network using bit as the metric parameter.
3. TRADITIONAL ADAPTIVE BASED HANDOFF SIMULATION
3. 1. Signal-to-Noise Ratio (SNR) Based Handoff
The changing parameter is chosen to be Signal-to-Noise Ratio (SNR) while the speed,
throughput (best-case scenario), and Mobile Transmitting Power (MTP) are 0 km/h, 100%,
and 20 dBm, respectively as shown in Fig. 2.
Figure 2. Varying SNR while M.T.P., velocity and throughput are kept constant
0 10 20 30 40 50 60 70 80 900
10
20
30
40
50
60
70
80
90
100
No of Sample
SN
R/M
.T.P
./V
elo
city/T
hro
ugh
pu
t
SNR (dB)
M.T.P. (dBm)
Velocity (dBm)
Throughput (%)
World Scientific News 120(2) (2019) 154-167
-158-
Figure 3. Handover Threshold vs SNR
Figure 4. Handover Hysteresis vs SNR
When the speed is maintained at 0 km/h, the throughput is kept at 100%, the mobile
Transmitting Power is also kept at 20 dBm and SNR changes, the handoff threshold vary
between -87.5 dBm to -107 dBm as shown in Fig. 3. In addition, the Hysteresis changes along
with respect to the SNR as shown in Fig. 4.
3. 2. Throughput Based Handoff
Fig. 5 shows the graph of SNR, MTP, and Velocity at 35 dB, 20 dBm, 0 km/h (best case
Scenario), respectively while the Throughput varies. It is worth noting that, the Handoff
10 15 20 25 30 35-110
-105
-100
-95
-90
-85
SNR (dB)
Ha
nd
ov
er T
hre
sho
ld (
dB
)
Graph of Handover Threshold Against SNR
10 15 20 25 30 356.5
7
7.5
8
8.5
9
9.5
Graph of Handover Hysteresis Against SNR
SNR (dB)
Han
do
ver
Hy
ster
esis
(d
B)
World Scientific News 120(2) (2019) 154-167
-159-
threshold and hysteresis continuously vary with respect to the variation in the throughput as
shown in Fig. 6 and Fig. 7, respectively. It is also seen that a sharp fall exist in the handoff
threshold of Throughput-based handoff when compared with SNR-based handoff which shows
the critical nature of the Throughput in the algorithm.
Figure 5. Varying Throughput while M.T.P., Velocity and SNR are kept constant
Figure 6. Handoff Threshold vs Throughput
World Scientific News 120(2) (2019) 154-167
-160-
Figure 7. Handoff Hysteresis vs Throughput
3. 3. Methodology
In this paper, four parameters were used as opposed to single parameter-based algorithms
used in traditional adaptive handoff systems. The parameters are Signal to noise ratio (SNR),
Mobile Transmission Power (MTP), Speed, and Throughput. These parameters are the inputs
to the Fuzzy Inference System to produce Handoff Hysteresis and Threshold as the output of
the fuzzy logic shown in Fig. 8. The implementation of this algorithm of four input parameters
in deciding the two metrics for handoff decision is as shown in Fig. 9.
Figure 8. Block diagram of the Methodology
World Scientific News 120(2) (2019) 154-167
-161-
Figure 9. FIS for threshold and Hysteresis
Each of the input variables is divided into three fuzzy sets each defined with three
linguistic variables, which are LOW, NEUTRAL, and HIGH. Seven linguistic variables were
used for the outputs, which are LOWEST, LOWER, LOW, NEUTRAL, HIGH, HIGHER, and
HIGHEST. The membership functions used are Gauss and Triangular Membership Functions.
These membership functions were used because they are suitable for the complex situation at
hand as well as for efficient computation.
The philosophy behind the fuzzy rules:
i. If the Throughput is “LOW”, the SNR is “LOW”, the MTP is “HIGH”, and the Speed
is “HIGH”. For this, the handoff should be encouraged as fast as possible. Therefore,
the Threshold should be increased to “HIGHEST” and the Hysteresis should be
“LOWEST”.
ii. If the Throughput is “HIGH”, the SNR is “HIGH”, the MTP is “LOW”, and the Speed
is “LOW”. For this, the handoff should be discouraged. Therefore, the threshold should
be decreased to “LOWEST” and the Hysteresis should be increased to “HIGHEST”.
Other rules philosophy states that if there is more support for encouraging a handoff, the
threshold is improved and the hysteresis is decreased. Otherwise, there is more support for
handover discouragement by the input parameter, there would be decrease in threshold and
increase in hysteresis. The degree of how FIS output change for encouragement or
discouragement is dependent on the number of input variables that support the decision. The
higher the rules the better performance, which reduces the quantity of unnecessary handover
and delay, involved in handover [3].
The membership functions (MF) used in the proposed algorithm for inputs and output are
Gauss and Triangular membership functions, respectively because they are more suitable to
model the possibility distributions derived from the natural phenomena of the parameters under
consideration. The mathematical model of Gauss MF and Triangular MF are given in Eqn. (1)
and (2), respectively, and Table 1 represents the simulation parameters:
World Scientific News 120(2) (2019) 154-167
-162-
𝑓(𝑥; 𝜎, 𝑐) = 𝑒−(𝑥−𝑐)2
2𝜎2 (1)
𝑇𝑟𝑖𝑎𝑛𝑔𝑙𝑒 (𝑥; 𝛼, 𝛽, 𝛾) = max (min (𝑥−𝛼
𝛽−𝛼,
𝛾−𝑥
𝛾−𝛽) , 0) (2)
where:
𝜎 = 𝑤𝑖𝑑𝑡ℎ of the Gaussian function, 𝑐 = 𝑐𝑒𝑛𝑡𝑒𝑟 𝑜𝑓 𝑡ℎ𝑒 𝑔𝑢𝑎𝑠𝑠𝑖𝑎𝑛 𝑐𝑢𝑟𝑣𝑒)
α, β, γ represent starting point, peak point and final point respectively which are x coordinates
where (α<β< γ).
Table 1. Simulation parameters
Parameter Configuration
FIS type mamdani
AND method min
OR method max
Implication min
Aggregation max
Defuzzification centroid
SNR range [10 35] dB
MTP [20 32] dBm
Velocity [0 130] Km/Hr
Throughput [0 100] %
Threshold [-110 -40]
Hysteresis [0 10]
4. RESULTS AND DISCUSSION
The proposed algorithm focuses on the removal of the Ping-Pong problem, corner effect
and to avoid the risk of unnecessary handoffs by optimizing the values of the threshold and
hysteresis based on the value of the input parameters. It is worthy of note that once the threshold
value for handoff is met, the effect of a mobile device spontaneously roaming between
overlapping coverage zones will be eliminated. This is largely because it will stay connected in
either a coverage zone or handoff to another coverage zone absolutely, thereby eliminating the
ping-pong effect. Further to this, the allowable signal strength difference between the serving
World Scientific News 120(2) (2019) 154-167
-163-
Node B and the monitoring Node B needed to initiate handoff, that is; the hysteresis value must
also be satisfied to absolutely mitigate this effect. The same goes for other handoff limiting
parameters. The results show inputs-Outputs variation as follows; when the SNR is 22.5 dBm,
MTP is 26 dBm, Velocity is 65 km/h and the throughput is 50%, the resultant handoff Threshold
and Hysteresis is 74.1 dBm and 4.94 dBm.
The Input-Output relation is modeled as presented in Eqn. (3):
𝑦𝑗𝑖 = ∑ ∏ ℎ𝑚𝑥𝑖𝑛𝑚=1
𝑏𝑖=1 (3)
where:
𝑏 = 𝑘𝑛 is the number of fuzzy rule generated
ℎ𝑚 is the mth row of FIS matrices of the rules in the FIS
𝑘 is the number of fuzzy membership function
(In this case, k = 3 that is; Low, Normal and High)
𝑛 is the number of inputs (in this case, n = 4)
𝑗 is the number of outputs for 1 ≤ 𝑗 ≤ 2
𝑥𝑖 is ith column of the input matrix for 1 ≤ 𝑖 ≤ 𝑏
𝑦𝑗𝑖 is the possible output response of the FIS
ℎ = [
ℎ11 ⋯ ℎ1,𝑛
⋮ ⋱ ⋮ℎ𝑏1, ⋯ ℎ𝑏𝑛
] (4)
𝑥 = [𝜃11 ⋯ 𝜃1𝑏
⋮ ⋱ ⋮𝜃𝑛1 ⋯ 𝜃𝑛𝑏
] (5)
𝑦 = [
𝑦𝑗1
:𝑦𝑗𝑏
] (6)
where:
ℎ is the matrix of the FIS rules and the matrix element represent the strength of each rule as
stated in the MF plane, x is the matrix of the input combination, and y is the output matrix
combination. It should be noted that x is a Gaussian function and y is a triangular function.
Therefore, Eqn. (3) is treated fuzzifully.
Furthermore, evaluation of the Algorithm was carried out using “eval” function by
randomly selecting the input sequence. The variation of the input parameters is as shown in Fig.
10. The resultant outputs (that is, Handoff threshold and Hysteresis) are as shown in Fig. 11.
The outcome can be seen to be highly adaptive in that variation in the input parameters causes
a corresponding change in the Handoff Threshold and Hysteresis as compared to traditional
handoff where both the Handoff threshold and hysteresis remain unchanged irrespective of the
variation in the network parameters. The combination of the four parameters in making handoff
initiation decision improves the Quality of Service as noise, speed, and throughput are put into
consideration. The use of Mobile Transmission Power also enhanced the power consumption
World Scientific News 120(2) (2019) 154-167
-164-
by mobile users which in turn optimized the overall power consumption via mobile
communication.
Figure 10. Input sequences for Performance Evaluation
Figure 11. Resultant Adaptive Handoff Threshold and Hysteresis
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 200
20
40
60
80
100
120
140
No of Sample
SN
R/M
.T.P
./V
elo
city/T
hro
ugh
pu
t
SNR (dB)
M.T.P. (dBm)
Velocity (Km/h)
Throughput (%)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20-120
-110
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
10
No of Sample
Ha
nd
ov
er T
hresh
old
/Hy
ste
resis
Handover Threshold (dBm)
Handover Hysteresis (dBm)
World Scientific News 120(2) (2019) 154-167
-165-
5. CONCLUSION
Though several works have been reported in this research direction, the application of
soft computing in this area provides a better way of handling problems associated with mobility
within a wireless mobile network. The proposed methodology provides an effective way of
handling handoff issues like Ping-Pong, handoff failure rate, and network latency as it allows
only the necessary handoff without Ping-Pong with a focus on the quality of service. More so,
the parameters under consideration provide a holistic quality of service as the link quality,
channel throughput, UE velocity is been monitored. Future work would focus on the
implementation of this algorithm in the existing wireless communication systems.
Acknowledgements
Special thanks to Covenant University and University of Lagos for the sponsorship of this research work.
References
[1] E. A. Oboyerulu, A. E. Ibhaze and O. B. Idowu-Bismark. Performance Analysis of
Wireless Mobile network. International Journal of Applied Engineering Research, vol.
11, no. 8, pp. 6058-6062, 2016.
[2] B. J. Chang, and J. F. Chen. Cross-Layer-Based Adaptive Vertical Handoff With
Predictive RSS in Heterogeneous Wireless Networks. IEEE Transactions on Vehicular
Technology, vol. 57, no. 6, pp. 3679-3692, 2008
[3] S. M. Phemina and S. Sendhilnathan. Fuzzy Based Mobility Management in 4G
Wireless Networks. International Journal of Brazilian Archives of Biology and
Technology, vol. 59, no. 2, pp. 1-13, 2016.
[4] E. Ibhaze, A. L. Imoize, S. O. Ajose, S. N. John, C. U. Ndujiuba and F. E. Idachaba. An
Empirical Propagation Model for Path Loss Prediction at 2100MHz in a Dense Urban
Environment. Indian Journal of Science and Technology, vol. 10, no. 5, pp. 1-9, 2017.
[5] S. Hussain, S. R. Zaidi, R. Dorsinville, M. A. Ali and G. Ellinas. Handoff in mobile
WiMAX: Forced handoff scheme with load balancing in mobile WiMAX networks. In
Proceedings 37th Annual IEEE Conference on Local Computer Networks - Workshops,
22-25 October 2012, Clearwater, Florida, USA, vol. 8, p. 666-672, 2012.
[6] O. B. Idowu-Bismark, A. E. Ibhaze and A. A. Atayero. MIMO Optimization
Techniques and Their Application in Maximizing Throughput for 3GPP HSPA+.
Journal of Wireless Networking and Communications, vol. 7, no. 1, pp. 1-8, 2017.
[7] N. A. Hamad. A New Proposal of Handover Algorithm between Cellular Mobile and
Mobile WiMAX Systems. International Journal of Applied Engineering Research, vol.
11, no. 24, pp. 11894-11900, 2016.
[8] M. Beach. Channel Characterisation for Advanced Wireless Access. IEEE
Loughborough Antennas & Propagation Conference (LAPC), 14-15 November 2011,
Loughborough, UK, vol. 1, p. 1-1, 2011.
World Scientific News 120(2) (2019) 154-167
-166-
[9] M. Kassar, K. Brigitte and P. Guy. An overview of vertical handover decision strategies
in heterogeneous wireless networks. Computer Communications, vol. 31, no. 10, pp.
2607-2620, 2008.
[10] M. Sauter. From GSM to LTE: an introduction to mobile networks and mobile
broadband. John Wiley & Sons, 2010.
[11] G. P. Pollini. Trends in handover design. IEEE Communications Magazine, vol. 34, no.
3, p. 82–90, 1996.
[12] N. Zhang and J. M. Holtzman. Analysis of handoff algorithms using both absolute and
relative measurements. IEEE Transaction on Vehicular Technology, vol. 45, no. 1, p.
174–179, 1996.
[13] P. Marichamy, S. Chakrabarti and S. L. Maskara. Performance evaluation of handoff
detection schemes. Proceeding of IEEE Conference on Convergence Technologies,
Asia-Pacific Region, 15-17 October 2003, Bangalore, India, vol. 2, p. 643–646, 2003.
[14] K. Pahlavan, P. Krishnamurthy, A. Hatami, M. Ylianttila, J. P. Makela, R. Pichna and J.
Vallstron. Handoff in hybrid mobile data networks. IEEE Personal Communications
Magazine, vol. 7, no. 2, p. 34–47, 2000.
[15] L. A. Zadeh. Some reflections on soft computing, granular computing and their roles in
the conception, design and utilization of information/intelligent systems. Soft
computing, vol. 2, no. 1, pp. 23-25, 1998.
[16] S. Lal and D. K. Panwar. Coverage Analysis of handoff Algorithm with Adaptive
Hysteresis Margin. In IEEE 10th International Conference on Information Technology
(ICIT 2007), 17-20 December 2007, Orissa, India, pp. 133 - 138, 2007.
[17] C. Çeken, Y. Serhan and A. Huseyin. Interference aware vertical Handoff Decision
Algorithm for QoS support in Wireless Heterogeneous Networks. Journal of Computer
Networks, vol. 54, no. 5, pp. 726-740, 2010.
[18] U. Mir and Z. A. Bhatti. Time triggered handoff schemes in cognitive radio networks: A
survey. Journal of Network and Computer Applications, vol. 102, pp. 71-85, 2018.
[19] M. Stemm and R. Katz. Vertical handoff in wireless overlay networks. ACM Journal on
Mobile Networks and Applications, vol. 3, no. 4, pp. 335-350, 1998.
[20] Y. Min-Hua, L. Yu, and Z. Hui-min. The mobile IP handoff between hybrid networks.
The 13th IEEE International Symposium on Personal, Indoor and Mobile Radio
Communications, 18-18 September 2002, Pavilhao Altantico, Lisboa, Portugal, vol. 1,
pp. 265-269, 2002.
[21] P. M. L. Chan, R. E. Sheriff, Y. F. Hu, P. Conforto and C. Tocci. Mobility management
incorporating fuzzy logic for a heterogeneous IP environment. IEEE Communication
Magazine, vol. 39, no. 12, pp. 42-51, 2001.
[22] I. Val, A. Etxabe, R. Torrego, P.M. Rodriguez, C. Cruces, V. Diez, M. Mendicute and
A. Arriola. Design, analysis and implementation of a time-bounded spectrum handoff
algorithm for real-time industrial wireless sensor and actuator networks. Journal of
Network and Computer Applications, vol. 125, pp. 1-16, 2019.
World Scientific News 120(2) (2019) 154-167
-167-
[23] A.M. Aibinu, A.J. Onumanyi, A.P. Adedigba, M. Ipinyomi, T.A. Folorunso and M.J.E.
Salami. Development of hybrid artificial intelligent based handover decision algorithm.
Engineering Science and Technology, an International Journal, vol. 20, no. 2, pp. 381-
390, 2017.
[24] A. Çalhan and C. Çeken. Speed sensitive-energy aware adaptive fuzzy logic based
vertical handoff decision algorithm. In IEEE 18th International Conference on Systems,
Signals and Image Processing (IWSSIP), 16-18 June 2011, Sarajevo, Bosnia-
Herzegovina, vol. 18, pp. 1-4, 2011.
[25] A. Jain and S. Tokekar. Application Based Vertical Handoff Decision in Heterogeneous
Network. In Elsevier 3rd International Conference on Recent Trends in Computing,
Procedia Computer Science 57, pp. 782-788, 2015