See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/319092802

Airport Gate Assignment for Improving Terminals' Internal Gate Efficiency

Article in The International Journal of Industrial Engineering: Theory, Applications and Practice · December 2016

CITATIONS READS
6 2,911

5 authors, including:

Hyeonu Im Chulung Lee

Korea University Korea University
7 PUBLICATIONS 35 CITATIONS 107 PUBLICATIONS 1,275 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Hyeonu Im on 26 July 2019.

The user has requested enhancement of the downloaded file.

International Journal of Industrial Engineering, 23(6), 431-444, 2016

AIRPORT GATE ASSIGNMENT FOR IMPROVING TERMINALS’ INTERNAL

GATE EFFICIENCY
Jaehwan Lee1, Hyeonu Im1, Ki Hong Kim1, Sha Xi2, and Chulung Lee3,*
1
Department of Industrial Management Engineering, Korea University
Seoul, Korea
2
School of Economics and Management, Tianjin Chengjian University
Tianjin, China
3
Division of Industrial Management Engineering, Korea University
Seoul, Korea
*
Corresponding author’s e-mail: [email protected]

This paper considers airport gate assignment (AGA) to evenly distribute passengers to airport internal gates (landside gate),
where critical processes such as security check, immigration, and customs are performed. The paper focuses on AGA and the
efficiency of landside gates. The AGA determines departure passengers’ entering gates and arrival passengers’ initial location
in an airport. The initial location may settle passengers’ landside gate selection depending on their moving distances, which
make congestion or idling on landside gate. To improve this inefficiency, we tried to adapt and modify AGA model by using
quadratic mixed integer programming to assign gates with balanced passenger flow on each gate. We then performed a
simulation experiment to verify the effect of proposed model on the internal gate efficiency. The result shows that the
proposed model reduces the passenger processing time.

Keywords: airport; gate assignment; line balancing; utilization; mixed integer programming

(Received on November 30, 2013; Accepted on December 10, 2016)

1. INTRODUCTION

After the deregulation, airlines employed a hub-and-spoke system to meet the significantly growing market demand. A
hub-and-spoke system in fact does improve airport capacity utilization and increase carriers` frequencies between origins and
destinations at lower costs. Such transformation, however, increases congestion in landside operations since the
non-expandable facilities cannot meet the increased number of the passengers.
At a major hub airport, such rapid growth could lead to airport congestion, particularly for the landside gates. In any
peak time (bank), a large number of flights arrives at the airport, allowing passengers to transfer to another flight, to complete
their arrival procedure, and to start boarding in a short period of time. A great number of passengers simultaneously landing,
transferring, and boarding challenges the airport operation system, especially for the landside gate processes. Thus, even
without the expansion of airport facilities, the landside process must still be more efficient than that of general airport gate
assignment (AGA).
Furthermore, some major airports, designed and equipped with sufficient facilities, fail to handle passenger flows in an
efficient manner. Figure 1, for example, shows unbalanced workload on landside gate. Most of the managers worry about
handling the passengers during the peak time, and managers try to reduce process time from security check, customs,
immigration, etc. To solve the problem in integrated manner, however, the unbalanced usage of landside gate must be
equalized. Because, when the unbalanced usage of landside gates becomes balanced, the throughput time of passengers can
be reduced. This paper considers an AGA model to equalize the passengers’ biased usage of landside gates and verifies that an
AGA model can increase both landside gate efficiency and throughput time of passengers. To solve the problem by balancing
usage of landside gates with AGA, we reviewed previous literatures that are related to improving airport operations efficiency.
The AGA for airport operation can be divided into two segments. The first one is airside oriented, and the other one is
landside oriented. Both of the classifications obtain a great interest from many researchers (Ding et al. 2004).

ISSN 1943-670X INTERNATIONAL JOURNAL OF INDUSTRIAL ENGINEERING

Lee et al. Airport Gate Assignment

Figure 1. The Number of Passengers in Each Landside Gate and Time

Regarding the airside oriented AGA, Lee and Steinberg (1972) studied AGA to get feasible solutions with chip selection
in a reasonable amount of time. However, the AGA cannot be handled with heuristic approach. Ohtsuki et al. (1979) solved
AGA problem with one-dimensional logic and graph theory, and this paper became a classic work for today’s scholars. Fujii
et al. (1987) studied taxi way layout to decrease conflicts, which occur in gate approaching within AGA schedule. However,
the limitation of this work is that they considered horizontal tracks only. Vanderstraeten and Bergeron (1988) and Drexl and
Nikulin (2008) studied AGA to minimize the number of idle gates, but the increased traffic volume gives only a few idle gates
in a hub airport. Wirasinghe and Bandara (1990) and Cheng (1997) solved AGA problem for minimized airport flight fee,
which is strongly related to revenue and has influence on ticket price. In addition, airport fee has strong effect on airport
management strategy. Bolat (2000, 2001), Lim et al. (2005), and Zhu et al. (2003) researched about minimizing the variance
of idle time of gates and flights. These studies are aimed at achieving robustness in operation scheduling. Unfortunately, the
robust schedules are slack in time utilization, and airport managers do not wish to use these strategies. Şeker and Noyan (2012)
generalized AGA model that has been studied in previous researches. Their work is valuable in studying AGA and provides a
way to translate AGA problem from deterministic to stochastic model. But empirical study was not carried out in their paper.
Regarding the landside oriented AGA, Bihr (1990) and Haghani and Chen (1998) looked for minimizing total passenger
walking distance with binary integer programming, and Pintea et al. (2008) and Diepen et al. (2009) studied in minimizing
passenger connection. Based on these approaches, passengers feel more convenient and satisfied, but the airport loses
operational efficiency on overall perspective, making application of their results limited in real operation. To solve the
limitations, Wei and Liu (2009) studied in multi objective AGA. Their objects are minimizing passenger walking distance and
maximizing schedule robustness at the same time. However, the maximized robust schedule wrecks efficiency. Yamada et al.
(1989) researched about optimal permutation of landside gates to minimize the chip area. It is a form of landside gate
assignment for revenue management, and recently, it attracts a lot of attention from airport managers.
In this paper, to increase airport internal efficiency, we designed a quadratic mixed integer programming to determine
gates assignment for a hub airport. We used discrete time simulation with real data to give empirical verifications.
The remainder of this paper is organized as follows. Objective of this paper is described in section 2. A mathematical
model is presented in section 3. Section 4 shows a real-life example as a simulation experiment, which is designed based on a
major airport to verify proposed model. Finally, our conclusion is presented in section 5.

2. PROBLEM DESCRIPTION

In this paper, we have to consider two things: the AGA (flight gate assignment) and passenger processes that occur in landside
gates.
The AGA assigns a flight to a gate in 1:1 match. The assignment is divided into two different situations shown below in
Figure 2. After landing, a flight is assigned to a gate for deplaning passengers. Then, if necessary, the flight moves into
another gate for boarding passengers. Between the first gate assignment and the second gate assignment, there are
maintenances such as refueling, catering, etc. As a result, the AGA must consider two situations, departure and arrival, at the
same time.

432
Lee et al. Airport Gate Assignment

Flight 1 : Gate assignment for Flight 1 : Turn around for the next Flight 1 : Gate assignment for
passenger deplaning flight passenger boarding
Flight 2 : Gate assignment for
boarding or deplaning

Figure 2. Concept of Gate Assignment

Figure 3 shows the passenger processes, which occur in landside gates. Before boarding for departure, passengers need
to pass through procedures like customs declaration, security check, and immigration. After landing, passengers proceed
through quarantine, immigration, and customs inspection for arrival, or security check for transfer. Passengers must pass
these processes before leaving the airport terminal, boarding a flight, or transferring.

Figure 3. Flow Chart of Passenger Processes

However, AGA and landside gate process are strongly related to each other. Buckelin’s work (1971) shows passengers
consider the distance between their current location and destination when moving in an airport. This is the reason why the
usage of airport landside gates is biased, which makes processes inefficient (Choi and Lee, 2012). Thus, we need to equalize
the biased usage among these processes to increase average utilization of airport facilities and solve the congestion problem.
The properties of the problem, which is described above, can be recognized as a manufacturing system with parallel
processes. The biased usage of landside gate can be seen as unbalanced input (passenger) flow, and landside gates are similar
to workstations in a manufacturing system. The utility of some parallel processes can be increased in restricted situations
where the input is balanced. This area is studied by Sarin et al. (1999) and Corrado and Mattey (1997). The passenger flows of
each process depend on flight scheduling and passenger behavior. Generally, passengers are inclined to minimize their
walking distance and reduce waiting time. These tendencies lead to intensive congestion at landside gates that are relatively
close to each other. Thus, designing a model that disperses or controls the passenger flow equally is an effective way of
solving the problem.
In the proposing modified AGA model, passenger flows can be more equally allocated to appropriate landside gates to
prevent congestion.

433
Lee et al. Airport Gate Assignment

3. MATHEMATICAL MODEL

The AGA model is well-described in Şeker and Noyan (2012). They explained that the objective function is specified
according to the decision maker’s preferences including minimizing the total passenger walking distance, non-occupied time
during the operation period, the variance of idle time, etc. They divided constraints into two parts. One is hard constraints,
which force assigning, and the other one is soft constraints, which represent gate type and operational rules that differ from
airport to airport. This general AGA model solves real time AGA problem by using integer programming (IP) model.
Passenger flows at landside gates, however, are not considered in existing research.

3.1 Modified AGA Model

In this paper, we propose an AGA model, which can control passenger flow to provide landside gates equal number of
passengers. The proposed model minimizes the difference of number of passengers on each landside gates.

Figure 4. Concept of Proposed Model

Figure 4 shows the concept of the proposed model. Flight schedule tells the time the flight must be assigned and
detached from a gate, which ensures timely assignment of the gates. We can calculate the number of passengers that should be
handled on each gate by using the number of passengers and the rate of transfer passengers. The important point is that there
are uncertainties in the passenger behavior. Thus, stochastic calculation is needed.
Furthermore, to design mathematical model for AGA, we set some assumptions as in Table 1 to satisfy
airport-dependent characteristics and to convert continuous operations to discrete planning.

Table 1. Assumptions in Designing Mathematical AGA Model

Flight Schedule Airport cannot change the flight schedule, which is already set by airline.
Unit Time Unit time is discretized in 5 min.
Flight Delay Does not consider the flight delay.

The passengers decide to use the nearest landside gate by following the retail gravity model
(Buckelin, 1971).
Retail Gravity Model :
Decision of Passengers’ D 
Paw  n aw w  w  1, , n
Landside Gate 
 aw
D
a 1
The customer(passenger)’s decision follows distance between current location and
potential market location , and the proportion is determined  .

Transfer Flight The transfer flight is the earliest one of the flights that is scheduled 1 hour after the arrival.

434
Lee et al. Airport Gate Assignment

The restriction about the flight type and gate type.

Airport Soft Constraints
The restriction about adjacent parking.
Landside Gate Operations Landside gate serves passenger proper time.

Passenger Traveling Speed The mean values of the distribution of passenger’s walking speed.

The definitions of notations are shown as follows:

i : Index of flight, iI

I : Set of flights
j : Index of gate, j  J
Ji : Subset of gates that is assignable for flight i , Ji  J
J : Set of gates
k : Index of internal-gate, k  K
l : Subset index of landside gates where l  1 is arrival landside gates, l2 is
departure landside gates, l  3 is transfer landside gates
Kl : Set of landside gate l  1, 2,3
t : Index of time (unit time), t T
T : Time interval
Sit : Flight schedule of flight i at time t
Arrt : Set of arrival flight at time t
Dept : Set of departure flight at time t
T jk : Traveling time of passenger from gate j to internal-gate k
A : Set of adjacent gates, A  {{1, 2},{1, 2,3}, ,{ j  1, j, j  1}, }
Ni : The number of passengers of flight i
Pr jk : Probability of passenger flow between gate j and internal-gate k
PTi : The proportion of transfer passengers from arriving flight i to other flight
PDi : The proportion of transfer passengers to departing flight i from other flight
Glkt : The number of passengers who use k internal-gate to pass customs at time t to go to
element of l gate set
xijt : Decision variables, if flight i is assigned to gate j at time t , then 1, otherwise 0

yij : Dummy decision variables, if flight i is assigned to gate j , then 1, otherwise 0

(This variable prevents overlapping the aircraft that is already assigned.)
M : A big number

The proposed model can be appreciated as a variation of AGA. This proposed model is composed of quadratic objective
function, which can be divided into landside gate usage equalizing function part and subsidiarity function part that makes
AGA solution feasible. Proposed mathematical model formulation is as follows.

435
Lee et al. Airport Gate Assignment

Minimize   (Gl kt )  M  yij

2 2
kK l ij
(1)

Subject to

xijt {0,1} i, j, t (2)

yij {0,1} i, j (3)

 xijt  yij i, j (4)
t
 xijt  1 i, j  J i , t  Sit (5)
j

 xijt  1 j  ji , t (6)
i
 xijt  1 t , A  {{1, 2},{1, 2,3}, ,{ j  1, j, j  1}, } (7)
i , jA

 (1  PTi ) Ni Pr jk xij (t T jk )  G1kT t  T , k  K1 (8)

iArrt , j

 (1  PDi ) Ni Pr jk xij (t T jk )  G2kT t  T , k  K2 (9)

iDept , j

 PTi Ni Pr jk xij (t T jk )  G3kT t  T , k  K3 (10)

iArrt , j

We employ the Lagrange Multipliers method to build the objective function (1), which makes the passenger flows of
landside gates balanced (refer George et al., 2010, Lin, 2009, Lin, 2003, and Boonprasurt and Nanthavanij, 2012). The
summation term has no effect on landside gate type because constraints (8), (9), and (10) control the number of passengers.
The transfer passengers are calculated based on arrival flight schedule. At the same time, the squared M depending on sum of
yij guarantees consecutive assignment of a flight to a gate. Constraints (2) and (3) are the decision variable constraints. The
decision variables have binary value, which represents whether decision variables are assigned or not. Constraint (4) works
with objective function to continuously assign during scheduled time without any change of gate. Constraints (5) and (6)
logically guarantee that one flight is assigned to one gate, and they partially contain soft constraints for airport gate occupancy
rule depending on flight type and airline. Constraint (7) is a soft constraint, which is set by assumption that prohibits the
adjacent gates from being assigned simultaneously. Constraints (8), (9), and (10) logically generate the passenger flow, which
heads to landside gates with some probability. The number of passengers that flows into internal arrival gate is the difference
between the number of passengers from scheduled arrival flight and transfer passengers, and the internal departure gates are
similar to the gate assignment. The internal transfer gate should serve partial passenger flows from arrival flight schedule.

4. REAL-LIFE EXAMPLES

4.1 Simulation Design

We designed a simulation to verify the feasibility and effectiveness of the model. The simulation is designed to compare the
performance between the real AGA operations and the newly proposed AGA operations, which is driven by proposed AGA
model. The simulation is designed as follows.

 Designed as a discrete time event simulation.

 The performance measures of simulation are cycle time per individual passenger, queue length, and equalized level.
The equalized level shows the differences existing on each gate. It is calculated with the following equation (11).

Max ( k ) 1
 (Glkt  Gl ( k 1)t ) 2
t , k 1
l (11)
1
t
 The simulation module structure is shown in Figure 6, and internal operation flow is shown in Figure 7.
436
Lee et al. Airport Gate Assignment

 The simulation time interval is set during 15:00 to 18:00, the time period where one of the banks occurs. Actual
performance is measured from 16:00 to 18:00.
 The input data of flight schedule and the number of passengers are based on Saturdays in the last 37 weeks.
(Saturdays are considered to be peak periods.)
 There are maximum of four departure gates, six transfer security checks, and six arrival gates.

The first simulation model is based on a major international hub airport. The layout is drawn as a flowing figure to
calculate distances between gates and landside gates we use in a major airport layout shown in Figure 5. The distance between
airside gate and landside gate is calculated on this layout as well. The layout has tripartite structure. The first layout is used by
departure passengers. The second layout is the arrival layout, where the passengers move to finish their processes. The last
one is used for transfer passengers. Each layout is mutually exclusive except the airside gates.

< Departure Layout > < Arrival Layout > < Transfer Layout >

Figure 5. Simulation Layout

Figure 6 shows the structure of simulation, which consists of 5 individual modules: AGA module, flight schedule
module, probabilities and distribution module, controller module, and simulation module. The role of each module is as
follows.

Figure 6. Simulation Module Structure

The flight schedule module contains flight information in Table 2 as flight type, which should be considered in the
proposed AGA model. The number of passengers is not only arrival and departure passengers, but also the rate of the transfer
passengers. The gate assign time window comes from historical data, and the reason proposed model needs that information is
included in the flight schedule module.

437
Lee et al. Airport Gate Assignment

Table 2. Input Flight Schedule Including AGA Information (partial)

FLT STA/D ETA/D ARR/DEP ORG/DES PAK

KE913 0:05 0:04 ARR DOH 21
SQ603 0:20 0:29 ARR PUS 124
OZ6783 0:20 0:29 DEP NNG 124
QR883 0:35 0:36 ARR PVG 111
OZ6889 0:35 0:36 DEP FRA 111

FLT : Flight ARR/DEP : Arrival operation / Departure operation

STA/D : Scheduled time of arrival/departure ORG/DES : Origin/Destination
ETA/D : Entering time (time of real operation) PAK : Parking gate number (Gate assignment information)

AGA module provides two things. One is historical AGA plan, and the other one is an AGA plan from the proposed
model. The two AGA plans are used for comparing the effect of AGA plan from proposed model to real AGA operations,
which are former operations. The AGA plan from proposed model is computed with IBM ILOG Cplex library.
The controller module controls manpower schedule in landside gate to handle passenger flows with optimal service
performance (Seo et al., 2012).
The probabilities and distribution module is directly linked with simulation module. This module provides uncertainty
to events as passenger generation, passenger walking speed, and service rates. This makes up the limitation of deterministic
model that we proposed and guarantees the practical effects of the proposed model. Table 3, on the next page, shows the
random numbers generated by probabilities and distribution module, which has the following distributions.

Table 3. Input Parameters and Distributions

Operations Distribution Parameters

Arrival : (Start of Assign, End of Assign)
Passenger generation Uniform
Departure : (-60 min, Start of Assign)
Passenger walking speed Uniform (100m/min, 160m/min)
Check-in process service rate Triangular (0.15 min, 0.3 min, 0.5 min}
Immigration process service rate Triangular (0.75 min, 1 min, 1.3 min)
Customs process service rate Triangular (0.5 min, 1.3 min, 3 min)
Security check process service rate Triangular (0.5 min, 0.85 min, 1.5 min)
Shuttle train service rate Triangular (0 min, 5 min, 10 min}
Uniform (Min, Max)
Triangular (Min, Mean, Max)

Figure 7 shows simulation module. This module interacts with other modules. It generates passengers with flight
schedule and AGA module, and calculates passenger travel time form gate to landside gate for arrival, landside gate to gate
for departure, and gate to gate via landside gate for transfer. Each process has queue and service rate, which receives inputs
from probabilities and distribution module. The performance measures, which we have set, are calculated in this module. (Lee
et al., 2003, Johnson, 2012)

438
Lee et al. Airport Gate Assignment

Figure 7. Simulation Structure

4.2 Simulation Model Verification and Validation

To verify the simulation model, we confirmed that optimization model shows feasible solutions, and simulation model gives
proper values without unusual value for the process as follows.

Figure 8. Simulation Model Verification

To validate the simulation model, we tested the null hypothesis with real data(A0) and simulation results(A1) using a
t-test. The test result shows that the null hypothesis is accepted.

Table 4. Simulation Model Validation

A1 A0
Average 47.30658 47.37854
Variation 337.966 351.3256
# of Observation 12 12
Pearson Val. 0.997012
Hypothesis. Diff. Avg. 1
Degree of Freedom 11
t Stat -2.51011
P(T<=t) One-tailed 0.01449
t Rejct One-Tailed 1.795885
P(T<=t) Two-tailed 0.028981
t Rejct Two-tailed 2.200985

439
Lee et al. Airport Gate Assignment

4.3 Simulation Result and Analysis

The simulation result shows a few interesting things. Figure 9 shows the cycle time of historical real AGA and proposed AGA
situation. At first, the average cycle time is decreased by 24% in arrival process with similar standard deviation. It means
arrival process is improved with similar fairness compared to the current AGA operations. In transfer process, the average is
increased by 15%, and it shows that proposed model makes transfer process inefficient. But the maximum cycle time is below
1 hour, and it can be regarded reasonable with our assumptions. The decreased departure cycle time can be considered that
proposed model increases the efficiency.

Figure 9. Cycle Time Per Passenger

Figure 10 shows queue length, attesting that the proposed model increases internal efficiency with AGA, which
equalizes the biased usage of landside gate and reduces the waiting time on queue. The problem, however, is that reducing
queue length, which means shorter waiting time on line, cannot guarantee decreasing transfer passengers’ cycle time. To
equalize passenger flows heading landside gate, this proposed model assigns gate to detour passengers.

Figure 10. Queue Length of Landside Gate

The proposed model equalizes the passenger flow effectively. Figure 11 shows that the figures are halved in arrival
landside gates and decreased over 76% in transfer landside gates. The departure process, however, is not easy to equalize due
to the assumptions made in departure passenger generation.

440
Lee et al. Airport Gate Assignment

Figure 11. Equalized Level

The number of passengers is needed to see the overall effect. Figure 12 shows the average passenger numbers, during
the 37 weeks, which are used in simulation model. The percentage of transfer passengers is around 15, and departure
passengers are more than that of arrival passengers. The biased number is caused by passenger dispersion, described in Figure
1. The arrival is evenly distributed compared to departure.

Figure 12. Average Number of Passengers During Simulation Period

The overall cycle time, described in Figure 13, shows that proposed model increases the efficiency of landside gates
operations.

Figure 13. Overall Cycle Time

The sum of overall passenger cycle time goes down more than 11%, and it represents that the increase in cycle time for
transfer passenger does not have an effect. In other words, the increase in transfer passenger cycle time is overwhelmed by
decrease in arrival and departure cycle times.
The simulation was utilized for Visual Basic, Excel VBA, and IBM Cplex. We repeated each experiment 30 times to
calculate a reliable average value of the AGA results. The experiments, which were performed on the Intel Core i3 2GB
environment, spent 1,491 minutes.

441
Lee et al. Airport Gate Assignment

Table 5. Benchmarking with Alternative Approaches

Reference Objective Methodology Limitation

The performance of the new algorithm may
Yamada et Optimize permutation of landside Heuristic become the same as that of the greedy
al. (1989) gates to minimize the chip area algorithm algorithm, since the original problem cannot
be hierarchically partitioned.
Bihr (1990)
Minimize total passenger walking Binary integer The approached model results frequent
Haghani and distance programming disruption of gate assignments
Chen (1998)
Hybrid ant system
Pintea et al.
(combinatorial The airport loses operational efficiency in
(2008)
Minimize passenger connection time optimization) overall perspective compared to increase in
Diepen et al. Integer linear passenger satisfaction
(2009) programming
Minimize passenger walking
Wei and Liu Maximizing the schedule robustness results in
distance / Maximize schedule Fuzzy model
(2009) loss of operational efficiency.
robustness

Previous research results do not satisfy both operational efficiency and schedule robustness at the same time.
Benchmarking with alternative approaches shown in Table 5, the objective of the proposal model is to maximize both
landside gate efficiency and throughput time of passengers by using quadratic mixed integer programming. By equalizing the
passengers’ biased usage of landside gates, the simulation result showed that the proposed model reduces the passenger
processing time and increases landside gate efficiency by reducing the waiting time on queue.

5. CONCLUSION

In this paper, we proposed an AGA model to increase landside gate efficiency by equalizing biased usage. We designed AGA
mathematical model with quadratic objective function based on Lagrange Multipliers method. In real-life example, it is
verified that, through this simulation, the proposed model is feasible and has effectiveness in real operation. The result
contains potential chance even though some part of the result shows that the model improves all internal operations. For
example, in the transfer process case, increased cycle time and decreased queue length could mean the chance of revenues
from transfer passengers in the airport revenue management aspect. The decreased queue length on departure process shows
that the proposed model slightly controls departure process. Furthermore, the slight change in departure process with
decreased equalize level represents that some facility extensions are needed.
The contributions provided in this paper are as follows. First, we achieved the passenger flow control with AGA and
validated it. This concept looks most useful to service industry where customer flows exist. Second, the results have academic
value in convergence study of customer behavior and industrial engineering. Third, the proposed model can be applied for
facility evaluation. The proposed model makes system extremely efficient. Thus, it can be a decision support tool for deciding
whether a system needs extension or not.
For further research, this paper can be extended in several ways. First, we can develop a robust rule to make passenger
flows balanced. Second, facility allocation and layout in airport terminals for balancing passenger flow can be considered.
Lastly, extending the proposed model into a cover-all model including aircraft surface control will be a priceless attempt in
improving the entire airport operations.

ACKNOWLEDGEMENT

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF)
funded by the Ministry of Education (2015R1D1A1A09058818).

442
Lee et al. Airport Gate Assignment

REFERENCES

Bihr, R.A. (1990). A conceptual solution to the aircraft gate assignment problem using 0, 1 linear programming, Computers &
Industrial Engineering, 19(1-4), 280-284.

Bolat, A. (2000). Procedures for providing robust gate assignments for arriving aircrafts, European Journal of Operational
Research, 120(1), Pages 63-80.

Bolat, A. (2001). Models and a Genetic Algorithm for Static Aircraft-Gate Assignment Problem, The Journal of the
Operational Research Society, 52(10), 1107-1120.

Boonprasurt, P. and Nanthavanij, S. (2012). Optimal fleet size, delivery routes, and workforce assignments for the vehicle
routing problem with manual materials handling. International Journal of Industrial Engineering: Theory, Applications and
Practice, 19(6), 252-263.

Bucklin, L.P. (1971). Retail gravity models and consumer choice: a theoretical and empirical critique. Economic Geography,
47(4), 489-497.

Cheng, Y. (1997). A knowledge-based airport gate assignment system integrated with mathematical programming,
Computers & Industrial Engineering, 32(4), 837-852.

Choi, S.G. and Lee, C. (2012). A study for improving passenger service level at the airport security checks by using
simulation. Journal of the Korea Society of Computer and Information, accepted.

Corrado, C. and Mattey, J. (1997). Capacity Utilization. The Journal of Economic Perspectives. 11: 151-167.

Diepen, G., Akker, J.M., Hoogeveen, J.A. (2009). Integrated Gate and Bus Assignment at Amsterdam Airport Schiphol.
Lecture Notes in Computer Science, 338-353.

Ding, H., Lim, A., Rodrigues, B. and Zhu, Y. (2004). Aircraft and gate scheduling optimization at airports. Proceedings of the
37th Annual Hawaii International Conference, System Sciences pp. 5-8.

Drexl, A. and Nikulin, Y. (2008). Multicriteria airport gate assignment and Pareto simulated annealing, IIE Transactions,
40(4).

Fujii, T., Horikawa, H., Kikuno, T. and Yoshida, N. (1987). A Heuristic Algorithm for Gate Assignment in One-Dimensional
Array Approach, Computer-Aided Design of Integrated Circuits and Systems, (6), 159-164.

George, B.T.Jr, Maurice, D.W. and Joel, .H. (2010). Thomas’ Calculus. Pearson, pp. 411-424.

Haghani, A. and Chen, M.-C. (1998) Optimizing gate assignments at airport terminals. Transportation Research Part A:
Policy and Practice, 32: 437-454.

Johnson, A. (2012). An Investigation of internal logistics of a lean bus assembly system via simulation: a case study.
International Journal of Industrial Engineering: Theory, Applications and Practice, 19(1).

Lee, L.H., Huang, H.C., Lee, C., Chew, E.P., Jaruphongsa, W., Yong, Y.Y., Ling, Y.Y., Liang Z., Leong, C.H., Tan, Y. P.,
Namburi, K., Johnson E. and Banks, J. (2003). Simulation of airports/aviation systems: discrete event simulation model for
airline operations: SIMAIR. In Proceedings of the 35th conference on Winter simulation: driving innovation, pp. 1656-1662.

Lee, M. and Leon, S. (1972). Techniques of gate assignment, Proceeding in DAC '72 Proceedings of the 9th Design
Automation Workshop, Association for computing machinery, pp. 63-71.

Lim, A., Rodrigues, B. and Zhu, Y. (2005). Airport Gate Scheduling with Time Windows, Artificial Intelligence Review,
24(1), 5-31.
443
 Lee et al. Airport Gate Assignment

Lin, Y.K. (2003). Flow assignment of a stochastic flow network with multiple node pairs. International Journal of Industrial
Engineering: Theory, Applications and Practice, 10(2), 167-174.

Lin, Y.K. (2009). A MP-based algorithm for a multicommodity stochastic-flow network with capacity weights. International
Journal of Industrial Engineering: Theory, Applications and Practice, 16(4).

Ohtsuki, T., Mori, H., Kuh, E.S., Kashiwabara, T. and Fujisawa, T. (1979). One-dimensional logic gate assignment and
interval graphs, Circuits and Systems, 26: 675-684.

Pintea, C.-M., Pop, P.C., Chira, C. and Dumitrescu, D. (2008). A Hybrid Ant-Based System for Gate Assignment Problem,
proceedings in: Third International Workshop, HAIS 2008, 273-280.

Sarin, S.C., Erel, E. and Dar-El, E.M. (1999). A methodology for solving single-model, stochastic assembly line balancing
problem, Omega, 27: 525-535.

Şeker, M. and Noyan, N. (2012). Stochastic optimization models for the airport gate assignment problem. Transportation
Research Part E: Logistics and Transportation Review, 48: 438-459.

Seo, S., Choi, S. and Lee, C. (2012), Security Manpower Scheduling for Smart Airports. Lecture Notes in Electrical
Engineering, 114: 519-527.

Varberg, D., Tec L. and Purcell E.J. (1997). Calculus with Analytic Geometry: Student Solutions Manual. Allyn & Bacon.

Vanderstraeten, B. and Bergeron, M. (1988). Automatic assignment of aircraft to gates at a terminal, Computers & Industrial
Engineering, 14(1), 15-25.

Wirasinghe, S.C. and Bandara, S. (1990). Airport gate position estimation for minimum total costs—Approximate closed
form solution, Transportation Research Part B: Methodological, 24(4), 287-297.

Wei, D. and Liu, C. (2009). Fuzzy model and optimization for airport gate assignment problem, Intelligent Computing and
Intelligent Systems, 2, 828-832.

Yamada, S., Okude, H. and Kasai, T. (1989). A hierarchical algorithm for one-dimensional gate assignment based on
contraction of nets, Computer-Aided Design of Integrated Circuits and Systems, 8: 622-629.

Zhu, Y., Lim, A. and Rodrigues, B. (2003). Aircraft and gate scheduling with time windows, Proceeding in : Tools with
Artificial Intelligence, 2003, 189-193.

444

View publication stats