A Novel Downlink Interference Management Mechanism for Two-Tier OFDMA Femtocell Networks

Full text


Abstract—In wireless broadband access networks, almost
indoor environment encounters serious coverage problem due to
non-line of sight transmission. To improve the coverage,
femtocell has been introduced as a promising technology.
However, the deployment of femtocells may introduce extra
interference to macrocell base stations. An effective interference
management mechanism is required to optimize the system
performance. In this paper, a resource allocation algorithm is
proposed to resolve the two-tier downlink interference problem.
The proposed scheme combines fractional frequency reuse (FFR)
in macrocell and a heuristic graph-based channel assignment
algorithm in femtocell to reduce the interference problem.
Simulation results reveal that the proposed approach
significantly improves the system throughput of the two-tier
OFDMA networks.
Index Terms—Femtocells, interference management,
resource allocation, Frational Frequence Reuse (FFR).
I. INTRODUCTION
Recently, the telecommunications industry has experienced
fantastic growth in mobile Internet traffic, which is expected
to continue in the coming years. Based on the Cisco VNI, the
overall mobile data traffic will grow to 11.2 exabyte per
month by 2017 [1]. Such growth is referred to as the Mobile
Data Explosion. Since 66% of voice and 90% of data traffics
are generated indoors [2], optimizing system performance
implies improving indoor coverage and capacity. Among new
emerging wireless technologies, femtocell is a cost effective
approach for operators nowadays. A network which includes
both macro base station and femtocells is commonly referred
as heterogeneous networks (HetNet).
Femtocells consists of femto base stations (FBS) which are
small, short-range (10~30 m), and low-power cellular base
station (BS) (10~100 mW) and typically are designed for use
in a home or small business. Femtocells operate in a licensed
spectrum owned by the mobile operator and enable fixed
mobile convergence (FMC) service. FBS connects to core
network via broadband technologies such as digital subscriber
line (DSL) or cable modem [3].
Besides the benefits that femtocell can bring on for both
operators and subscribers, there are some challenges which
need be solved. Among the challenges, interference
management is a key issue for researchers and mobile
operators, especially in case of high network density in
heterogeneous networks in next few years. Due to the
Manuscript received October 1, 2014; revised March 25, 2016.
Gwo-Jong Yu and Hoai Nam Bui Khac are with the Department of
Computer Science and Information Engineering, Aletheia University, New
Taipei City, 251 Taiwan (e-mail: yugj@mail.au.edu.tw,
hoainam.bk2012@gmail.com).
interference problem in uplink and downlink is quite different,
to simply the problem, the downlink interference problem in
femtocell network is taken into account in this paper.
Orthogonal frequency division multiple access (OFDMA) is a
multicarrier transmission technology. Due to its highly
spectral efficiency and low implementation cost, it has been
adopted in many high rate wireless transmission standards
(LTE, WiMax, and WiFi). This paper takes LTE as an
example to investigate the interference problem in
OFDMA-based femtocell networks [4].
In heterogeneous networks, there are two types of
interferences. The first one is cross-tier interference which
occurs by the network elements from different tiers. For
example, the interference between macrocell evolved NodeB
(eNB) and femtocell home evolved NodeB (HeNB) is
cross-tier interference. The second one is co-tier interference
which occurs by the network elements from the same tier. For
example, the interference between two HeNB is co-tier
interference. According to the involved network elements
(sender/receiver), interference types (cross-tier
interference/co-tier interference), and transmission directions
(uplink/downlink), interference scenario in HetNet can be
summarized as shown in Fig. 1 [5].
Fig. 1. Interference scenarios of each cell in heterogeneous network.
In this paper, an effective radio resource allocation
mechanism is proposed to mitigate the interference in an
OFDMA two-tier network. Since femtocells are managed by
Femtocell Gateway (FGW) through the Internet, the FGW is
designed to support resource assignment. The main
contributions of this paper include: 1) A novel solution is
proposed to solve the downlink interference problem in
OFDMA-based femtocell network; 2) Both cross-tier and
co-tier interferences are taken into consideration to achieve
high system throughput for both macrocell and femtocell
users; 3) The computational complexity of the proposed
Gwo-Jong Yu and Hoai Nam Bui Khac
A Novel Downlink Interference Management Mechanism
for Two-Tier OFDMA Femtocell Networks
Journal
of Advances in Computer Networks, Vol. 4, No. 2, June 2016
80
doi: 10.18178/jacn.2016.4.2.208

algorithm is not too complicated, so that the computation
overhead is acceptable.
The remaining parts of this paper are organized as follows.
In section II, related works about interference management is
presented. Section III includes model problem formulation as
well as the assumptions of the interference problem. In
Section IV, the proposed algorithm FFR-GC is presented to
solve the interference problem. Then, the simulation is
performed to evaluate the performance of our algorithm in
Section V. Finally, we conclude this paper in Section VI.
II. RELATED WORK
In order to reduce the interference, improve the link
reliability, increase the capacity and enhance the system
performance of femtocell networks, different kinds of
interference management schemes have been investigated. In
[3], Chandrasekhar and Andrews provided a good survey of
femtocell technologies and provided guidelines about
interference managements. Saquib et al. [5] summarized
about the prominent issues and feasible approaches about
femtocell interference problem as well as provided qualitative
comparison among different schemes. In [6], Chandrasekhar
and Andrews proposed to use sectorial antenna and multiple
radio paths to enhance system capacity and to avoid uplink
interference. Park et al. [7] used beam subset selection
strategy and proposed an orthogonal random beam-forming
based cross-tier interference reduction scheme in
closed-access two-tier femtocell networks. Power control
approaches are proposed as promising cost-effective methods
[8]-[10].
In order to improve the resource allocation efficiency in
OFDMA systems, the concept of frequency reuse have been
introduced in [11]. In [12], Lopez-Perez et al. proposed two
novel approaches for the self-organization of OFDMA
femtocells, so that femtocells are able to dynamically sense
the air interface and tune sub-channel allocation to reduce
mutual interference. Tan et al. [13] used graph coloring
algorithm and proposed a novel dynamic sub-band allocation
technique (GC-DSA) to avoid downlink interference problem.
GC-DSA exhibit high throughput in femtocell networks.
However, this approach is not effective in case of large
network with the high density of femtocell base stations. To
deal with inefficient problem in high density networks, Zhang
et al. [14] suggested to use cognitive approach to mitigate
interference. The importance of cross-tier interference
problem has also been discussed in literature. In [15], Lee et
al. proposed to use fractional frequency reuse (FFR) in
macrocell and assign unused sub-band for femtocell to
minimize the interference between both cells. However, this
approach only focuses in mitigating cross-tier interference
between macrocell and femtocells. The co-tier interference
among femtocells may also degrade system performance in
dense networks.
III. SYSTEM MODEL AND PROBLEM FORMULATION
A. System Model
The objective of this paper proposed a novel mechanism to
reduce interference in OFDMA Femtocell networks so that
the throughput for the network can be improved. The network
model that this paper investigates is illustrated in Fig. 2.
Fig. 2. Network model of OFDMA-based femtocell.
The target network is divided into 7 macrocells in which
each macrocell was served by one eNB. A number of HeNB
are deployed within a macrocell to enhance signal quality and
system throughput. With the deployment of femtocells, if two
network elements share the same radio resource, the
interference in macrocell/femtocell pair or and
femtocell/femtocell pair will occur. The objective of this
paper is to investigate how to reduce both cross-tier
interference and co-tier interference to enhance capacity of
the network. The problem is formulated in the following.
B. Problem Formulation
Assume there are total NM macrocells in the network and
each macrocell is served by an eNB. A number of femtocells
are randomly distributed in entire network. The user
equipment (UE) which is located outdoor and served by
macrocell eNB is referred as macrocell user equipment (MUE)
while the user equipment which is located indoor and served
by femtocell HeNB is referred as femtocell user equipment
(FUE). Assume the network is deployed in urban area, then
the path loss model which described in [16] by 3GPP in
LTE-advanced standard can be used to model the signal
degradation. Let the network operates in 2GHz, then the path
loss (PL) between macrocell and MUE can be calculated as:
2
10
( ) 15.3 37.6log ( )
outdoor
PL dB d

  
(1)
In Eq. (1), d is the distance (meter) between MUE m and
macrocell M.

2 is a factor which represents the outdoor
log-normal shadowing (in dB) and is characterized by the
Gaussian distribution with zero mean and standard deviation
[17]. Similarly, the path loss between FUE f and femtocell F
can be calculated as:
10
( ) 38.46 20 log
indoor walls
PL dB d L  
(2)
The value of Lwall depends on the distance between UE and
BS (both macrocell and femtocell). The value, Lwall equals
7/10/15 (dB) if d is within (0-10 m)/(10 m -20 m)/ (20 m -30
m), respectively. So, the channel gain (G) for i-th UE can be
Journal
of Advances in Computer Networks, Vol. 4, No. 2, June 2016
81

calculated as:
10
10 PL
G

(3)
In LTE standard, resource block (RB) is the minimum unit
for radio resource allocation, the downlink signal to
interference plus noise ratio (SINR) value on RB a of MUE
m:
, , ,
,,
0 , ', , ', , , , ,
'
M a m M a
m M a
m a M a m M a f a F a m F a
MF
PG
SINR
N f P G P G


  

(4)
In Eq. (4), PM,a and PM’,a are transmit powers of serving
macrocell and neighboring macrocell set M’ on resource
block a. PF,a is the set transmit power of femtocells which
interfere to MUE m. G is channel gain which can be computed
by (3). M’ and F are sets of neighboring macrocells and
femtocells, respectively. βm,a and βf,a denote the resource
block a is assigned for MUE m and FUE f. βm(f),a= 1 if
resource block a is assigned for MUF m ( or FUE f), otherwise
βm(f),a = 0. Similarly, SINR of FUEs f is calculated:
, , ,
,,
0 , ', , ', , , , ,
'
F a m F a
f F a
f a F a f F a m a M a f M a
FM
PG
SINR
N f P G P G


  

(5)
The capacity of MUE m (or FUE f) on resource block a can
be given by (6):
( ), 2 ( ),
*log (1 )
m f a m f a
C f SINR

  
(6)
In Eq. (6), α is a constant derived from bit error rate (BER)
and can be defined α = -1.5/ln(5BER) with the setting of BER
to 10-6.
Thus, the total throughput for each macrocell M and
femtocell F can be calculated:
()
( ) ( ), ( ),
( ) 1 1
mf a
NN
M F m f a m f a
m f a
TC



(7)
In generation, the objective of this paper is how to optimal
the total throughput by allocating sub-bands for each
femtocell after using FFR to assign sub-band for MUEs. Thus,
we can formulate the problem of allocating sub-bands to
femtocell as follows:
,,
;
, , , ,
1 1 1 1 1 1
f
m a a
MF
m a f a
NN
N
N N N
m a m a f a f a
M m a F f a
CC
Max


     









  
(8)
Subject to:
1) βm,a x βf,a = 0
a
aN
2) βf,a x βf’,a = 0 if df,f’ < dthreshold
3) SINRa,m>SINRthreshold
In Eq. (8), NM is the total number of macrocells in entire
networks and NF is the total number of femtocells within each
macrocell coverage area. Nm and Nf are total number of MUEs
and FUEs in each macrocell and femtocell, respectively.
Subject 1 makes sure that there is no channel that assign for
both MUE and FUE , and subject 2 dedicates that there is no
channel assign for both femtocells which near to each other.
In this way, the resource allocation problem becomes
calculating the optimal parameter set
af ,

, such that the total
throughput in Eq. (8) can be maximized. The proposed
resource allocation mechanism is presented in next section.
IV. PROPOSED METHOD
In this section, a fractional frequency reuse-based graph
connectivity (FFR-GC) mechanism is proposed to solve the
interference problem. FFR-GC scheme can be partitioned into
two phases. In phase 1, the frequency sub-bands for macrocell
users are allocated based on FFR to reduce interference
experienced by MUEs. Then, a graph-coloring based
algorithm which reduce the co-tier interference among
femtocells and cross-tier interference between macrocell and
femtocells is adopted in phase 2. This resource assignment
sequence is reasonable since the traffic demand of MUEs
should be fulfill first, then the throughput of FUE can be
enhanced.
A. Phase 1: Macrocell Radio Resource Allocation
The FFR-based sub-band allocation mechanism proposed
by Assaad [18] is adopted to assign sub-bands in macrocell.
Each macrocell is partitioned into center zone and three edge
zones (FFR-3) as show in Fig. 3. It is investigated in previous
works that if the number of sector is larger than 3, there is no
significant performance improvement [19]. The simulation
results presented in [18] show that when 2/3 available
bandwidth are allocated to center zone and the remaining 1/3
available bandwidth is allocated to edge zones then the
overall performance is the best. The size of sub-bands and the
assignment of sub-bands to sub-area in macrocell using FFR
mechanism illustrated in Fig. 3. Four frequency sub-bands
(A,B,C,D) are allocated to sub-areas of macrocell coverage
area (C1, C2, C3 for center zone and E1, E2, E3 for edge
zones). The detail sub-band assignment is illustrated in Fig.
3(b). Using FFR method, the interference for MUEs which
located in edge zones will be reduced. For example, MUEs in
edge zone E1 can be interfered by macrocell 4 and macrocell
5 as shown in Fig. 3(a). Moreover, according to the above
FFR method, the cross-tier interference between macrocell
and femtocell are also avoided in each cell.
Fig. 3. Illustration of FFR macrocell frequency sub-bands assignment. (a)
Sub-bands assignment in macrocell. (b) The sub-area layout and allocation
for UEs using sectored-FFR.
As shown in Fig. 3(b), the MUEs and FUEs will be
assigned by different sub-band in each sub-area. For example,
Journal
of Advances in Computer Networks, Vol. 4, No. 2, June 2016
82

in edge zone E1, the sub-band B is assigned for macrocell, the
remaining sub-bands, {A, C, D}, can be assigned for
femtocell. In the center zone C1, the femtocell can use
sub-band C and sub-band D since sub-band A has already
been assigned to macrocell in center zone. The sub-band B is
assigned for macrocell edge zone E1 area where is near to C1
area. By this way, the cross-tier interference will be mitigated
and the throughput over entire network will be enhanced.
B. Phase 2: Frequency Sub-band Algorithm for Femtocell
1) Sub-bands for femtocells
Fig. 4. The frequency allocation for femtocells.
As show in Section III, the resource allocation for
femtocells is NP-Problem. Using the optimal algorithm to get
the optimum resource assignment is not adequate since it
requires the high complexity. Therefore, we propose the
heuristic algorithm as a promising solution to solve with this
problem. Thus, the propose algorithm may provide a
non-optimal algorithm, but it give an efficient way, instead
high complexity computation of the optimal algorithm.
To simplify the problem, we defined again the sub-bands
for femtocell in each sub-area after using FFR for MUEs as
shown in Fig. 3. Accordingly, the sub-band A will be divided
into the six smaller sub-bands {A1, A2, A4, A5, A6}.
Consequently, total bandwidth is partitioned into nine
sub-bands equally (because sub-band A equal 2/3 total
bandwidth).
The sub-bands which assigned for FUEs in each sub-area
are shown in Fig. 4. Here, we can see the number of sub-bands
that assigned for the FUEs in each sub-area of center zone
only equal two while in each edge zone, there are eight
sub-bands that are able to assign for UEs. This problem is
feasible in practical system since almost the UEs in center
zone will require data rates to macrocell base station to get
stable signal instead using femtocell.
The resource allocation problem can be translated into
undirected graph problem where each femtocell is a vertex
and the edge that connecting two vertices if they may interfere
to each other. In this paper, we set the edge between two
vertices if the distance between them smaller than dthreshold
which be defined up to the cell radius R (cell coverage) of a
femtocell base station:
'
()
F1 F1,f
02
10
F1 F1,f
02
10 F2 F'2,f
Pg
log (1
2 N f
arg Pg
log (1
N f P g )
f
Rm
Nf
f
threshold N
f
B
d
B







   








  









(9)
In Eq. 9, B0 is total bandwidth for two femtocells F1, F2
which are neighboring femtocells. Thus, the allocation
problem transform into graph problem based on modified the
maximal independent set (MIS). As the observation that the
MIS algorithm usually gives a good result for reducing
interference due to the set of femtocells using the same
sub-bands is an independent set which not share the same
channel. Moreover, we use iteration MIS for each sub-bands,
so femtocell are able to reuse more than one sub-bands so the
throughput will be enhanced. The process of the algorithm is
given as follows:
1) The first step of the algorithm is to transform the system
mode into undirected graph G(V,E) with V is the set of
femtocell and E is the set of interferences.
2) In second step, with each sub-band s(s

S is total
sub-band)), we find a graph Gs(Vs, Es) (Gs

G) obtains
all the femtocell that the sub-band s is able to assigned.
For example, in Fig. 4, the femtocells which located in
sub-area E1 can use sub-bands {A1,A2,A3,A4,
A5,A6,C,D}. This thing is supported by FGW since the
femtocells are managed by FGW where contain all
information of each femtocell such as the locations, the
neighboring femtocells and the sub-bands that able to be
assigned.
3) The third step is based on modified maximal independent
set algorithm for graph Gs to find a set K which including
all the independent femtocells (no interference) which
assigned by sub-band s (detail in algorithm 1).
4) Final step is repeating from step 2 until the last element of
the set of sub-band S.
3) Practical operation algorithm
Since all the femtocells in network are managed by FGW,
so we can know all information of each femtocell i.e locations,
sub-bands are able to assigned for each femtocell.
Algorithm 1 The Modified MIS Algorithm
INPUT: A graph Gs(Vs,Es), an adjacency matrix

S,V of all
sub-bands S and femtocells v.
OUTPUT: set K with maximal independent set.
1:while

v

Vs do
2: find v= min
,SV
sS


;
3:

is set of independent vertices with v;
4: set K= M_I_S (Gs,v,

);
5: return K ;
Function M_I_S(Gs,v,

)
6:if

=

then return set temp = {v};
7: else while

i


do
8: find i = min
,SV
sS


;
9:temp = {v}i;
10:

temp is set of independent vertices with v;
11: recall M_I_S(Gs,temp,

temp);
UEs in each sub-areas after allocating for macrocell UEs
(Fig. 4). Therefore, for each sub-band, we can find all the
femtocells in network which it is able to assign as example in
Journal
of Advances in Computer Networks, Vol. 4, No. 2, June 2016
83
2) Proposed algorithm for femtocell assignment

Fig. 5(b,c,d). After defining sub-graphs for each sub-bands,
the algorithm 1 presented which femtocell can use those
sub-bands to avoid co-tier interference of femtocells as Fig. 5
(e).
The inputs are the sub-graph of each sub-bands and an
adjacency matrix

which indicates the number of sub-bands
assigning for each femtocell (

s,v =1 if sub-band s

S assign
for vertex v

V). Typically, to make sure all femtocells can
use frequency band (assuming each femtocell have the same
data requirement) in each sub-graph, we will find femtocell v
is assigned least sub-bands (min Σ

S,V) with S is total
sub-band and beginning find the MIS from femtocell v
(function M-I-S) .
Fig. 5. An example of the modified MIS algorithm.
V. PERFORMANCE EVALUATION
In this paper, we evaluate the proposed schemes in terms of
throughput. We also concentrate on the performance of
femtocell UEs and total UEs individually to get the fair
evaluation about the proposed approach. We adopt the system
parameters of 3GPP LTE, which are listed in Table I. The
scenario is consisting of an area of 150 square meters. We
vary the number of femtocells from 20 to 120 in one
macrocell coverage in order to figure out the influence of
variation of femtocell density. The macrocell UEs and
femtocell UEs are randomly distributed in the overall
network.
TABLE I: SIMULATION PARAMETERS
Parameters
Macrocell
Femtocell
Number of cell
7
20-120/cell
Cell Coverage
400 m
30 m
Transmit Power
20W
20mW
Number of User per cell
120 UEs
120 UEs
Map Range per cell
150m × 150m
Channel Bandwidth
20 MHz
Carrier Frequency
2GHz
Number of Resource Blocks
100 RBs
Spectrum of RB
180 kHz
White Noise
-174 dBm
Lognormal Shadowing
10 dB
To show the benefits of the proposed approach, we
compare the capacity of the proposed Fractional Frequency
Reuse based on Graph Connectivity (FFR-GC) with other two
schemes such as the Optimal-FFR scheme and Graph
Coloring based Cognitive Spectrum Allocation (GC-CSA)
scheme. For the optimal-FFR scheme in [18], they take an
investigation in how to assign sub-bands for UEs using FFR
method to get the best result. In this regards, we use the best
result from this article to compare with our approach. The
GC-CSA scheme was proposed in [13], in this article, they
using graph coloring algorithm to assign sub-bands for
femtocell.
To evaluate the proposed approach carefully, we firstly
compare separately average throughput of macrocell UE (Fig.
6) and femtocell UE (Fig. 7). Then, Fig. 8 shows average
throughput of UE in total network (both macrocell UEs and
femtocell UEs).
Fig. 6 shows the average throughput of macrocell UEs,
since using FFR method, cross-tier interference of both
FFR-optimal approach and proposed approach (FFR-GC)
will be reduced more effective than GC-CSA approach.
However, when the network gets high density of femtocells,
the interference will be more complicated. The macrocell UEs
in FFR Optimal approach will get more interference from
neighboring femtocells. On the other hand, in our approach,
we use graph algorithm to assign the sub-band for femtocell,
the downlink interference from femtocell to macrocell UE
will be reduce.
20 30 40 50 60 70 80 90 100 110 120
10
20
30
40
50
60
70
80
Number of Femtocells
Average Through put (Mbps)
GC-CSA
Optimal FFR
FFR-GC
Fig. 6. Average throughput of macrocell UEs.
Fig. 7 shows the average throughput of femtocell UEs in
each cell. We can see the domination of the proposed
approach since it is able to get the higher SINR for each UE.
The optimal-FFR only focus on using FFR for macrocell UEs
and then reusing remaining sub-bands for femtocell UEs as
much as possible, this thing bring on the high interference
among femtocells. Meanwhile, the GC-CSA approach can
make high results due to this approach use graph coloring to
assign sub-bands for femtocell. Accordingly, the femtocells
where nearby to each other will use different color to reduce
the interference. But they do not consider interference from
eNBs to FUEs so the throughput will be lower.
In Fig. 8, the average throughput of both MUEs and FUEs
in total network is described. Using propose approach
FFR-GC, The throughput of propose approach is always
higher. Since it is not only mitigating cross-tier interference
but also reducing co-tier interference.
Journal of Advances in Computer Networks, Vol. 4, No. 2, June 2016
84

20 40 60 80 100 120
0
5
10
15
20
25
30
Number of Femtocell
Average of Throug hput (Mbps)
GC-CSA
Optimal FFR
FFR-GC
Fig. 7. Average throughput of femtocell UEs.
20 40 60 80 100 120
5
10
15
20
25
30
35
40
45
Number of Femtocells
Average Through put (Mpbs)
GC-CSA
Optimal FFR
FFR-GC
Fig. 8. Average throughput of UEs.
VI. CONCLUSION
Recent years, femtocell is proposed as promising issues to
provide higher throughput service with low cost in wireless
cellular system. However, the interference in two-tier network
in femtocell systems is a main challenge for the researchers as
well as mobile operators. In this paper we propose an
approach to deal with this problem. Firstly, we use FFR
method to assign sub-bands for macrocell UEs to avoid
cross-tier interference in each cell. Then, we propose a
heuristic algorithm based on modified Maximal Independent
Set algorithm to assign the sub-bands for Femtocell UEs to
reduce co-tier interference as well as improve throughput. As
results are shown in section V, the proposed algorithm is not
only reducing the interference in two-tier network but also
improve throughput compare with previous approaches.
Moreover, this approach also can be used in the large size of
the network.
ACKNOWLEDGMENT
The authors would like to thanks the partial financial
support from National Science Council in Taiwan with
project number: NSC-103-2221-E-156-009.
REFERENCES
[1] Cisco White Paper, Cisco Visual Networking Index: Global Mobile
Data Traffic Forecast 2012–2017, February 6, 2013.
[2] S. Carlaw, “IPR and the potential effect on femtocell markets,”
FemtocellsEurope, ABIresearch, 2008.
[3] V. Chandrasekhar and J. Andrews, “Femtocell networks: A survey,”
IEEE Commun. Mag., vol. 46, no. 9, pp. 59-67, Sept, 2008.
[4] 3GPP, “Ts 36.201 v8.1.0,” Tech. Rep., April 2009.
[5] N. Saquib, E. Hossain, L. B. Le, and D. I. Kim, “Interference
management in OFDMA femtocell networks: Issues and approaches,”
IEEE Wireless Commun. Mag., vol. 19, no. 3, pp. 86-95, June 2012.
[6] V. Chandrasekhar and J. G. Andrews, “Uplink capacity and
interference avoidance for two-tier femtocell networks,” IEEE Trans.
Wireless Commun., vol. 8, no. 7, pp. 3498–3509, Jul. 2009.
[7] S. Park, W. Seo, Y. Kim, S. Lim, and D. Hong, “Beam subset selection
strategy for interference reduction in two-tier femtocell networks,”
IEEE Trans. Wireless Commun., vol. 9, no. 11, pp. 3440-3449, Nov.
2009.
[8] M. S. Jin, S. Chae, and D. I. Kim, “Per cluster based opportunistic
power control for heterogeneous networks,” in Proc. IEEE VTC’
11-Spring, Buadapes, Hugary, May 2011.
[9] S. Gurauacharya, D. Niyato, E. Hossain, and D. I. Kim, “Hierarchical
competition in femtocell-based cellular networks,” in Proc. IEEE
Globe-com’ 10, Miami, FL, USA, 6-10 Dec., 2010.
[10] H. Jung, J. Han, and J. Lee, “Interference avoidance and resource
allocation for OFDMA downlink of femtocells with distributed power
control,” IEICE Trans. Commun., vol. E93b, no. 4, pp. 1061-1064,
Apr. 2010.
[11] D. Lopez-Perez, A. Valcarce, G. de la Roche, and J. Zhang, “OFDMA
Femtocells: A roadmap on interference avoidance,” IEEE Commu,
Mag., vol. 47, pp. 41-48, Sept. 2009.
[12] D. Lopez-Perez, A. Ladanyi, A. Juttner, and J. Zhang, “OFDMA
Femtocells: A-self-organizing approach for frequency assignment,” in
Proc. PIMRC, 2009, pp. 2202-2207.
[13] L. Tan, Z. Feng, W. Li, Z. Jing, and T. A. Gulliver, “Graph coloring
based spectrum allocation for femtocell downlink interference
mitigation,” in Proc. IEEE Wireless Commun. Netw. Conf., Mar. 2011,
pp. 1248–1252.
[14] L. Zhang, L. Yang, and T. Yang, “Cognitive interference management
for LTE-A femtocells with distributed carrier selection,” in Proc. IEEE
VTC’, Sept. 2010, pp. 1–5.
[15] T. Lee, J. Yoon, S. Lee, and J. Shin, “Resource allocation analysis in
OFDMA femtocells using fractional frequency reuse,” in Proc.
PIMRC, Sept. 2010, pp. 1224-1229.
[16] 3GPP, “Tr 36.814 v9.0.0,” Tech. Rep., March 2010.
[17] 3GPP, “ETSTTr 136.931 v9.0.0,” Tech. Rep., May 2011.
[18] M. Assaad, “Optimal fractional frequency reuse (FFR) in
Multi-cellular OFDMA System,” in Proc. IEEE VTC’, Sept. 2008.
[19] D. Lee, C. Xu, U. Mayekar, and M. Mohile, “Frequency reuse factor vs.
pathloss exponent and sectorization,” IEEE MTT-S Symp. Tech.
Wireless App. Digest, pp. 109-112, Feb. 1997.
Gwo-Jong Yu received the B.S. degree in computer
science from Christian University, Zhongli, Taiwan, in
1989, and the Ph.D. degree in computer science from
the National Central University, Zhongli, Taiwan, in
2001.
Since August 2001, he was with the Faculty of
Department of Computer Science and Information
Engineering (CSIE), Aletheia University, New Taipei,
Taiwan. He became a Professor at the Department of
CSIE, Aletheia University, in 2011. His current research interests include
wireless sensor networks, ad hoc networks, WiMAX, and LTE.
Hoai Nam Bui Khac was born in Quang Binh,
Vietnam in 1988. He received the M.S. degree in the
Department of Computer Science and Information
Engineering (CSIE) in 2014 from Aletheia University,
New Taipei, Taiwan. Currently, He is working in the
Department of Information and Communications
Technology in Quang Binh, Vietnam.
His research interests include radio resource
management, interference management, and LTE.
Journal of Advances in Computer Networks, Vol. 4, No. 2, June 2016
85