A mode selection algorithm for mitigating interference in D2D enabled next-

generation heterogeneous cellular networks

Md Kamruzzaman1, Nurul I Sarkar1, Jairo Gutierrez1, and Sayan Kumar Ray2

1Department of IT and Software Engineering, Auckland University of Technology, Auckland, New Zealand
2Faculty of Business and Information Technology, Manukau Institute of Technology, Auckland, New Zealand.

{zkamruzz,nurul.sarkar,jairo.gutierrez}@aut.ac.nz,[email protected]

Abstract— Device-to-Device (D2D) communication is the end of 2016. It is expected to grow further sevenfold
becoming more popular research topic among the network between 2016 and 2021, reaching 49.0 Exabyte per month by
researchers and engineers across the globe This popularity is 2021.
due to various key applications of D2D communications
including 4G LTE-A envisioning the continuous evolution of Besides these, in 3rd Generation Partnership Project (3GPP)
current cellular networks into 5G to support a much larger and Release 12 for proximity-based services (ProSe) and group
diverse range of applications. D2D communications play an communication system enablers (GCSE) D2D communication
important role to fulfill the quality of service (QoS) is an integrated module in Forth Generation (4G) Long Term
requirements for 5G networks. However, to realize the Evolution (LTE) Advanced standard. During natural disasters
performance gain of D2D communications, various open like earthquake or hurricane, with the help of D2D
challenges need to be resolved including mode selection (MS). functionality, a replacement of the traditional network can be
This paper, therefore proposes a new mode selection algorithm set up within a short period of time. Furthermore, multi-hop
for D2D enabled heterogeneous 5G cellular network. By cooperation between devices can help to enhance the coverage
quantifying the key parameters we mitigate interference that which may be the only mode of communication in case of none
affect the system performance, and finally proposed an coverage area as well as emergency situation. Hence, D2D
algorithm for optimal selection of transmission mode in a communications play important role for traffic off-loading,
heterogeneous cellular network. The findings reported in this improving spectrum efficiency and latency, capacity and
paper provide some insights into the design of 5G networks that coverage enhancements, and energy efficiency in the future
may help network designer and engineers to contribute further cellular network [3].
towards the development of next generation wireless networks. However, to maximize the performance gain of D2D
communication, there are many open challenges that need to be
Keywords—Device to Device communication, D2D, Mode thoroughly addressed [1]. These challenges include mode
Selection, HetNets, interference management Public Safety. selection, neighbor discovery, interference and radio resource
management, energy consumption, the coexistence of D2D
with small cells, mobility management, and network security.
I. INTRODUCTION Among them, mode selection (MS) in a heterogeneous scenario
In cellular network, Device-to-Device (D2D) communication is comparing all tiers simultaneously is very important and
an emerging technology where two proximity mobile UEs complex.
(User Equipment) communicate each other without any Base
Station (BS) or core network. Due to the short communication 1st Tier
range between a D2D pair, D2D communication provides
several advantages in terms of spectrum efficiency, throughput, 1st <–> 2nd tier int.
CUE
delay, power management, coverage expansion, and capacity eNB (Scenario-1)

improvement by reusing radio resources and allowing network 1st <–> 3rd tier int.
(Scenario-2)
functionalities to the devices.
Furthermore, D2D communication enables new services such 2nd Tier
as public safety, location-based commercial proximity, content 3rd Tier

sharing like files, videos or pictures, gaming, connectivity SCeNB SUE

DUE1 2nd <–> 3rd tier int.
extension, and traffic offloading [1]. Owing to these benefits, (Scenario-3)
D2D communication is considered one of the key technologies
DUE2
in next-generation networks.
Last decade experienced an unexpected growth both in mobile
broadband traffic as well as end-user demand for faster
accessing the data. According to the latest Cisco visual
networking index [2], mobile data traffic was grown up 18-fold, Figure 1: Interference in D2D communication in 3-tier networks
400 petabytes per month in 2011 to 7.2 Exabyte per month at
Figure 1 shows the possible interference scenario in a 3-tier where devices form a coalition to share the spectrum among
cellular network like interference between the 1st-tier and the DUEs and CUEs.
2nd -tier, the 1st-tier and the 3rd-tier, and interference between
the 2nd-tier and 3rd-tier are. Hence, to avail the benefits of D2D All the aforementioned state-of-the-art techniques describe how
communication in cellular network interference management is to control interference in single or 2-tier cellular networks.
essential and by selecting an appropriate mode of transmission Moreover, most of the MS or resource allocation schemes in the
we can achieve this. In spectrum sharing scenario there are two above literature consider only one D2D pair and both of them
modes of operation either cellular mode (CM) where DUEs are in the same cell. Very few papers focus where DUEs in a
exchange the information via eNB by using orthogonal D2D pair are attached to different eNBs. But in practice,
spectrum or sharing mode (SM) where CUE and DUE both multiple CUEs may associate with single eNB and form
share the same spectrum resources. multiple D2D pairs. Not only multi-cell but in future a large
number of small cells, especially femtocells will be deployed
II. RELATED WORK into the cellular network (5G and beyond) which will create a
The state-of-the-art researchers proposed various methods for multi-tier network. In that multi-tier network the most
interference management in D2D enabled 2-tier cellular prominent challenge is how to minimize the interference by
network such as split frequency, selection of transmission selecting appropriate mode of transmission.
mode, power control, resource allocation, cognitive radio, game
theoretic etc., and their working principles as well as techniques Despite the fact, there are very few on-going research efforts to
are also different. address the MS as well as resource allocation for interference
problems in 3-tiers cellular networks. Future D2D enabled
Cross-tier and co-tier interference both can be avoided by using cellular network will be heterogeneous and the complete
proper mode selection (MS) algorithm. In the-state-of-art work, distributed solutions for the mode selection as well as radio
many researchers focus on MS with different like path loss [4], resource allocation problems in a D2D-enabled multi-tier
based on received signal strength over the D2D link [5] or the scenario has not been studied comprehensively in the literature.
physical distance among the devices [6-8]. However, these Nonetheless, none of the authors examines the worst and, at the
schemes do not reflect exact channel quality or interference same time, the most challenging interference case, where all 3-
issues. tiers of the networks (i.e. eNB, SCeNB and D2D
communication) mutually interfere. The main difference
In [9, 10] MS has been performed based on the channel quality. between the exiting approach and our proposal is mode
In particular [10], P. Janis et al. consider the sum rate of the selection in a 3-tier heterogeneous networks.
connection between D2D pairs as well as CUEs as the MS
criterion. In [11], Fan et al. worked on mode selection and
resource allocation issues for D2D communications. In their III. PROPOSED MODE SELECTION ALGORITHM FOR
proposal, they ensured the proper transmission of cellular users INTERFERENCE MITIGATION
and derived under what conditions D2D underlay mode can be For an effective D2D communication in a heterogeneous
used by maintaining the channel quality good enough for CUEs. cellular network considering the impact of interference from all
But the main constraints of this proposal is they considered all 3-tiers simultaneously, our techniques will work on two phases.
CUEs and D2D pairs are located in a single cell. Firstly, based on the proposed model, select the optimal mode
of transmission for minimizing interference. In the analytical
In [12], a distributed joint mode selection and power control model, we will first develop a mode selection mechanism
scheme is proposed in an iterative manner by using Monte (Figure-2) and finally construct a resource allocation model to
Carlo method. Based on effective interference on each RB quantify the performance gain for uplink/downlink
(Resource Block), D2D pairs select its mode and adjusts power communication in D2D-enabled multi-tier cellular networks.
to achieve the targeted throughput while CUEs performance
will be unaffected. Like [11], their proposed scheme is only In a multi-tier network, co-tier, and cross-tier both interferences
validated in a single cell but different radius. exist. Therefore, designing the optimal network geometry
By using the real-time information of local traffic channel and model is an important issue to overcome the interference
surrounding nodes, a distributed approach for a smart mode and problem in dense HetNets. For design and analysis of cellular
power selection is proposed in[13]. Here dynamic switching is networks deterministic and stochastic geometry approaches are
adopted between D2D transmissions, when the interference is used. Deterministic methods are suitable only for topologies
limited otherwise continue the communication using cellular with fixed cell sizes not for heterogeneous network topology.
mode. On the other hand, stochastic geometry is a promising approach
to predict probabilistic parameters of randomly designed
To resolve the issues arising from spectrum sharing in D2D HetNets such as SINR (signal-to-interference-plus-noise ratio),
enabled cellular network, a location related strategy for mode transmission power, throughput, coverage probability and load
selection and spectrum sharing algorithm is proposed in[14] distribution [15].
 Start

Which performance has

Cellular D2D
higher priority? Cellular
or D2D ?

Locations of DUEs for

Yes D2D pair are far away
from each other ? No

Locations of DUEs for Explore D2D Resource reuse

CM No Yes
D2D pair close to each Communication direction, UL or DL?
other ?
Resource reuse
direction? UL or DL
UL
No

DL DL
D2D pairs close No Yes D2D pairs Close
to eNB? to CUE UL

SM No D2D pairs Close

No to CUE

Yes DM D2D pairs close

Yes
to eNB?

Yes

Figure 2: Flowchart of the proposed mode selection algorithm for

Figure 3: Interferences in a multi-tier network scenario
D2D communications

In our proposal, we can use the stochastic geometry model to

Therefore, the reliability of link i-j, Prel can be written as
select the appropriate mode of transmission for D2d pairs.
Different properties of a heterogeneous network like device Prel  Pr ( j   th )        (2)
location, the position of macro and small cells, the mobility of Where th is the minimum required SINR for an effective
user etc. are considered as a random (stochastic) process of communication.
specified probability distribution [16]. To determine the spatial
distribution of BS and users, we will use the Poisson Point Moreover, based on the Shannon theory, we can respectively
Process (PPP). derive the achievable data rates of that particular link, expressed
as
r  log 2 (1   )          (3)
A. Analytical Model
For modeling a 3-tier cellular network let us consider a macro In an underlaying heterogeneous cellular network, for an
cell, eNB at the center of the coverage area which is surrounded effective D2D communication we have to fulfill the QoS for
by several small cells. Small cells and UEs are assumed to be D2D link, a cellular link, and link under the small cell services.
randomly deployed in the cellular network. Further assumed In D2D communication, D2D link shares cellular network
that the channel between any two UEs is reciprocal, UEs are resources and small cells also reuse the same resources. So in
limited to half-duplex communication, and all UEs are the service coverage area of macro cell M for any random D2D
synchronized with the macro eNB. Also, assume that all spatial communication, for example, D2D link i-j, received signal will
coexistence transmission links between UEs are mutually be interfered by the transmitting signals from macro eNB, M
independent and path-loss is dominated by multipath fading. and small cell SCeNB, S in a Downlink (DL) sharing model.
Hence, the SINR at the receiver j of the i-j D2D link can be
Therefore, for a random communication link i-j Signal-to- expressed are
Interference-Noise Ratio (SINR), 𝛾 𝑗 for a random receiver j can PD Gij d ij
be expressed as  D
        ( 4)
I M  I S  No
j

Pi Gij d
  ij
       (1)
I  No
j

Where PD is the emission power of D2D link transmitter, IM and

Where α is the path loss coefficient, I is the interference at the IS are the interference due to the power emission of macrocell
receiver j, the distance between the transmitter, i to the receiver, and small cell. 𝐼𝑀 = 𝑃𝑀 𝐺𝑀𝑗 𝑑𝑀𝑗 −𝛼 −𝛼
and 𝐼𝑆 = ∑𝐼∈𝑆 𝑃𝑠𝐺𝑖𝑗 𝑑𝑖𝑗
j is dij, channel coefficient between these two random UEs is
respectively. The transmitted power from eNB and SCeNB are
Gij.
PM and PS respectively.
Here the complex noise at the receiver is modeled as Additive
After substituting these values into the above equation, the
White Gaussian Noise (AWGN) with zero mean and variance
SINR 𝛾𝑗𝐷 at the receiver of the D2D link can be represented as
No, the power of thermal noise density.
PDGij dij
 Dj       (5)

PM GMj d Mj   PS Glj dlj  N o
l S
Thus, the reliability of the D2D link can be written as
TABLE 1. PARAMETERS USED IN THE SIMULATION
D
Prel  Pr ( D j   th )        (6)
Symbol Description Values
Similarly, for a random D2D links M-j in macro cell, D2D pairs α Path loss exponent 4
receiver signal-to-interference-noise ratio, 𝛾𝑗𝑀 can be expressed No Power density spectrum -174dBm
PM Transmit power of eNB 46dBm
as

PM GMj d Mj
PS Transmit power of SCeNB 33dBm
 Dj     (7 ) PD Transmit power of DUEs 13dBm
M 
PDGDj d Dj   PS Glj d lj  N o
l S

Figure 4 represents the CDFs (Cumulative Distribution

Where D2D link transmitter interference for the underlaying Functions) of SINR for different links. We observe that by
𝑀−𝛼 increasing the required SINR decreases the outage probability
macro link, M-j is 𝐼𝐷 = 𝑃𝐷 𝐺𝐷𝑗 𝑑𝐷𝑗
And the reliability of the macro link M  j can be written as
of D2D link sharply after a certain level whereas macro and
small cell links can sustain further which represent the
M
Prel  Pr ( Mj   th )        (8) vulnerability of D2D link compare to cellular links with higher
SINR requirement in a heterogeneous cellular network. Figure
In the heterogeneous cellular network, the link performance for 5 shows that D2D link reliability increases while the distance
all 3-tiers needs to be guaranteed. Let us consider a random UE between D2D pairs and eNB (macro and Small cell base
j is associated with the small cell S0. In that case, the link from station) increases and reciprocal while the distance between
the small cell eNB to this UE S0 →j will be interfered by the D2D pairs increases.
signals from other SCeNBs, 𝑆 ′ = 𝑆/𝑆0 ; eNB M of the macro
cell where S0 belonging to and DUE transmitter of the D2D pair
which is sharing its resources. Therefore, for the small cell, the
SINR at the cellular UE j will be as

PS GS 0 j d S0j
 Sj 
I M  I S'  I D  N o
PS GS 0 j d S0j
  (9)

PM GMj d Mj  P G d
l S
S lj

lj
S 
 PDGDj d Dj  No
l  S0

Hence the reliability of small cell link S0 → j can be expressed

as
S
Prel  Pr ( Sj   th )              10

By analyzing equations (4) to (10) and solving them under PPP Figure 4: SINR for different links in the D2D communications
distribution in stochastic geometry, we can explore at various
conditions which MS is causing minimum interference to the
existing CUEs. We can also derive the distance requirement
criteria for the D2D communication with optimal data rates by
guaranteeing the cellular communication.

Once appropriate communication mode is selected for D2D

communication, co-channel interference will be introduced due
to resource reuse. One can further investigate the interference
strategy by allocating cellular resource efficiently and
guaranteeing QoS for cellular links and to maximize the system
throughput.

IV. RESULTS AND DISCUSSION

To get numerical results, we used a MATLAB tool with the Figure 5: Impact of distance for various nodes in D2D
initial parameters as shown in Table 1. communications
 V. CONCLUSIONS underlaying cellular communications systems,"
In this paper, we proposed a new mode selection algorithm for International Journal of Communications, Network and
D2D enabled heterogeneous 5G cellular networks. These System Sciences, vol. 2, p. 169, 2009.
results are basic outcomes of our analytical model that linking [11] F. Jiang, B. Wang, C. Sun, Y. Liu, and R. Wang, "Mode
the theoretical analysis of D2D communication in 3-tier cellular selection and resource allocation for device-to-device
network. Results obtained have shown that the outage communications in 5G cellular networks," China
probability of D2D links sharply increased while participating Communications, vol. 13, pp. 32-47, 2016.
DUEs are moving farther from each other whereas reciprocal
[12] E. Naghipour and M. Rasti, "A distributed joint power
while D2D pairs are moving away from the cellular base
control and mode selection scheme for D2D
station. An investigation of D2D link reliability and throughput
communication underlaying LTE-A networks," in
is suggested as future work.
Wireless Communications and Networking Conference
(WCNC), 2016 IEEE, 2016, pp. 1-6.
REFERENCES [13] F. Librino and G. Quer, "Distributed Mode and Power
[1] X. Lin, J. G. Andrews, A. Ghosh, and R. Ratasuk, "An Selection for non-Orthogonal D2D Communications: a
overview of 3GPP device-to-device proximity services," Stochastic Approach," IEEE Transactions on Cognitive
IEEE Communications Magazine, vol. 52, pp. 40-48, 2014. Communications and Networking, 2018.
[2] C. V. Networking Index, "Cisco Visual Networking Index: [14] H. Chen, X. Deng, M. Gao, L. Yang, L. Guo, and M. Chi,
Global Mobile Data Traffic Forecast Update, 2016–2021 "Location Related Communication Mode Selection and
White Paper," ed,2017. Spectrum Sharing for D2D Communications in Cellular
[3] H. A. U. Mustafa, M. A. Imran, M. Z. Shakir, A. Imran, Networks," in 2018 International Conference on
and R. Tafazolli, "Separation Framework: An Enabler for Intelligent Transportation, Big Data & Smart City
Cooperative and D2D Communication for Future 5G (ICITBS), 2018, pp. 169-173.
Networks," IEEE Communications Surveys & Tutorials, [15] M. Haenggi, Stochastic Geometry for Wireless Networks:
vol. 18, pp. 419-445, 2016. Cambridge, UK: Cambridge University Press, 2012.
[4] X. Hongnian and S. Hakola, "The investigation of power [16] H. S. Dhillon, R. K. Ganti, F. Baccelli, and J. G.
control schemes for a device-to-device communication Andrews, "Modeling and Analysis of K-Tier Downlink
integrated into OFDMA cellular system," in 21st Annual Heterogeneous Cellular Networks," IEEE Journal on
IEEE International Symposium on Personal, Indoor and Selected Areas in Communications, vol. 30, pp. 550-560,
Mobile Radio Communications, 2010, pp. 1775-1780. 2012.
[5] Z. Liu, T. Peng, S. Xiang, and W. Wang, "Mode selection
for Device-to-Device (D2D) communication under LTE-
Advanced networks," in 2012 IEEE International
Conference on Communications (ICC), 2012, pp. 5563-
5567.
[6] H. ElSawy, E. Hossain, and M. S. Alouini, "Analytical
Modeling of Mode Selection and Power Control for
Underlay D2D Communication in Cellular Networks,"
IEEE Transactions on Communications, vol. 62, pp. 4147-
4161, 2014.
[7] X. Lin, J. G. Andrews, and A. Ghosh, "Spectrum Sharing
for Device-to-Device Communication in Cellular
Networks," IEEE Transactions on Wireless
Communications, vol. 13, pp. 6727-6740, 2014.
[8] E. Frlan, "Direct communication wireless radio system,"
ed: Google Patents, 2000.
[9] C. H. Yu, O. Tirkkonen, K. Doppler, and C. Ribeiro, "On
the Performance of Device-to-Device Underlay
Communication with Simple Power Control," in VTC
Spring 2009 - IEEE 69th Vehicular Technology
Conference, 2009, pp. 1-5.
[10] P. Janis, Y. Chia-Hao, K. Doppler, C. Ribeiro, C. Wijting,
H. Klaus, et al., "Device-to-device communication