distributed cell configuration strategy for macro/micro overlaid ...€¦ · the hot spot is...
TRANSCRIPT
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015), pp. 177-188
http://dx.doi.org/10.14257/ijseia.2015.9.9.15
ISSN: 1738-9984 IJSEIA
Copyright ⓒ 2015 SERSC
Distributed Cell Configuration Strategy for Macro/Micro
Overlaid Hierarchical Cellular Systems
Eunsung Oh
Department of Electronics Engineering, Hanseo University,
Chungcheongnam-do, Korea, 360-706
Abstract
This paper presents cell configuration strategies for orthogonal frequency division
multiple access (OFDMA) based hierarchical cellular systems (HCS). We consider two
cases: the symmetric case that HCS is used to extend the data rate coverage and the
asymmetric case that HCS is applied to cover the hot spot. We formulate the cell
configuration strategy problem with spectral reuse planning and channel allocation
constraints, and suggest solutions for each case. The numerical results show that HCS
can improve the system stability in the symmetric case and enhance the system
performance about two times rather than that of the conventional cellular system, when
the hot spot is created at the macro-cell edge in the asymmetric case. In addition, we
describe that the co-channel interference from macro-cell base stations (BSs) are the
dominant factor of system performance, but that from microcell, BSs can be negligible.
Keywords: Cell configuration, radio resource allocation, hierarchical cellular system,
orthogonal frequency division multiple access, co-channel interference, quality of service
1. Introduction
Because of the increasing the mobile multimedia service e.g., gaming, video streaming,
and high-speed Internet access, the required data rate has been rapidly growing. However,
the macro-cell data rate coverage is less economically viable with an increasing demand
of high data rate services. To reduce this problem, the cellular structure of micro-cells
work overlaying existing macro-cells, called the overlaid cellular systems or hierarchical
cellular systems (HCS) [1, 2]. In HCS, two or more hierarchical cells are considered.
They consist of pico-cells to serve indoor user equipments (UEs), and micro-cells to
provide services to indoor or outdoor UEs. Both are overlaid by macro-cells. Small cells
provide greater spectral reuse and larger capacity, and allow the use of low power. They
also extend the data rate coverage that larger cells cannot serve. Large cells are used to
cover larger areas with low-cost implementation and to provide overflow groups of
channels for clusters of small cells when heavily loaded.
For enhancing the performance of HCS, the efficient cell configuration strategy is
required. The cell configuration algorithms have been proposed based on load balancing
[3–5]. In [3], it is introduced that load balancing can improve the system capacity in
CDMA based HCS systems. Jeong et al., presented a load sharing strategy controlled by
the serving cell selection [4]. The serving cell is selected based on measurement of the
channel quality and the load condition. It is shown that the macro-cell backs up the micro-
cell well in unusual heavy traffic load condition. Kwon and Cho proposed the resource
management algorithm in CDMA systems [5]. The proposed algorithm solves the
resource shortage in micro-cells by increasing the resource usage, and the resource
shortage is resolved by decreasing traffic in macro-cells. The computer simulation shows
that HCS systems can improve utilization of resource compared with conventional
systems.
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
178 Copyright ⓒ 2015 SERSC
Also, the frequency planning in HCS systems has been researched [6]–[8]. It presented
a frequency planning between macro-cells and micro-cells in [6]. They compared the per-
cell capacity to share the frequency, and showed that the capacity of the frequency sharing
system is poor because of the large amount of the co-channel interference (CCI) between
the macro-cell and the micro-cell. Similar results are shown in [7]. Kim et al. showed that
the cell capacity is maximized by co-locating macro-cell and micro-cell sites with
different frequencies in CDMA systems [7]. However, if the interference avoidance
technique is considered, the frequency split does not maximize the system capacity [8].
Chandrasekhar and Andrews illustrated that, considering the worst-case interference at a
corner micro-cell, interference avoidance through a time-hopped CDMA and sector
antennas allows about a 7x higher micro-cell BS density, relative to a split spectrum
network with omnidirectional micro-cell antennas. These results provide guidelines for
the design of robust shared spectrum in HCS systems.
In this paper, we establish the cell configuration strategy in HCS downlink systems.
The goal of the strategy is to maximize the total transmitted data rate. The cell
configuration strategy in HCS environments is introduced considering the uniform traffic
case only [9]. This paper considers two HCS usage models: First, HCS systems are used
for the coverage extension. In this case, it is assumed that the traffic is uniformly
distributed. We define it as the symmetric case. HCS systems are also applied to the
asymmetric case such as the hot-spot case. In the hot-spot case, micro-cells must serve
small regions of high demand traffic with in the macro-cell coverage area. To handle
various cases, orthogonal frequency division multiple access (OFDMA) is considered [10].
To solve the problem, the optimization problem for maximizing the total transmitted data
rate is formulated. From the second-order condition for convexity, it is shown that the
problem has a global optimum point. Using some assumptions, the optimal cell
configuration strategy is proposed for the total transmitted data rate maximization.
This paper is organized as follows: In Section II, system models used in this study are
explained and the optimization problem for the total transmitted data rate maximization is
formulated. Section III presents the optimal cell configuration strategies. Section IV
details the simulation results of proposed cell configuration strategies, and discusses the
performance of HCS systems. Our conclusions are summarized in Section V.
2. System Model and Problem Formulation
2.1. System Model
Figure 1. A Macro/Micro-Cell HCS System Model. This Figure Illustrates that Three Microcell BSs Subsists in One Macro-Cell BS Service Area. Micro-cell
BSs are Posited with the Equal Distance
We consider a system where its service area is divided into hexagonal macro-cells of
equal size with the base station (BS) at the center of each macro-cell, ,
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
Copyright ⓒ 2015 SERSC 179
and micro-cells are overlaid on each macro-cell service area, , as shown
in Figure 1. The 3-sector antenna for macro-cell BSs and the omni-directional antenna for
micro-cell BSs are assumed, respectively. Macro/micro-cells use same frequency
assignment (FA). In other words, the same radio channel is reused in every cell.
Subchannelized OFDMA with Nt subcarriers is used for the multiple access
schemes. The subcarrier space is divided into a number Ng of successive groups.
Each group contains Ns subcarriers. A subchannel has one element from each group
allocated, so N is the number of subchannel elements, N = Ng [11]. Subchannels are
randomly allocated to each UE. We assume partial CSI at the BS and per fect CSI at
the UE. This means that BSs only know the average channel gain of each
subchannel. The BS with the maximum signal strength, which then considers the
best serving cell selection, serves each UE.
A full buffer case is assumed for each UE. It means that the serving UEs are
always in the active state. The sets of UEs served by macro-cells and microcells are
defined as U1 and U2
, and represent the sets of UEs as for the
macro-cell and for the micro-cell, respectively. It is assumed that
the overall system is homogeneous in statistical equilibrium. In a homogeneous
system, one cell is statistically the same as any other cell. Using this observation,
the problem can be decoupled as a cell from the rest of the system, and the system
performance can be evaluated by analyzing the performance of the cell.
2.2. Problem Formulation
The goal of this work is to maximize the total transmitted data rate. The objective
function is formulated as
f Ru( ) = RuuÎU1
å + RuuÎU2
å , (1)
where Ru [bits/sec] is the transmitted data rate of a UE u. Using the Shannon
capacity, it is calculated as,
Ru
=BW
Nnulog
21+g
u
Rx( ). (2)
As shown in (2), the transmitted data rate is determined by the amount of
allocated channel resource to a UE u, nu , and the unit data rate constructed as the
transmitted signal to interference ratio (SIR) of a UE u, gu,
gu
Tx =Gubpi
ai1
Gub
1
p1
b1ÎB
1
å +ai2
Gub
2
p2
b2ÎB
2
å for uÎ U1,U2{ } and iÎ 1,2{ }, (3)
where Gub is the channel gain from BS b to UE u including the path loss and
shadowing impact, p1 and p2
are the transmitted power at macro/micro-cell.
Macro/micro-cells use the same FA; therefore the co-channel interference is
affected by both macro-cells and micro-cells. aij is the channel orthogonal factor
between cell j and cell i [6]. Because the subchannel is constructed as an interleaved
type in subchannelized OFDMA systems, the SIR of a UE u needs not be calculated
in each subcarrier [12]
We firstly consider the minimum required data rate as the quality of service
(QoS) guarantee constraint,
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
180 Copyright ⓒ 2015 SERSC
Ru
³hu for uÎ U1,U2{ }, (4)
where hu [bits/sec] is the minimum required data rate of UE u.
Constraints for the cell configuration are added. The spectral reuse planning, ri ,
is considered. The spectral reuse planning is related to the traffic load and CCI,
because aij of (3) can be calculated as ri ´ r j
when random allocation is assumed
[13]. The constraint is expressed as
0 £ ri£1 for iÎ 1,2{ }, (5)
where the lower index 1 and 2 are presented for the macro-cell and the micro-cell,
respectively. In (5), ri =1 means that all channels are used at the cell, and the cell is
shut down if ri = 0.
And, the channel allocation constraint, nu, for each user is inserted for the link
performance,
nu
uÎUi
å / N £ ri
for iÎ 1,2{ }, (6)
Using the constraints explained above, the optimization problem is formulated as
maxr ,n
Ru
uÎU1
å + Ru
uÎU2
åìíï
îï
üýï
þï
s.t. Ru
³hu, "uÎ U
1,U
2{ }
0 £ ri£1, i Î 1,2{ }
nu
uÎUi
å / N £ ri, i Î 1,2{ }
(7)
The optimization problem in (7) means that the total transmitted data rate is
maximized by the cell configuration strategy of the spectral reuse planning and the
channel allocation considering the minimum required data rate.
3. Distributed Cell Configuration Strategy
In this section, the cell configuration strategies of HCS downlink systems are
determined. We firstly reformulate the optimization problem considering the signal
model. Based on that, we propose the cell configuration strategy for the total transmitted
data rate maximization.
3.1. Symmetric Case
The first usage of HCS is to increase the transmitted data rate by enhancing the
coverage. In this case, it is assumed that UEs are uniformly distributed, which is defined
as the symmetric case, and the spectral reuse planning for macro-cell and micro-cell are
equally controlled, r1 = r2.
Based on the above assumption, the transmitted SIRs of (3) are calculated as
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
Copyright ⓒ 2015 SERSC 181
gu
1
Rx =Gubp
1
r1r
1Gub
1
p1
b1ÎB
1
å + r1r
2Gub
2
p2
b2ÎB
2
å=
1
r1
2gu
1
Tx for uÎU1
gu
2
Rx =Gubp
2
r2r
1Gub
1
p1
b1ÎB
1
å + r2r
2Gub
2
p2
b2ÎB
2
å=
1
r2
2gu
2
Tx for uÎU2
(8)
And the problem in (7) is reformulated as
maxr ,n
R̂u
uÎU1
å + R̂u
uÎU2
åìíï
îï
üýï
þï
s .t. R̂u
³hu, "uÎ U
1,U
2{ }
0 £ x £1,
x £ ri, i Î 1,2{ }
nu
uÎUi
å / N £ ri, i Î 1,2{ }
(9)
where R̂u
=BW
Nnulog
21+g
u
Tx / ri
2( ) . The object function is monotonic decreasing in ri . Thus,
the second constraint for the spectral reuse planning in (7) is modified as the second and
third constraints in (9).
In (9), the constraints satisfy the convexity, and the objective function is a concave
function, so the modified problem of (9) becomes a convex optimization problem. A
convex optimization problem can achieve a global optimized solution through Lagrangian
relaxation [14].
To solve the problem, the second and third constraints are relaxed using the Lagrangian
relaxation,
L n,r,x,l,m( ) = Ru
uÎUi
å + l 1- x( ) +m1
r1- x( ) +m
2r
2- x( )
(10)
The relaxed problem is decomposed into two parts. The first part is the problem of the
global spectral reuse planning, x,
maxx
- l +m1+m
2( ) x+ l (11)
Considering the slackness condition and the dual values, the solution of (8) is
determined as
x = min 1,r1,r
2( ) (12)
It is said that the system is not limited by CCI when x =1. However, if r1 =1 or r2 =1
then the system is restricted by CCI effected to the macro- or the micro-cell.
The problem of the second part is constructed by the residual part of (10) with the first
and fourth constraints. It is decomposed to the independent problem per each cell.
Therefore, we can do the distributed cell configure. In this paper, the way to solve the
problem of the second part for the macro-cell configuration is presented. Without loss of
generality, the micro-cell configuration can be gotten in the same way.
Similar to solving the first part, the first and fourth constraints in (7) is relaxed using
the Lagrangian relaxation,
L n,r,m,n ,w( ) = Ru
uÎU1
å +m1r
1+ n
uRu-h
u( )uÎU
1
å +w1Nr
1- n
u
uÎU1
åæ
è
çç
ö
ø
÷÷ (13)
From KKT and slackness condition, it is calculated as
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
182 Copyright ⓒ 2015 SERSC
2 12
1
1 lo g 1 0
T x
u
u
u
L B W
n N
1
1 13
1 1 1
120
lo g 2
T x
u u u
T x
u U u
nL B WN
N
1
1 10
u
u U
N n
1
0u u u
u U
R
(14)
Under the characteristic of dual variables, nu,w
1> 0 , the first equation of (14) is
expressed as
nu
=w1
N
BW
log2
log 1+gu
Tx
r1
2
æ
èçç
ö
ø÷÷
-1³ 0 w
1³BW
Nlog
21+
gu
Tx
r1
2
æ
èçç
ö
ø÷÷, "u ÎU
1
(15)
Considering the dual problem, w1 is calculated as
w1=BW
Nlog
21+
gu*
Tx
r1
2
æ
è
çç
ö
ø
÷÷ (16)
where u* = arg maxuÎU
1
gu
Tx( ) . Using (14) and (16), the channel for each UE except UE u*
is
allocated to
2 2
1
lo g 1
u
u T x
u
n
N
B W
(17)
And, the UE u*
is residually allocated to satisfying the condition,
nu
gu
Tx / r1
2
1+gu
Tx / r1
2
1
log 1+gu
Tx / r1
2( )-
1
2
æ
è
çç
ö
ø
÷÷
uÎU1
å = 0 (18)
From results, the cell configuration strategy is summarized as follows:
[The cell configuration strategy at the symmetric case]
1) At each cell, the local spectral reuse planning, ri , is calculated by (17) and (18).
2) The information of the local spectral reuse planning is exchanged, and the global
spectral reuse planning is decided from (12).
3) The channel is allocated based on the global spectral reuse planning using (17) and
(18) at each cell.
3.2. Asymmetric Case
HCS can be applied at the asymmetric case such as covering hot spots. In this case, the
micro-cell operation is restricted because the power of the micro-cell has a very low value
compared with that of the macro-cell, p2 ≪ p1. Therefore, we assume that the spectral
reuse is only planned at the macro-cell, r1, and the spectral reuse factor for the micro-cell
is fixed to one.
Under the environments, the optimization problem of (7) is reformulated as
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
Copyright ⓒ 2015 SERSC 183
maxr ,n
Ru
uÎU1
å + Ru
uÎU2
åìíï
îï
üýï
þï
s.t. Ru
³hu, "uÎ U
1,U
2{ }
0 £ r
1£1,
nu
uÎU1
å / N £ r1,
n
u
uÎU2
å / N £1
(19)
Similar to the symmetric case, the optimization problem of (19) is also the convex
optimization problem
Applying the Lagrangian relaxation, (19) can be expressed as
L nu,r
1,m
1,nu,w
i( ) = Ru
uÎ U1,U
2{ }
å + nuRu-h
u( )uÎ U
1,U
2{ }
å +m1
1- r1( )
+w1Nr
1- n
u
uÎU1
åæ
è
çç
ö
ø
÷÷+w
2N - n
u
uÎU2
åæ
è
çç
ö
ø
÷÷
(20)
From KKT and slackness conditions, the channel for each UE is allocated as
2 2
1
lo g 1
u
u T x
u
n
N
B W
for uÎU1
2 2
1
lo g 1
u
u T x
u
n
N
B W
for uÎU1 /u2
*
nu
2* = N - n
u
uÎU2/u
2*
å .
(21)
The spectral reuse planning of macro-cell is obtained to satisfying the follow condition,
nu
gu
Tx / r1
2
1+gu
Tx / r1
2
log 1+gu
1*
Tx / r1
2( )log 1+g
u
Tx / r1
2( )-
1
2
æ
è
ççç
ö
ø
÷÷÷+
uÎU1
ånu
2
gu
Tx / r1
2
1+gu
Tx / r1
2
log 1+gu
2*
Tx / r1( )
log 1+gu
Tx / r1( )uÎU 2
å = 0 (22)
Comparing to (18), equation (22) is shown that, at the asymmetric case, the spectral
reuse can be less than that at the symmetric case because of the micro-cell. At the
asymmetric case, the micro-cell is dependent on the macro-cell, thus the cell
configuration is processed at the macro-cell.
The cell configuration strategy at the asymmetric case is summarized as follows:
[The cell configuration strategy at the asymmetric case]
1) The spectral reuse planning, ri , is initialized.
2) The channel is allocated based on the spectral reuse planning using (21).
3) The spectral reuse planning is determined through (26).
4) The step 2) and 3) is repeated until the value is converged.
4. Simulation Results
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
184 Copyright ⓒ 2015 SERSC
In our simulation, we considered the subchannelized OFDMA with 1024
subcarriers. The number of subchannel sets is 30 subchannels, and each data channel
set has 24 subcarriers [15]. The cellular system being simulated consists of 19 two-
tier hexagonal cells. In order to avoid the boundary effect, the results from the
center hexagonal cell are used.
4.1. System Throughput
To fairly compare, we calculated the spectral efficiency per a macro-cell area,
i.e., one macro-cell plus micro-cells. In figures, d means the distance from a macro-
cell BS to a micro-cell BS at symmetric case and a hot spot at asymmetric case,
respectively.
2 6 10 14 18 22 26 301
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
Number of user equipment
Spe
ctra
l eff
icie
ncy
[b
its/s
ec/
Hz]
without microcells
with 5 microcells at d=200m
with 5 microcells at d=600m
with 3 microcells at d=200m
with 3 microcells at d=600m
d=600m
d=200m
Figure 2. Spectral Efficiency versus the Number of Micro-Cell and UE at the Symmetric Case with hu = 256[kbps / sec]
Figure 2 shows spectral efficiencies when the number of UE and the distance
between a macro-cell BS and a micro-cell BS are varying at symmetric case. At the
symmetric case, spectral efficiency without micro-cells is increased in low traffic
region, but is decreased in medium and high traffic regions. That is why in low
traffic cases, the system can obtain the multi-user diversity when the number of UEs
is increased, but CCI constraints the spectral efficiency in medium and high traffic
regions. However, spectral efficiencies with micro-cells are maintained in medium
and high traffic regions. This means that HCS systems can enhance the system
stability in high traffic region. Also, the spectral efficiency is improved when the
number of micro-cell is increasing and the distance between the macro-cell BS and
the micro-cell BS becomes far off. It is shown that HCS systems obtain performance
enhancement by spreading the traffic. Without loss of generality, when the micro-
cell is posited at the macro-cell edge, the enhancement is maximized in medium and
high traffic regions. Switching off BS without serving UE occurs in the low traffic
region. Therefore, the spectral efficiency has the maximum value when the
microcell is posited near the macro-cell BS.
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
Copyright ⓒ 2015 SERSC 185
Figure 3. Spectral Efficiency versus the Number of Micro-Cell and UE at the Asymmetric Case with hu = 256[kbps / sec]
Figure 3 illustrates spectral efficiencies versus the number of UEs at asymmetric
case. Hot spot is modeled that 20 UEs are randomly generated in a circular region
within 150m, and a micro-cell is posited at the center of circle. In this case, the
spectral efficiency without micro-cells has a similar value to the symmetric case, but
that with micro-cells are enhanced in all traffic regions. This means that micro-cells
support all UEs in hot-spot area, and the spectral efficiency is dependent on that in
HCS systems.
As a result, the HCS system can improve the system stability at high traffic and
high-required data rate at the symmetric case, and can dramatically enhanced the
spectral efficiency at asymmetric case.
4.2. Optimum Spectral Reuse Planning
Figure 4 shows the optimum spectral reuse planning at symmetric case. If the
minimum required data rate case is lower than the optimum, spectral reuse planning
without micro-cells is one at the most traffic region. It means that CCI does not
limit the system performance in the low requirement case. However, increasing the
traffic in the high requirement case decreases the optimum spectral reuse planning.
That is why CCI constraints the system performance. On the contrary, the effect of
traffic can be neglected by the optimum spectral reuse planning with micro-cells.
The optimum spectral reuse planning with micro-cells depends on the distance
between the macro-cell BS and the micro-cell BS. The longer distance case has a
larger optimum spectral reuse planning. It means that the amount of CCI from the
macro-cell BS is seriously affected in HCS systems. Because the optimum spectral
reuse planning has a similar value at 3 and 5 microcell cases, it is said that the effect
of CCI from the microcell BS is neglectful.
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
186 Copyright ⓒ 2015 SERSC
2 6 10 14 18 22 26 300.7
0.75
0.8
0.85
0.9
0.95
1
1.05
Number of user equipment
Op
tim
um
rh
o, r
*
without microcells
with 5 microcells at d=200m
with 5 microcells at d=600m
with 3 microcells at d=200m
with 3 microcells at d=600m
d=600m
d=200m
Figure 4. Optimum Spectral Reuse Planning versus the Number of Micro-Cell and UE at Symmetric Case with hu = 256[kbps / sec]
2 6 10 14 18 22 26 300.75
0.8
0.85
0.9
0.95
1
1.05
Number of user equipment
Op
tim
um
rh
o, r
*
without microcells at d=200m
without microcells at d=400m
without microcells at d=600m
Figure 5. Optimum Spectral Reuse Planning versus the Number of Micro-Cell and UE at the Asymmetric Case with hu = 256[kbps / sec]
At the asymmetric case, the optimum spectral reuse planning without micro-cells
is illustrated in Figure 5. That has the similar value at symmetric case. It means that
the distribution of UE’s position doesn’t affect the system performance. It is also
shown in results of the spectral efficiency. However, the optimum spectral reuse
planning with micro-cells at asymmetric case has the value of minimum spectral
reuse planning. It is said that the system performance is decided by the performance
of micro-cell.
As a result, it is said that all spectral resource must be used to obtain the optimum
performance in low requirement regions, but resource partitioning is optimum in
high requirement region without micro-cells. However, because HCS systems can
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
Copyright ⓒ 2015 SERSC 187
separate the traffic, the optimum spectral reuse planning with microcells is
dependent on CCI from macro-cell BSs at symmetric case. This has the minimum
spectral reuse planning because the microcell performance constraints the system
performance at asymmetric case.
5. Conclusions
In this paper, we proposed cell configuration strategies for OFDMA based HCS
downlink systems at symmetric case and asymmetric case. First, we formulated the
optimization problem with spectral reuse planning and channel allocation
constraints, and obtained the optimum solution using the Lagrangian relaxation
because the optimization problem is the convex problem. The proposed solutions are
only required partial CSI and are able to be distributed cell configuration. Results
showed that HCS systems can enhance the system stability at symmetric case and
improve the system performance at asymmetric case, and verified that the CCI from
macro-cells is the important parameter in HSC systems. Furthermore, we outlined
the relationship among resource parameters. Further research is needed to consider
the dynamic resource management, i.e. power control, MIMO.
Acknowledgements
This work was supported by 2014 Research Grant of Hanseo University.
References
[1] H. Claussen, L. T. W. Ho, and L. G. Samuel, “Financial analysis of a pico-cellular home network
deployment”, Proceedings of The IEEE International Conference on Communications, Glasgow,
Scotland, (2007) June 24-28.
[2] K. Son, E. Oh and B. Krishnamachari, “Energy-efficient design of heterogeneous cellular networks from
deployment to operation”, Computer Networks, vol. 78, (2015).
[3] A. Jirattitichareon, M. Hatori and K. Aizawa, “Integrated macrocell/microcell (IMM) for traffic
balancing in CDMA cellular system”, Proceedings of IEEE International Conference on Universal
Personal Communications, San Diego, USA, (1996) September 29-October 2.
[4] D. G. Jeong and W. S. Jeon, “Load sharing in hierarchical cell structure for high speed downlink packet
transmission”, Proceedings of IEEE Global Telecommunications Conference, Taipei, Taiwan, (2002)
November 17-21.
[5] T. Kwon and D.-H. Cho, “Adaptive radio resource management based on cell load in CDMA-based
hierarchical cell structure”, Proceedings of IEEE Vehicular Technology Conference, Birmingham, USA,
(2002) September 24-28.
[6] C.-L. I, L. J. Greenstein and R. D. Gitlin, “A microcell/macrocell cellular architecture for low- and high-
mobility wireless users”, IEEE Journal on Selected Areas in Communications, vol. 6, no. 11, (1993).
[7] S. Kim, D. Hong and J. Cho, “Hierarchical cell deployment for high speed data CDMA systems”,
Proceedings IEEE Wireless Communications and Networking Conference, Orlando, USA, (2002)
March 17-21.
[8] V. Chandrasekhar and J. G. Andrews, “Uplink capacity and interference avoidance for two-tier cellular
networks”, Proceedings IEEE Global Telecommunications Conference, Washington, D.C., USA, (2007)
November 26-30.
[9] E. Oh, “Cell Configuration Strategy for OFDMA Based Hierarchical Cell Structure”, Proceedings
International Workshop Mechanical Engineering 2015, Jeju Island, Korea, (2015) August 19-22.
[10] Y.-J. Choi, K. B. Lee and S. Bahk, “All-IP 4G network architecture for efficient mobility and resource
management”, IEEE Wireless Communication Magazine, vol. 2, no. 14, (2007).
[11] I. Koffman and V. Roman, “Broadband wireless access solutions based on OFDM access in IEEE
802.16”, IEEE Communications Magazine, vol. 4, no. 40, (2002).
[12] A. Ghosh, D. R. Wolter, J. G. Andrews and R. Chen, “Broadband wireless access with WiMax/802.16:
current performance benchmarks and future potential”, IEEE Communications Magazine, vol. 2, no. 43,
(2005).
[13] E. Oh, M.-G. Cho, S. Han, C. Woo and D. Hong, “Performance analysis of reuse-partitioning-based
subchannelized OFDMA uplink systems in multicell environments”, IEEE Transactions on Vehicular
Technology, vol. 5, no. 57, (2008).
[14] G. Li and H. Liu, “On the optimality of the OFDMA network”, IEEE Communication Letters, vol. 5, no.
9, (2005).
International Journal of Software Engineering and Its Applications
Vol. 9, No. 9 (2015)
188 Copyright ⓒ 2015 SERSC
[15] Part 16: Air Interface for Fixed and Mobile Broadband Wireless Access Systems Amendment for
Physical and Medium Access Control Layers for Combined Fixed and Mobile Operation in Licensed
Bands, IEEE Standard 802.16 (2006).
Author
Eunsung Oh, received his B.S., M.S. and Ph.D. degrees in
Electrical Engineering at Yonsei University, Seoul, Korea, in 2003,
2006 and 2009, respectively. From 2009 to 2011, he was a post-
doctoral researcher in the Department of Electrical Engineering at the
University of Southern California's Viterbi School of Engineering.
From 2011 to 2012, he was a senior researcher at Korea Institute of
Energy Technology Evaluation and Planning, Korea. From 2012 to
2013, he was a research professor in the Department of Electrical
Engineering at Konkuk University, Korea. He is currently an assistant
professor in the Department of Electrical and Computer Engineering
at Hanseo University, Korea. His main research interests include the
design and analysis of algorithms for green communication networks
and smart grid.