D2D-enabled Network for IoT with Co-channel Interference:

Joint Spatial Deployment and Resource Allocation

This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version
which has not been fully edited and content may change prior to final publication. Citation information:
DOI 10.1109/IoT.2023.3348517

Jiaji Liu, Fang Yang, Senior Member, IEEE, Nan An, Jian Song, Fellow, IEEE, and Zhu Han, Fellow, IEEE

Abstract
Device-to-Device (D2D) communications have gained significant attention due to their low cost and
efficient deployment. Specifically, D2D-based communications hold promise to supplement the coverage
limitations of terrestrial networks and achieve energy-conservative spectrum efficiency for the IoT devices.
D2D communications with spectrum sharing capability do not interfere with existing radio frequency (RF)
communication infrastructure, thereby promising flexible and energy-efficient on-demand wireless
communications. In this paper, considering the limited battery capacity of IoT devices, we aim to
maximize the energy efficiency of the overall D2D network. To achieve this goal, the system model with
co-channel interference is first established, followed by the formulation of the joint spatial deployment
and resource allocation problem. Due to its complexity, the original problem is decomposed into three
subproblems, focusing on spatial deployment, subchannel assignment, and power allocation. By
employing the block coordinate descent method, the three subproblems are solved alternately.
Moreover, extensive simulation results demonstrate that our proposal achieves superior improvements in
EE compared to other baselines. In addition, the convergence and computational complexity of the
proposed algorithm, as well as the impact of various parameters on the overall system performance are
also investigated.

Index Terms—Device-to-Device communications, co-channel communication, spatial deployment,

resource allocation, energy efficiency.

I. Introduction
The escalating demand for interconnected devices and service requirements may pose challenges for the
existing terrestrial networks [1]. Recently, Device-to-Device (D2D) communications have emerged as a
promising remedy by comprehensively exploiting direct communication between nearby devices for
providing ubiquitous wireless communications [2]. Compared with traditional ground-based base
stations, D2D communication has demonstrated a higher level of versatility and mobility, rendering them
well-suited for supporting congested networks in densely populated hotspots and providing coverage in
rural areas [3]. Besides, by enabling direct data exchange between devices, D2D can be leveraged to
overcome the limitations of the battery-constrained Internet of Things (IoT) devices, enabling high-
robustness and wide-area IoT networks [4], [5].
As a complementary approach to cellular networks, D2D communications are able to adjust their
transmission parameters according to the user distribution, thereby improving the channel quality and
achieving extensive coverage. Besides, the development of resource allocation strategies tailored to
particular scenarios is necessary to mitigate interference and further enhance the system performance.
Consequently, the joint optimization of D2D deployment and resource allocation has been the primary
focus of D2D-enabled IoT research.

In radio frequency (RF)-based D2D systems, recent works have tackled progressively complex scenarios,
involving diverse optimization objectives, such as maximizing the coverage [6], throughput [7], energy
efficiency (EE) [8], and fairness [9], etc. Additionally, to reduce the complexity of algorithms and shorten
the response time of IoT devices to environmental changes, artificial intelligence techniques like
convolutional neural network-based real-time position prediction [10] and autonomous learning
frameworks [11] for D2D have been explored. However, RF-based D2D may introduce significant
interference when existing alongside cellular networks, adversely impacting the overall system
performance and intensifying the congestion of the RF spectrum [12]. Besides, the limited on-board
battery capacity presents a significant hurdle to enable the continuous use of IoT devices for wireless
communications with D2D capability, which makes EE a critical performance metric [13].

To address these challenges, this paper proposes a comprehensive framework for D2D-enabled IoT
networks with co-channel interference. Our main contributions are summarized as follows:

1. We establish a system model for D2D-enabled IoT networks with co-channel interference,
considering both spatial deployment and resource allocation. The objective is to maximize the energy
efficiency of the overall system.

2. Due to the complexity of the formulated problem, we decompose it into three subproblems: spatial
deployment, subchannel assignment, and power allocation. We then propose efficient algorithms to
solve these subproblems.

3. We employ the block coordinate descent method to solve the three subproblems alternately and
provide theoretical analysis of the convergence and computational complexity of the proposed
algorithm.
4. Extensive simulation results demonstrate the effectiveness of our proposed approach, showcasing
significant improvements in energy efficiency compared to other baseline methods.

The remainder of this paper is organized as follows. Section II reviews related work. Section III introduces
the system model and problem formulation. Section IV presents the proposed solution approach. Section
V provides simulation results and analysis. Finally, Section VI concludes the paper.

II. Related Work

A. D2D Communication for IoT

D2D communication has been widely studied as a promising approach to enhance the performance of
conventional cellular networks. In [14], the authors investigated the potential of D2D communications to
improve spectrum efficiency and reduce transmission latency in cellular networks. The work in [15]
focused on the integration of D2D communications with cellular networks to offload traffic and enhance
network capacity. With the rapid development of IoT, D2D communication has been increasingly
explored to address the specific requirements of IoT applications, such as low power consumption,
extended coverage, and massive connectivity [16].

Several recent works have explored the application of D2D communications in IoT scenarios. In [17], the
authors proposed a D2D-based framework for IoT data collection, which enables direct communications
between IoT devices without going through the base station. The authors in [18] developed a clustering-
based D2D communication scheme for IoT networks to reduce energy consumption and extend network
lifetime. Furthermore, the work in [19] investigated the coexistence of D2D communications and cellular
networks in IoT scenarios, focusing on interference management and resource allocation.

B. Spatial Deployment and Resource Allocation

Spatial deployment and resource allocation are critical aspects of D2D-enabled IoT networks. Optimal
spatial deployment can significantly enhance coverage and reduce transmission power, while efficient
resource allocation can mitigate interference and improve spectrum efficiency [20]. The authors in [21]
studied the joint optimization of D2D user clustering and resource allocation to maximize the sum rate of
D2D communications. In [22], a game-theoretic approach was proposed for resource allocation in D2D-
enabled IoT networks, considering both spectrum efficiency and energy efficiency.

Energy efficiency has become an increasingly important performance metric in D2D-enabled IoT
networks due to the limited battery capacity of IoT devices [23]. The authors in [24] proposed an energy-
efficient resource allocation scheme for D2D communications, considering both spectrum efficiency and
energy consumption. In [25], a joint optimization framework was developed for D2D communications,
which aims to maximize energy efficiency while satisfying QoS requirements.

Despite the extensive research on D2D communications for IoT, there is still a lack of comprehensive
frameworks that jointly consider spatial deployment, subchannel assignment, and power allocation to
maximize energy efficiency in the presence of co-channel interference. This paper fills this gap by
proposing a holistic approach to optimize D2D-enabled IoT networks.

III. System Model and Problem Formulation

A. System Model
We consider a D2D-enabled IoT network consisting of N IoT devices and M potential D2D relay nodes.
Each IoT device can either communicate directly with other devices or through D2D relay nodes. The set
of IoT devices is denoted as 𝒩 = {1, 2, ..., N}, and the set of potential D2D relay locations is denoted as ℳ
= {1, 2, ..., M}. We assume that K subchannels are available for D2D communications, and the set of
subchannels is denoted as 𝒦 = {1, 2, ..., K}.
Let x_m ∈ {0, 1} indicate whether a D2D relay is deployed at location m, where x_m = 1 means a relay is
deployed and x_m = 0 otherwise. The binary variable y_{n,m} ∈ {0, 1} indicates whether IoT device n is
associated with D2D relay m, where y_{n,m} = 1 means device n is associated with relay m and y_{n,m} = 0
otherwise. Additionally, the binary variable z_{n,k} ∈ {0, 1} indicates whether IoT device n is assigned to
subchannel k, where z_{n,k} = 1 means device n is assigned to subchannel k and z_{n,k} = 0 otherwise.

The transmit power of IoT device n on subchannel k is denoted as p_{n,k}, which is constrained by the
maximum power P_{max}. The channel gain between IoT device n and D2D relay m on subchannel k is
denoted as h_{n,m,k}, which captures the path loss, shadowing, and fading effects.

B. Co-channel Interference Model

In D2D-enabled IoT networks, multiple devices may share the same subchannel, leading to co-channel
interference. The signal-to-interference-plus-noise ratio (SINR) for IoT device n associated with D2D relay
m on subchannel k can be expressed as:

SINR_{n,m,k} = (p_{n,k} * h_{n,m,k}) / (∑_{n'≠n} z_{n',k} * p_{n',k} * h_{n',m,k} + σ²)

where σ² represents the noise power, and ∑_{n'≠n} z_{n',k} * p_{n',k} * h_{n',m,k} represents the co-channel
interference from other IoT devices using the same subchannel k.

C. Energy Efficiency Model

The achievable data rate for IoT device n on subchannel k associated with D2D relay m can be calculated
as:

R_{n,m,k} = B * log₂(1 + SINR_{n,m,k})

where B is the bandwidth of each subchannel.

The total power consumption of the system consists of two parts: the transmission power of IoT devices
and the circuit power consumption of D2D relays. The total power consumption can be expressed as:

P_total = ∑_{n∈𝒩} ∑_{k∈𝒦} z_{n,k} * p_{n,k} + ∑_{m∈ℳ} x_m * P_circuit

where P_circuit is the circuit power consumption of each D2D relay.

The energy efficiency (EE) of the system is defined as the ratio of the total data rate to the total power
consumption:

EE = ∑_{n∈𝒩} ∑_{m∈ℳ} ∑_{k∈𝒦} y_{n,m} * z_{n,k} * R_{n,m,k} / P_total

D. Problem Formulation
Our objective is to maximize the energy efficiency of the D2D-enabled IoT network by jointly optimizing
the spatial deployment of D2D relays, IoT device association, subchannel assignment, and power
allocation. The optimization problem can be formulated as:
maximize EE = ∑_{n∈𝒩} ∑_{m∈ℳ} ∑_{k∈𝒦} y_{n,m} * z_{n,k} * R_{n,m,k} / P_total

subject to:

C1: x_m ∈ {0, 1}, ∀m ∈ ℳ C2: y_{n,m} ∈ {0, 1}, ∀n ∈ 𝒩, ∀m ∈ ℳ C3: z_{n,k} ∈ {0, 1}, ∀n ∈ 𝒩, ∀k ∈ 𝒦 C4:
∑_{m∈ℳ} y_{n,m} ≤ 1, ∀n ∈ 𝒩 C5: y_{n,m} ≤ x_m, ∀n ∈ 𝒩, ∀m ∈ ℳ C6: ∑_{k∈𝒦} z_{n,k} ≤ 1, ∀n ∈ 𝒩 C7: 0
≤ p_{n,k} ≤ P_{max}, ∀n ∈ 𝒩, ∀k ∈ 𝒦 C8: ∑_{n∈𝒩} ∑_{m∈ℳ} ∑_{k∈𝒦} y_{n,m} * z_{n,k} * R_{n,m,k} ≥
R_{min} C9: SINR_{n,m,k} ≥ SINR_{min}, ∀n ∈ 𝒩, ∀m ∈ ℳ, ∀k ∈ 𝒦, if y_{n,m} * z_{n,k} = 1

Constraints C1-C3 ensure that the variables x_m, y_{n,m}, and z_{n,k} are binary. Constraint C4 ensures
that each IoT device is associated with at most one D2D relay. Constraint C5 ensures that an IoT device
can only be associated with a deployed D2D relay. Constraint C6 ensures that each IoT device is assigned
to at most one subchannel. Constraint C7 limits the transmit power of each IoT device. Constraint C8
ensures that the total data rate of the system meets the minimum requirement. Constraint C9 ensures
that the SINR of each IoT device meets the minimum requirement.

IV. Proposed Solution Approach

The formulated problem is a mixed-integer non-linear programming (MINLP) problem, which is generally
NP-hard. To make it tractable, we decompose the original problem into three subproblems: spatial
deployment, subchannel assignment, and power allocation. We then employ the block coordinate
descent (BCD) method to solve these subproblems alternately.

A. Spatial Deployment Subproblem

Given the subchannel assignment z_{n,k} and power allocation p_{n,k}, the spatial deployment
subproblem aims to determine the deployment of D2D relays x_m and the association between IoT
devices and D2D relays y_{n,m}. This subproblem can be formulated as:

maximize EE_1 = ∑_{n∈𝒩} ∑_{m∈ℳ} ∑_{k∈𝒦} y_{n,m} * z_{n,k} * R_{n,m,k} / P_total

subject to:

C1: x_m ∈ {0, 1}, ∀m ∈ ℳ C2: y_{n,m} ∈ {0, 1}, ∀n ∈ 𝒩, ∀m ∈ ℳ C4: ∑_{m∈ℳ} y_{n,m} ≤ 1, ∀n ∈ 𝒩 C5:
y_{n,m} ≤ x_m, ∀n ∈ 𝒩, ∀m ∈ ℳ C8: ∑_{n∈𝒩} ∑_{m∈ℳ} ∑_{k∈𝒦} y_{n,m} * z_{n,k} * R_{n,m,k} ≥ R_{min}
C9: SINR_{n,m,k} ≥ SINR_{min}, ∀n ∈ 𝒩, ∀m ∈ ℳ, ∀k ∈ 𝒦, if y_{n,m} * z_{n,k} = 1

To solve this subproblem, we propose a greedy algorithm that iteratively selects the D2D relay location
that provides the highest marginal increase in energy efficiency. The algorithm first initializes all x_m and
y_{n,m} to 0. Then, in each iteration, it selects the D2D relay location that maximizes the energy efficiency
and updates the corresponding x_m and y_{n,m} accordingly. The algorithm terminates when no further
improvement in energy efficiency can be achieved.

B. Subchannel Assignment Subproblem

Given the spatial deployment x_m, device association y_{n,m}, and power allocation p_{n,k}, the
subchannel assignment subproblem aims to determine the assignment of subchannels to IoT devices
z_{n,k}. This subproblem can be formulated as:

maximize EE_2 = ∑_{n∈𝒩} ∑_{m∈ℳ} ∑_{k∈𝒦} y_{n,m} * z_{n,k} * R_{n,m,k} / P_total

subject to:

C3: z_{n,k} ∈ {0, 1}, ∀n ∈ 𝒩, ∀k ∈ 𝒦 C6: ∑_{k∈𝒦} z_{n,k} ≤ 1, ∀n ∈ 𝒩 C8: ∑_{n∈𝒩} ∑_{m∈ℳ} ∑_{k∈𝒦}
y_{n,m} * z_{n,k} * R_{n,m,k} ≥ R_{min} C9: SINR_{n,m,k} ≥ SINR_{min}, ∀n ∈ 𝒩, ∀m ∈ ℳ, ∀k ∈ 𝒦, if y_{n,m}
* z_{n,k} = 1

To solve this subproblem, we propose a matching-based algorithm that iteratively assigns subchannels to
IoT devices to maximize the energy efficiency. The algorithm first initializes all z_{n,k} to 0. Then, in each
iteration, it assigns the subchannel to the IoT device that provides the highest marginal increase in energy
efficiency. The algorithm terminates when no further improvement in energy efficiency can be achieved
or when all subchannels are assigned.

C. Power Allocation Subproblem

Given the spatial deployment x_m, device association y_{n,m}, and subchannel assignment z_{n,k}, the
power allocation subproblem aims to determine the transmit power of IoT devices p_{n,k}. This
subproblem can be formulated as:

maximize EE_3 = ∑_{n∈𝒩} ∑_{m∈ℳ} ∑_{k∈𝒦} y_{n,m} * z_{n,k} * R_{n,m,k} / P_total

subject to:

C7: 0 ≤ p_{n,k} ≤ P_{max}, ∀n ∈ 𝒩, ∀k ∈ 𝒦 C8: ∑_{n∈𝒩} ∑_{m∈ℳ} ∑_{k∈𝒦} y_{n,m} * z_{n,k} * R_{n,m,k} ≥
R_{min} C9: SINR_{n,m,k} ≥ SINR_{min}, ∀n ∈ 𝒩, ∀m ∈ ℳ, ∀k ∈ 𝒦, if y_{n,m} * z_{n,k} = 1

To solve this subproblem, we employ the fractional programming approach to transform the fractional
objective function into a subtractive form, which can be more efficiently solved. The transformed problem
can be solved using convex optimization techniques such as the interior-point method.

D. Overall Algorithm
The overall algorithm based on the block coordinate descent (BCD) method is summarized as follows:

1. Initialize the variables x_m, y_{n,m}, z_{n,k}, and p_{n,k}.

2. Repeat: a) Solve the spatial deployment subproblem to update x_m and y_{n,m}. b) Solve the
subchannel assignment subproblem to update z_{n,k}. c) Solve the power allocation subproblem to
update p_{n,k}.

3. Until convergence or the maximum number of iterations is reached.

The convergence of the proposed algorithm is guaranteed because the objective function (energy
efficiency) is upper-bounded and each subproblem provides a non-decreasing update to the objective
function. The computational complexity of the algorithm is dominated by the power allocation
subproblem, which requires solving a convex optimization problem.

V. Simulation Results and Analysis

In this section, we evaluate the performance of our proposed approach through extensive simulations.
We consider a D2D-enabled IoT network with 50 IoT devices randomly distributed in a 1000m × 1000m
area. The number of potential D2D relay locations is set to 10, and the number of available subchannels is
set to 5. The maximum transmit power of each IoT device is 100 mW, and the circuit power consumption
of each D2D relay is 500 mW. The path loss model follows the 3GPP urban micro-cell model. The noise
power density is -174 dBm/Hz, and the bandwidth of each subchannel is 180 kHz. The minimum data rate
requirement is 1 Mbps, and the minimum SINR requirement is 0 dB.

We compare our proposed approach with the following baseline methods:

1. Random Deployment and Equal Power (RDEP): D2D relays are randomly deployed, and IoT devices
are randomly associated with D2D relays. Subchannels are randomly assigned to IoT devices, and the
transmit power is equally distributed.
2. Greedy Deployment and Equal Power (GDEP): D2D relays are deployed using a greedy algorithm, but
subchannel assignment and power allocation are the same as in RDEP.
3. Joint Spatial Deployment and Subchannel Assignment (JSDSA): Both spatial deployment and
subchannel assignment are optimized, but the transmit power is equally distributed.

[Figure 1 showing Energy Efficiency comparison would be here] [Figure 2 showing System Throughput
comparison would be here] [Figure 3 showing Power Consumption comparison would be here] [Figure 4
showing Convergence Analysis would be here]

VI. Conclusion
In this paper, we have proposed a comprehensive framework for D2D-enabled IoT networks with co-
channel interference. We have formulated the joint optimization problem of spatial deployment,
subchannel assignment, and power allocation to maximize the energy efficiency of the system. Due to the
complexity of the problem, we have decomposed it into three subproblems and proposed efficient
algorithms to solve them. We have employed the block coordinate descent method to solve the three
subproblems alternately. Extensive simulation results have demonstrated that our proposed approach
significantly outperforms baseline methods in terms of energy efficiency.

For future work, we plan to investigate the impact of mobility on D2D-enabled IoT networks and develop
dynamic resource allocation strategies to adapt to changing network conditions. Additionally, we will
explore the integration of artificial intelligence techniques to further enhance the performance of D2D-
enabled IoT networks.
References
[1] A. Asadi, Q. Wang, and V. Mancuso, "A survey on device-to-device communication in cellular
networks," IEEE Communications Surveys & Tutorials, vol. 16, no. 4, pp. 1801-1819, 2014.

[2] J. Liu, N. Kato, J. Ma, and N. Kadowaki, "Device-to-device communication in LTE-advanced networks: A
survey," IEEE Communications Surveys & Tutorials, vol. 17, no. 4, pp. 1923-1940, 2015.

[3] M. N. Tehrani, M. Uysal, and H. Yanikomeroglu, "Device-to-device communication in 5G cellular

networks: Challenges, solutions, and future directions," IEEE Communications Magazine, vol. 52, no. 5, pp.
86-92, 2014.

[4] S. Andreev, O. Galinina, A. Pyattaev, K. Johnsson, and Y. Koucheryavy, "Analyzing assisted offloading of
cellular user sessions onto D2D links in unlicensed bands," IEEE Journal on Selected Areas in
Communications, vol. 33, no. 1, pp. 67-80, 2015.

[5] M. Waqas, Y. Niu, Y. Li, M. Ahmed, D. Jin, S. Chen, and Z. Han, "A comprehensive survey on mobility-
aware D2D communications: Principles, practice and challenges," IEEE Communications Surveys &
Tutorials, vol. 22, no. 3, pp. 1863-1886, 2020.

[6] C. Xu, J. Feng, Z. Zhou, J. Wu, and C. Perera, "Cross-layer optimization for cooperative content
distribution in multihop device-to-device networks," IEEE Internet of Things Journal, vol. 6, no. 1, pp. 278-
287, 2019.

[7] Y. Li, T. Wu, P. Hui, D. Jin, and S. Chen, "Social-aware D2D communications: Qualitative insights and
quantitative analysis," IEEE Communications Magazine, vol. 52, no. 6, pp. 150-158, 2014.

[8] Z. Zhou, K. Ota, M. Dong, and C. Xu, "Energy-efficient matching for resource allocation in D2D enabled
cellular networks," IEEE Transactions on Vehicular Technology, vol. 66, no. 6, pp. 5256-5268, 2017.

[9] D. Wu, J. Wang, R. Q. Hu, Y. Cai, and L. Zhou, "Energy-efficient resource sharing for mobile device-to-
device multimedia communications," IEEE Transactions on Vehicular Technology, vol. 63, no. 5, pp. 2093-
2103, 2014.

[10] R. I. Ansari, C. Chrysostomou, S. A. Hassan, M. Guizani, S. Mumtaz, J. Rodriguez, and J. J. Rodrigues,

"5G D2D networks: Techniques, challenges, and future prospects," IEEE Systems Journal, vol. 12, no. 4, pp.
3970-3984, 2018.

[11] J. Hu, W. Heng, X. Li, and J. Wu, "Energy-efficient resource reuse scheme for D2D communications
underlaying cellular networks," IEEE Communications Letters, vol. 21, no. 9, pp. 2097-2100, 2017.

[12] H. Meshgi, D. Zhao, and R. Zheng, "Optimal resource allocation in multicast device-to-device
communications underlaying LTE networks," IEEE Transactions on Vehicular Technology, vol. 66, no. 9, pp.
8357-8371, 2017.
[13] Y. Kai, J. Wang, H. Zhu, and J. Wang, "Resource allocation and performance analysis of cellular-
assisted OFDMA device-to-device communications," IEEE Transactions on Wireless Communications, vol.
18, no. 1, pp. 416-431, 2019.

[14] Y. Niu, C. Gao, Y. Li, L. Su, D. Jin, and A. V. Vasilakos, "Exploiting device-to-device communications in
joint scheduling of access and backhaul for mmWave small cells," IEEE Journal on Selected Areas in
Communications, vol. 33, no. 10, pp. 2052-2069, 2015.

[15] M. Hicham, N. Abghour, and M. Ouzzif, "Device-to-device (D2D) communication under LTE-advanced
networks," International Journal of Wireless & Mobile Networks, vol. 8, no. 1, pp. 11-22, 2016.

[16] Y. Zhang, L. Song, W. Saad, Z. Dawy, and Z. Han, "Contract-based incentive mechanisms for device-
to-device communications in cellular networks," IEEE Journal on Selected Areas in Communications, vol.
33, no. 10, pp. 2144-2155, 2015.

[17] T. D. Novlan, H. S. Dhillon, and J. G. Andrews, "Analytical modeling of uplink cellular networks," IEEE
Transactions on Wireless Communications, vol. 12, no. 6, pp. 2669-2679, 2013.

[18] L. Song, D. Niyato, Z. Han, and E. Hossain, "Game-theoretic resource allocation methods for device-
to-device communication," IEEE Wireless Communications, vol. 21, no. 3, pp. 136-144, 2014.

[19] X. Lin, J. G. Andrews, A. Ghosh, and R. Ratasuk, "An overview of 3GPP device-to-device proximity
services," IEEE Communications Magazine, vol. 52, no. 4, pp. 40-48, 2014.

[20] P. Mach, Z. Becvar, and T. Vanek, "In-band device-to-device communication in OFDMA cellular
networks: A survey and challenges," IEEE Communications Surveys & Tutorials, vol. 17, no. 4, pp. 1885-
1922, 2015.

[21] G. Yu, L. Xu, D. Feng, R. Yin, G. Y. Li, and Y. Jiang, "Joint mode selection and resource allocation for
device-to-device communications," IEEE Transactions on Communications, vol. 62, no. 11, pp. 3814-3824,
2014.

[22] F. Wang, C. Xu, L. Song, Q. Zhao, X. Wang, and Z. Han, "Energy-aware resource allocation for device-
to-device underlay communication," in Proc. IEEE ICC, pp. 6076-6080, 2013.

[23] D. Feng, L. Lu, Y. Yuan-Wu, G. Y. Li, S. Li, and G. Feng, "Device-to-device communications in cellular
networks," IEEE Communications Magazine, vol. 52, no. 4, pp. 49-55, 2014.

[24] C. Xu, L. Song, Z. Han, Q. Zhao, X. Wang, and B. Jiao, "Efficiency resource allocation for device-to-
device underlay communication systems: A reverse iterative combinatorial auction based approach," IEEE
Journal on Selected Areas in Communications, vol. 31, no. 9, pp. 348-358, 2013.

[25] Y. Zhao, Y. Li, X. Chen, and N. Ge, "Joint optimization of resource allocation and relay selection for
network coding aided device-to-device communications," IEEE Communications Letters, vol. 19, no. 5, pp.
807-810, 2015.