Inter-cell interference coordination in LTE-A HetNets: A survey on self organizing approaches

Abstract

Heterogeneous Networks (HetNet) are multi-mode, multi-layer, and multi-band structured and utilize cells of varying sizes. The goal behind the implementation of HetNets involves incrementing the capacity of the established network, modifying spectrum use, lowering the capital and operating costs, and offering steady user-based experience of network architecture. However, these random small cell deployments cause severe problems and results interference in the network. Therefore, the ultimate is the total system performance degradation and this interference becomes a key challenge in HetNet. This article investigates the performances of the current trends and approaches of interference in self organizingnetwork for LTE-A.

Full text

2013 INTERNATIONAL CONFERENCE ON COMPUTING, ELECTRICAL AND ELECTRONIC ENGINEERING (ICCEEE)
978-1-4673-6232-0/13/$31.00 ©2013 IEEE 196
Inter-cell Interference Coordination in LTE-A
HetNets: A Survey on Self Organizing Approaches
M. K. Hasan, A.F. Ismail, Aisha H. Abdalla,
Khaizuran Abdullah, HAM. Ramli, Shayla Islam
Department of Electrical and Computer Engineering,.
International Islamic University Malaysia
hasankamrul@msn.com
Rashid A. Saeed
Department of Electronics Engineering
Sudan University of Science and technology
Faculty of Engineering
eng_rashid@ieee.org
Abstract—Heterogeneous Networks (HetNet) are multi-mode,
multi-layer, and multi-band structured and utilize cells of
varying sizes. The goal behind the implementation of HetNets
involves incrementing the capacity of the established network,
modifying spectrum use, lowering the capital and operating
costs, and offering steady user-based experience of network
architecture. However, these random small cell deployments
cause severe problems and results interference in the network.
Therefore, the ultimate is the total system performance
degradation and this interference becomes a key challenge in
HetNet. This article investigates the performances of the current
trends and approaches of interference in self organizingnetwork
for LTE-A.
Key Words—inter-cell interference;cognitive radio; LTE-A;
femtocell.
I. INTRODUCTION
The number of wireless subscribers have increased
exponentially with time and telecommunication companies
are perennially challenged to meet the customers’
requirements for increased coverage and faster data transfer
rates.The release 8/9 standard provides major benefits in High
Speed Packet Access (HSPA) which includes better spectral
efficiency, reduced latency on account of its Internet Protocol
(IP) architecture, and greater throughputs[1]. Nevertheless,
inspite of its enhanced performance, Release 8/9 is not up to
the par in terms of the advanced standards set by the
International Mobile Telecommunications (IMT) for fourth
generation mobile networks. Such networks were defined by
the International Telecommunication Union (ITU).
Accordingly, to meet such necessities which is for downlink
data rates of up to 500 Mb/s and 1 Gb/s for mobile and
nomadic users, respectively, LTE Release 10 is now under
standardization. In HetNet radio link quality can be enhanced
due to the reduced distance between transmitter and receiver,
and the larger number of cells allows for more efficient
spectrum reuse and therefore larger data rates. As a result,
HetNets are expected to be one of the major performance
enhancement enablers. Among the low-power nodes in a
HetNet, Pico-eNodeBs and HeNodeBs are very important part
for capacity enhancement and coverage extension within the
coverage areas of eNodeB. In Figure. 1 HetNet network
scenario is decomposed. LTE-A HeNodeBs are observed as a
promising alternative for mobile operators to develop
coverage and provide high-data-rate services in a cost-
effective manner by decreasing the macro-eNodeB traffic load
and offloading it over public broadband connections to core
network. In order to evade interferences, HeNodeBs
synchronization is veritably important. The co-channel
deployment in macro-eNodeB and HeNodeBs could increase
the capacity of the network manifold through high spatial
frequency reuse. However, co-channel deployment in macro-
eNodeB and HeNodeBs results interference in the network
which becomes a key challenge in HetNet.
The rest of the paper is organized as follows. Section 2 the
architecture, Section 3 related works, Section 4 Interference
approaches and performance analysis, and in section 5
summarizes the paper.
Fig. 1: Network structure of HetNet Scenario in LTE-A
II. LTE-A SYSTEM ARCHITECTURE
LTE-A is set with forceful performance requirements that
depend on physical layer technologies such as Orthogonal
Frequency Division Multiplexing (OFDM), Multiple-Input
Multiple-Output (MIMO) systems, and Smart Antennas for
accomplishing these targets. In Figure. 2 displays an OFDM
signal [2] with 5 MHz bandwidth. It is significant that the data

197
symbols are individually modulated and transmitted over a
densed spaced orthogonal sub-carriers.
Fig. 2: Frequency-time illustration of an OFDM Signal [2]
In downlink, the subcarriers are divided into resource
blocks which empower the system to be capable of arranging
the data across standard numbers of subcarriers compartment
wise. Resource blocks comprise of 12 subcarriers, one slot in
the time frame irrespective of the general LTE-A femtocell
signal bandwidth (See in Figure. 3). It can be understood that
dissimilar LTE-A signal bandwidths will have diverse
numbers of resource blocks. Furthermore, the subframes are
assembled in 10 msradio frames, which holds two 5 ms halves
containing the signals essential to acquire the physical identity
of the cell. The signals are the primary and secondary
synchronization signals for acquisition channels, also called
the physical cell identity (PCI) of the cell, and the physical
broadcast channel (PBCH), be responsible for some critical
system information such as the DL transmission bandwidth
and the number of DL antenna ports. The acquisition channels
share the property of spanning the middle six RBs of the
system band-width [3].
For the LTE uplink, a different perception uses of the
access technique while still using OFDMA technology, the
implementation is known as Single Carrier Frequency
Division Multiple Access (SC-FDMA). With the RF power
amplifier that transmits the radio frequency signal through the
antenna to the base station being the maximum power item
inside the mobile, it is essential that it functions as competent
mode as possible [4]. However, it can be meaningfully
affected by the procedure of radio frequency modulation and
signal format. Signals containing a large peak to average ratio
and necessitate linear amplification do not lend themselves to
the usage of efficient RF power amplifiers. Consequently the
implication of a transmission mode has a continuous power
level while in function. However, OFDM contains a high peak
to average ratio. While this is not a difficulty for the base
station where power is an imprecise problem, it is unsuitable
for the mobile. Thus, LTE-A makes use of a modulation
method addressed as SC-FDMA- Single Carrier Frequency
Division Multiplex. This is hybrid format and integrates the
low peak to average ratio offered by single-carrier systems
along with the multipath interference resilience as well as
flexible subcarrier frequency allocation offered by OFDM. In
LTE-A the guard interval is a cyclic prefix (CP) which is
inserted prior to each OFDM symbol.
Fig. 3: illustration of for OFDMA DL Physical layer arrangement [3]
In the time domain, adding a CP to each symbol can be
useful to mitigate inter-OFDM-symbol-interference due to
channel delay spread. The data throughput capacity will be
reduced once the CP length is too long. For LTE-A, the
standard length of the cyclic prefix has been chosen to be 4.69
µs . This enables the system to accommodate path variations
of up to 1.4 km. With the symbol length in LTE set to 66.7 µs
[2]. In Figure. 4 demonstrated the CP adding in a single
carrier transmission.
However, the block-wise single carrier generation
equalization need to most accurate and the channel should be
constant over time span corresponding to the size of the
processing block.
This constraint provides an upper limit on the block size N
that fully depends on the rate of the channel variations.
Additionally, The OFDM subcarrier spacing depending on the
rate of the channel variations [2].
Fig. 4: CP adding in a single carrier transmission[2]
III. LITERATURE REVIEW
In order to achieve a solution, researchers have considered
various types of industrial challenges towards large
deployment of HeNodeB, The interference occurs vastly in
macro-eNodeB-to-macro-eNodeB, HeNodeB-to-HeNodeB,
and macro-eNodeB-to-HeNodeB, thus aggravating the
performance of the system. The cell edges users suffer from

198
low throughput, with the increasing number of cells, which is
caused by interference [5].The authors in [6][7], studied open
vs. closed access for uplink of HeNodeB OFDMA based
networks. Open access reduces interference and endow witha
reasonable way to increase the capacity of the operator’s
network. Moreover, authors showed numerically and by
simulations that the certainty is more problematical and be
subject to dense cellular user. For instance, closed access is
characteristically superior at sky-scraping user concentrations
in orthogonal multiple access. The authors inspected the
impacts of access modes and directional antenna on link
consistency and ability for the OFDMA-based HeNodeB [8].
However, under the link steadfastnesscondition of in building
and outdoor users, the enhancement in femtocell capacity due
to open access mode is minor. The interference can be
categorized in to two section which is discussed as below:
A. Co-tier Interference
Co-tier interference is consigned based on the undesirable
signals received by UEs from the HeNodeB adjacent co-
channel. Co-tier refers to interfering signal that is received
from the similar network tier in close access mode. The
interference can be handled effectively with the ICIC
techniques standardized by 3GPP in LTE/LTE-A. HeNodeBs
are usually deployed by end users and has a lack of influential
backhauls, henceforth fast reacting ICIC techniques are
impervious [9][10]. Thereby, co-tier interference is formed by
HeNodeBs owing to the low isolation of walls or windows.
The avoidance of co-tier interference is achievable by
OFDMA HeNodeBs by properly assigning frequency
resources among users in a bigger time scale or by self-
organizing methods.
B. Cross-tier Interference
Conceptually, cross-tier interference is significant for
macro-eNodeB users near Closed Subscriber Group (CSG)
HeNodeBs, as they are not permitted to connect for closing
HeNodeBs with reduced path losses than their operating
macro-eNodeBs, which occurs due to the connectivity rights.
Thus, the macro-eNodeB users may experience outage due to
this reason and the general communication, which usually
takes place, may not be continued since there will be no
competent backhaul connection between macro-eNodeBs and
HeNodeBs [9]. The DL/UL cross-tier interference scenarios
has been shown in Figure. 5.
Fig. 5. DL/UL cross-tier interference scenarios in HetNet
Consequently the equipment becomes used to its ‘nature to the
local context. The authors presented access control
techniques[12][13][14][15][16], Sensing Approaches
[17],Hardware approach [18], interweave techniques[19] in
self organizing networks. The performance along with the
algorithm is described in the next sections.
IV. PERFORMANCE ANALYSIS OF CR BASED
APPROACHES
A. Access control Techniques
A prodigious compact of power allocation polices and
interference control approaches have been proposed for
spectrum sharing CRNs [12]. For instance, the optimal power
allocation schemes to maximize the capacity of the secondary
user with an effective protection of the primary user (PU) for
spectrum-sharing CRNs. wavelet energy entropy
basedspectrum sensing algorithm was proposed to sense the
spectrum in presence of PU comparing with signals to a
threshold [11][13]. However, in this algorithm sensing delay is
observed which is not applicable in the HetNets. To sense the
spectrum in CRN SNR based adaptive spectrum sensing
algorithm was proposed which was the mix techniques of
energy detector and cyclostationary detection. The algorithm
shows that it is working as like OR function with this two
detection [14].However, it can be seen that the lower SNR rate
this technique is not suitable for HetNets. In Figure.6(a) and
6(b) shows the block diagram of energy detector and
cyclostationary detection.
(a)
(b)
Fig. 6. Block diagram of (a) energy detector. (b) cyclostationarydetector[14]
A power controlled based spectrum sensing techniques
was proposed[15] using interference level threshold to
coordinate the interference. The technique was able to bind
the interference limit. However, the technique was proposed
for TV white space reuse which is imperfect for the licensed
spectrum such as HeNodeBs neighbor networks. The
calculated probability density ofentire cell throughput for
closed and open access at a home environment where
considered a largely deployment HeNodeB and macro-
eNodeB [16]. As a result of the critical interference produced
by HeNodeB to users which are in to the macro-eNodeB
coverage. It is found that open access performance satisfactory
than closed access. The closed access technique be inclined
tohave the whole cell throughput to 15% lower than open

199
access method. In hybrid access, the accessibility of HeNodeB
can be precise and it can be constructed to assurance a
leastrecital. Additionally, the allocation of HeNodeB resources
between subscribers and non-subscribers should be highly
tuned. However, the hybrid access techniquerequired for
adjustment to the HeNodeB arrangement developments.
B. Neighbour Network Sensing Approach
HeNodeB neighbour networks sensing approach comprises
of constructing the sensing capability into the HeNodeB
neighbour networks which are proficient of sensing the users
in range. Hence, it may accurately and proficientlyemploy its
spectrum by means of the channel sensing, which is alike to a
user terminal working in idle mode [17]. In Figure. 7, a
depiction of adjacent HeNodeBs sensing is shown.
Fig. 7. Neighbor Network sensing approach[17]
C. HeNodeB-HeNodeB interoperability Approach
HeNodeB-HeNodeB interoperability is an approach where
the neighboring HeNodeB require interlink with each other for
the aim of checking the user information and the mode of
operation. For this type of communication between neighbor
HeNodeBs can occur through a direct link between HeNodeB
namely HeNodeB gateways, or with the use of the UE (Figure.
3)[17]. Therefore, a caution is given by HeNodeBs before
there is an arrival of the current and any other future
movements of their neighboring cells. These messages can be
replaced through the HeNodeB gateway, by utilizing the X2
interface between cells. The broadcasting messages can also
be interchanged via the mobile terminals. Nonetheless, there is
not much competition present between them to operate in
certain circumstances where the HeNodeBs are out of range of
coverage with another HeNodeBs. In Figure. 8, user 3 has
been positioned at the cell edge of two overlapping HeNodeBs
in which they are unseen to each other. Such a user is likely to
suffer from this circumstance due to the interference, as
HeNodeBs are unable to coordinate their resource allocation
due to the vacancy of information acquired in the sensing
D. UE supported Sensing Approach
This kind of approach is capable of performing a side -step
HeNodeB range difficulties by the utilization of measurement
report that may be functioned by UEs and acknowledged by
the HeNodeBs. The UEs requires sending the information to
the HeNodeB regarding the current positioning, receive signal
strength as well as active sub channels of the operating and the
toughest adjacent cells (both macro-eNodeB and HeNodeB),
in order to function and self-optimize HeNodeB more
efficiently. This approach gives an opportunity to UEs for
further their information about their instantaneous radio
environment (user location) to their FAPs and aid in
diminishing interference according to Figure. 9 scenario[17].
.
Fig. 8. HeNodeB-HeNodeB interoperability [17]
Fig. 9. UE supported sensing [17]
E. Hardware Centric Approaches
A presentation of a hardware centric approach for interference
cancellation techniques employed in layer1 in order to
coordinate interference[18]. Practically, in the cellular system
the downlink and uplink characteristics are very different for
increasing the capacity of cellular systems. And from the
downlink, each receiver requires the decode of a single desired
signal from the K of intra-cell signals, while suppressing other
cell interference from a few dominant neighbour cells,
according to Figure.10. Due to the origin of K user signal is
from the base station, the link is synchronous and the K-1
intra-cell interference is able to be orthogonalized at the base
station transmitter [18]. However, some quadration is lost in
the channel. Besides that, the base station receiver must
decode all K desired users in the uplink, while suppressing
other cell interference from independent sources, as shown in
Figure. 11.
Fig. 10. Downlink interference scenario [18]
Fig. 11: Uplink interference scenarios [18]

200
F. Interweave Technique
In [19][20], authors have detailed models for cognitive radios
on the basis of the overlay technique through two switch
cognitive radio model. In where, HeNodeBs approves various
allocation methods to attain diverse spatial reuse levels.
Figure. 12 illustrated diverse standpoints on switch model in
cognitive radio with transmitter ST and receiver SR where
nodes are noted as A, B and C signifying the PU of the
spectrum. It can be observe when the transmitter and receiver
are far away, the PU actions are more distributed and lower
correlation. For capacity of the two switch model, authors
provided tight upper and lower bounds of cognitive radio
channel which is expressed in equation 1 [19].
Fig. 12. Two switch model cognitive radio transmitter and receiver [19]
(1)
In a lower bound capacity, a genie argument is used to attain a
lower bound capacity while considering the cognitive
transmitter receiver pair. Assumption is made on a genie
which offers side information to the receiver with the use of
every channel in a variable G. The genie bound result
demonstrates that the capacity increasing because of the genie
information G given to the receiver are not allowed to surpass
the entropy genie information rate. This is specified in
equation 2 and 3 [19].
(2)
(3)
where, is the entropy rate of the genie information G
specified the receiver state . When rises, equation 3 shows
that the genie lower bound swiftly approaches the capacity
with whole information at the receiver. Hence it is assumed
that the capacity of the two switch channel model equation
1.The positioning of the primary users in the system are
apprehended by a poisson point process with density of 
primary nodes per unit area where the probability of finding k
primary in the area can be stated in equation 4 and the
capacity of the secondary link can be foundthrough using the
equation 5[19]. Where, P is noted as the power constraint at
the secondary transmitter.
(4)
(5)
Figure. 13 illustrated the secondary user throughput
against the radius of the sensing regions Rsfor various primary
user densities λ. At the same time with the increase of sensing
region the sensitivity of detection keeps increasing, the
average number of communication opportunities reduce
causing a little throughput according to the expected level.
Furthermore, in these methods HeNodeB behave as secondary
user with resource level information that can look for
provisional frequency holes for ingenious allocation. So,
HeNodeB-UEs and macro-eNodeB UEs function on
orthogonal bands, and less cross-tier interference has
observed. Nevertheless, this approach limits the system
capacity as no spatial reuse is adopted.
Fig. 13. Throughput vs. sensing radius for different values of  [19]
Fig. 14. Illustration of transmission radius on transmission capacities [20]

201
V. CONCLUSION
Coordination of inter-cell interference in LTE-A HetNet is
urge for the operators in order to mitigate interferences as well
as channel capacity increase. In this paper, we presented the
architecture of the OFDMA techniques and investigated the
current proposed approaches and techniques with pros and
cons. However, it can be summarized that self organinzing
techniques are promising in terms of lessen interferences than
the other techniques.
REFERENCES
[1] Lopez-Perez, David, Ismail Guvenc, Guillaume de la Roche,
MariosKountouris, Tony Quek, and Jie Zhang. Enhanced
intercell interference coordination challenges in heterogeneous
networks, IEEE Wireless Communications, 2011.
[2] Erik D., S. Parkvall, J. Skold, 2011. 4G: LTE/LTE-Advanced for
Mobile Broadband: LTE/LTE-Advanced for Mobile Broadband,
Published by Elsevier.
[3] Aleksandar D., J. Montojo, Y. Wei, T. Ji, T. Luo, M.
Vajapeyam, T. Yoo, O. Song, and D. Malladi, A Survey
On3GPP Heterogeneousnetworks, IEEE Wireless
Communications, 8(3), pp.10-21, Jun. 2011,doi:
10.1109/MWC.2011.5876496.
[4] Holma H. and A. Toskala, 2011. LTE for UMTS Evolution to
LTE-Advanced. John Wiley & Sons, 2 edition.
[5] Mehmet Y., M. Farhad, N. Sanjiv, P. Akhilesh, J. Nick, R.
Balaji and R. Andy, 2009. Interference Management and
Performance Analysis of UMTS/HSPA+ Femtocells, IEEE
Communications Magazine, IEEE Journal and Magazine,
47(9),pp. 102 – 109, doi:10.1109/MCOM.2009.5277462
[6] Mhiri, F., K. Sethom, and R. Bouallegue. A survey on
interference management techniques in femtocell self-organizing
networks, Journal of Network and Computer Applications, 2013.
[7] Xia P., Chandrasekhar V, Andrews JG. , Open vs. closed access
femtocells in the uplink. IEEE Trans Wireless Commun
2010;9(12):3798–809
[8] Huang J., W. Li-Chun, C. Chung-Ju, S. Wen-Shan, Design of
optimal relay locations in two-hop cellular systems.
ACM/Springer Wireless Networks 2010;16(8)
[9] Jun Z., Hui T., Peng T.,Y. H., Liqi G., 2012. Dynamic
Frequency Reservation Scheme for Interference Coordination in
LTE-Advanced Heterogeneous Networks,IEEE Vehicular
Technology Conference (VTC Spring), pp. 1-5, doi:
10.1109/VETECS.2012.6239915.
[10] M. K. H., R. A. Saeed, A. A. Hashim, S. Islam, R.A. Alsaqour
and T. A. Alahdal, Femtocell Network Time Synchronization
Protocols and Schemes. Research Journal of Applied Sciences,
Engineering and Technology, 4(23): 5136-5143. 2012
[11] Feng G., R. Rangnekar, A. Radhakrishnan, S. Nair, A. Fayez, Q.
Chen, A. Young,Y. Wang, M. D. Silvius, T. Brisebois, G.
Marballie, X. Cheng, N. He, B. Li,C. W. Bostian, and M. Hsiao,
A heterogeneous cognitive radio network enabling dissimilar
cooperative spectrum sensing, dynamic spectrum access,
interoperability, IEEE ComSoc, DySPAN, 2008.
[12] Kang X, Zhang R, Liang Y-C, Garg HK., Optimal power
allocation strategies for fading cognitive radio channels with
primary users outage constraint. IEEE Journal of Selected
Areas Communication 2011;29(2):374–83.
[13] Z. Jiang, Q. Zhang ; Y. Wang, X. Shang, Wavelet packet
entropy based spectrum sensing in cognitive radio, IEEE
Conference on Communication Software and Networks, pp.
293-298, 2011, doi: 10.1109/ICCSN.2011.6014054
[14] E. Waleed, Najam H. and Hyung S. K., SNR-BASED
ADAPTIVE SPECTRUM SENSING FOR COGNITIVE,
International Journal of Innovative Computing, Information and
Control, 8(9), pp. 6095-6105, 2012
[15] R. Harish Kumar, T. SyamaSundara, Dr. N. Padmaja,
Department of Electronics and Communication Engineering,
Power Control Mechanism for Cognitive Radios via Spectrum
Sensing with Interference Management,International Journal of
Engineering Trends and Technology (IJETT) - 4(9), April 2013.
[16] Lopez-Perez D., Valcarce A., De La Roche G., Enjie L., Jie Z.,
Access methods to WiMAX femtocells: a downlink system-
level case study. In: Proceedings of the 11th IEEE Singapore
international conference on communication systems, 2008
[17] Lopez-Perez, D., Valcarce, A. , De la Roche, G.,
JieZhang,OFDMA Femtocells: A Roadmap on Interference
Avoidance. IEEE Communications Magazine , pp.41-48,
September 2009
[18] W. Yuanye, and I. P. Klaus, Performance Analysis of Enhanced
Inter-cell Interference Coordination in LTE-Advanced
Heterogeneous Networks, IEEE 75th Vehicular Technology
Conference (VTC Spring), pp. 1-5, Jul. 2012,
doi:: 10.1109/VETECS.2012.6240233
[19] Srinivasa S., SA. Jafar, The throughput potential of cognitive
radio: a theoretical perspective. IEEE Communication
Magazine,45(5), 2007pp.73–9.
[20] Shin-Ming C., Weng A. Chon, F. Tseng and K. Chen, Design
And Analysis Of Downlink Spectrum Sharing In Two-Tier
Cognitive Femtocell Networks, IEEE Transaction On Vehicular
Technology, Vol. 61, No. 5, pp. 2194-2207, 2012.