A new spectrum sharing protocol with simultaneous wireless information and power transfer (SWIPT) is proposed to cope with the increasingly prominent problem of spectrum and energy scarcity. It operates within a cognitive radio network (CRN) in the context of cellular IoV (C-IoV), enabling bidirectional communication between two vehicles parked within the base station coverage (VnBSs) while facilitating cooperation for a pair of primary users (PUs), i.e., VnBSs can act as relays to provide cooperation communication for the cell–edge vehicle user (eVU). Unlike most existing work, both VnBSs can use time switching (TS) to obtain energy from radio frequency (RF) signals emitted from the base station. In order to enhance the reliability of the network, this study incorporates the automatic repeat request (ARQ) technique in the CRN supported by the nonorthogonal multiple access (NOMA) and SWIPT, which has not been performed in other works. Based on this, the transmission is divided into one energy harvesting (EH) phase and three information processing (IP) phases. A new packet for PUs is transmitted in the first IP phase and is allowed to be retransmitted twice in the last two IP phases depending on the decoding. VnBSs act as relays to obtain energy in the EH phase to assist in retransmitting the PU’s packets and sending their own packets in the last two IP phases. The system states are analyzed by building a one-dimensional Markov chain, and the end-to-end outage probability (OP) is calculated for each state under the Nakagami-m fading channel. Using these two results, the OP of the primary and secondary networks, system throughput and energy efficiency (EE) are derived. Finally, the validity of the derived results is verified by Monte Carlo simulation using MATLAB and compared with the protocol without ARQ, and the protocol proposed shows a better performance.

1. Introduction

In the fifth generation of wireless networks, the internet of things (IoT) spectrum sharing technology is a very promising technology that enables automated devices to establish communications for IoT applications within the authorized spectrum [1]. IoT integrates a variety of physical entities, including smartphones, smart furniture, sensors, and vehicles. [2]. They are connected through communication networks to facilitate the exchange of information between themselves and their surroundings. The IoT is consolidated IoV [3], a communication system that integrates vehicles with people, roadside units, intelligent sensors, and other networks [4]. To enhance user experience and service quality, vehicles must gather and analyze intricate traffic information in real time, even from considerable distances. This presents a significant challenge for energy-constrained and distance-limited cell–edge vehicle users (eVUs). On the other hand, with the progress of society and the development of science and technology, the cost of automobile manufacturing is gradually reduced, and the number of vehicle users is increasing explosively, which makes the limited road and spectrum resources more tense [5]. This urges us to study and design new spectrum sharing protocols to further improve the utilization of available resources.

To address the above issues, we conduct an in-depth analysis of the existing work on the background of IoV and the existing work on IoT general scenarios that can be applied to IoV. The results prove that cognitive radio (CR), nonorthogonal multiple access (NOMA), and simultaneous wireless information and power transfer (SWIPT) technologies can provide effective solutions to improve the spectrum efficiency (SE) and energy efficiency (EE) of IoT/IoV. The authors in reference [6] proposed a cluster-based CR–assisted vehicular network (CRAV-Net) to improve SE and congestion perception. In [7], the authors proposed two cooperative spectrum sharing protocols. The protocol used a mixture of underlay and overlay to provide secondary users (SUs) more opportunities to access the spectrum, thereby improving spectrum utilization. In addition, they enhanced system reliability through automatic repeat requests (ARQs). While the scheme in [8] greatly increases the system throughput by combining CR and NOMA techniques, the authors in [9] studied the performance of the overlay cognitive NOMA network under the Rayleigh fading channel and compared it with the performance of the overlay cognitive OMA network and underlay cognitive NOMA network. The results show that the overlay cognitive NOMA network performs the best. In addition, the integration of SWIPT in NOMA systems has been recognized as a potential paradigm for enhancing the lifetime of energy-constrained IoT/IoV networks [10]. In [11], the authors proposed a SecGreen scheme with an efficient energy harvesting (EH) algorithm. The SecGreen scheme maximizes EE by using SWIPT. The results show that the EE of the proposed scheme is improved by 22.5% and 20.34%, respectively, compared to the other two schemes. In [12, 13], the authors investigated the system of SWIPT technology integrated with cooperative NOMA. They utilized dedicated relays and user relays to facilitate communication for distant users, aiming to enhance the performance of remote users. In addition, the integration of SWIPT with NOMA technology exhibits versatility in its applicability to various communication scenarios. In the unmanned aerial vehicle (UAV) communication scenario of [14], the authors studied two cooperative NOMA communication networks based on time switching (TS) and power splitting (PS), respectively. The authors in [15] applied SWIPT and cooperative NOMA techniques to heterogeneous networks (HetNets). Unlike other literature, they considered not only superimposed signals from macrobase stations (MBSs) but also interfering signals from neighboring small-base stations (SBSs) for EH with good results. In addition, in order to solve the interruption and fair data rate problems of cell–edge users, the authors in [16] investigated the performance of cell–edge users in multiple input single output NOMA (MISO–NOMA) systems and proposed a SWIPT protocol that hybridizes TS and PS for powering the relay. In [17], the authors implemented ARQ technology in the NOMA network with the EH relay for the first time, enhancing system reliability. In addition to integrating the two technologies, some experts have begun to study the organic combination of the three technologies. The authors in [18] researched a cooperative spectrum sharing transmission scheme with EH customized for IoT applications. In the protocol described in [19], the primary transmitter (PT) leverages the secondary transmitter (ST) serving as an IoT relay to facilitate communication with users experiencing poor signal reception. In a cooperative network with SWIPT, the CR relay uses a PS scheme to obtain transmit power from the ST. In the protocol discussed in [20], the SWIPT–NOMA strategy in cooperative CR networks (CRNs) introduces a minor degradation in overall outage performance relative to similar protocols utilizing dedicated batteries, while maintaining the same diversity order.

It is found that there is little work to consider the application of CR, NOMA, and SWIPT technologies in IoV to vehicle communication. While there are many related works in IoT that can realize direct communication between users under the architecture of device-to-device (D2D) communication [21], and IoV is implemented based on IoT, devices in IoT also include mobile vehicles [22]. Therefore, when analyzing the related work aiming at improving the spectrum utilization and data confidentiality of vehicle communication, the related work of IoT general scenario suitable for IoV is also considered. For example, a bidirectional DF strategy based on two-phase network coding has been proposed in [23]. In [24], the authors investigated a scenario in which two IoT devices act as relays to assist a pair of primary users (PUs) in CRNs. The IoT relay device harvests energy using the TS protocol and subsequently utilizes the NOMA protocol to exchange its own information while forwarding the PU’s information. The authors in [25] analyzed the performance of a two-way relay system using SWIPT under Nakagami-m fading channel. The bidirectional communication between two nodes is accomplished through three phases, where the relay relies on the PS strategy to EH in the first two phases and amplify-and-forward (AF) the information from the node in the third phase. The authors of [26] designed a new SWIPT scheme with hybrid PS and TS to achieve bidirectional communication between PUs and SUs through four phases. These literature give useful references in terms of protocol design regarding the fusion of the three techniques in the context of vehicle communication. In [27], a Lagrangian dyadic algorithm based on a first-order Taylor series expansion function is proposed to solve the transmit power optimization problem to maximize the EE of SU. To achieve in-band bidirectional D2D/vehicle-to-vehicle (V2V) communication, the authors of [28, 29] utilize CR to enable devices to share the licensed spectrum of the PU and rely on the SWIPT technique to power the devices. The simulation results in [28] illustrate the SE and EE advantages of this system over unidirectional relaying. Simulation results in [29] show that the system achieves about 68% and 277% gains in SE and EE, respectively, compared to another spectrum sharing system for two-way communication. Similarly, in [30], the authors examined the performance of AF and DF relaying techniques in networks supporting D2D communications using CR and SWIPT methods. The results show that the proposed spectrum sharing protocol improves the performance of DF relaying and AF relaying schemes by 209% and 49%, respectively. Therefore, the collaboration of D2D communication users as relays for data transmission can significantly enhance system performance by sharing the bandwidth utilized by cellular base stations [21]. In addition, the authors in [31] have also proved that using roadside parked vehicles as relays within cellular networks, that is, by leveraging the vehicle’s onboard connectivity capabilities and turning them (when parked) into vehicle relay nodes, can monitor operating costs and increase network coverage and capacity. As a result, it is feasible to use parked vehicles with wireless functions as relays, and it is more capable of overcoming the exponential growth of user data requirements than traditional fixed-base deployments.

As mentioned above, considering CRNs with SWIPT two-way V2V relay cooperation in the context of cellular IoV (C-IoV) has significant implications for SE and EE improvement. However, the OP increases in [20]. Furthermore, most studies are conducted under Rayleigh channels, and none of them has applied the ARQ technique to SWIPT–enabled CR–NOMA networks. Therefore, considering the key constraints of C-IoV deployment, this study proposes a DF–based bidirectional cooperative CRN that leverages the benefits of NOMA and SWIPT technologies, as shown in Figure 1. In this scenario, a pair of vehicles parked within base station coverage (VnBSs) harnesses energy from radio frequency (RF) signals using the TS technique. The harvested energy is then employed to facilitate relay cooperation with PU, while also supporting its own information exchange requirements. To incorporate ARQ, the information processing (IP) phase is divided into three parts and the PU is fixed to transmit new packets using the IP-1 phase. If PU successfully decodes, VnBSs are granted exclusive spectrum access for subsequent information exchange in the later phase. However, if PU fails to decode, VnBSs are prioritized for retransmission on behalf of PU and can simultaneously transmit their own information using the NOMA technique. Hence, in contrast to the cooperative scheme lacking ARQ, the protocol presented in this paper not only addresses the reliability concerns of PUs but also enhances spectrum access for VnBSs, thereby improving VnBSs′ communication quality while ensuring PU’s quality of service (QoS). Ultimately, due to its comprehensive nature and compatibility with various experimental data, we opt to employ the Nakagami-m fading channel model. Thus, in contrast to [9], the Nakagami-m fading channel model used in this paper offers a more realistic representation than the Rayleigh fading model. In addition, the CR–NOMA framework discussed here supports the integration of SWIPT and ARQ technologies, which is not addressed in [9]. Since the network architecture discussed in [9] is a typical CR system with a PUs pair and a SUs, it is consistent with the network architecture studied in this paper. Therefore, during the application of the above two technologies, the spectrum sharing scheme adopted by VnBSs as a cooperative relay to assist PU transmission utilizes the results of [9], i.e., the overlay spectrum sharing scheme is adopted. Thus, the performance of this scheme is compared with that of [9] during the numerical simulation process.

Details are in the caption following the image — **Figure 1**
Open in figure viewer PowerPoint

V2V communication in C-IoV.

In summary, the contributions of this paper are as follows:

•
A DF–based bidirectional cooperative CRN model is established that utilizes NOMA and SWIPT technologies for a new spectrum sharing protocol, where VnBSs harvest energy from RF signals using the TS technique to facilitate relay cooperation and conduct their own information exchange, thereby improving SE and EE.
•
The ARQ technique is applied to the cognitive NOMA network supported by SWIPT. Based on the feedback from PR and VnBSs, the system can employ varied retransmission strategies. To facilitate communication between primary and secondary systems, the transmission is divided into four phases, comprising one EH phase and three IP phases. If PU successfully decodes its own information in the IP-1 phase, VnBSs can exclusively utilize the remaining two phases for their own information exchange. If PU fails to decode, retransmission of PU’s information will occur in the subsequent two phases following a specific order: I VnBS₁, VnBS₂, and PT based on VnBSs′ decoding status until both retransmission opportunities are exhausted. This scheme enhances spectrum access opportunities for SUs, optimizes spectrum utilization, and bolsters the reliability of the primary network.
•
A one-dimensional Markov chain is established to analyze the system’s performances by examining the system state change. We derive the end-to-end outage probability (OP) and utilize it, along with the Markov chain, to obtain the OP, system throughput, and EE of both the primary and IoV systems. Ultimately, numerical simulations of the system performances demonstrate the impact of certain parameters. Furthermore, compared to the protocol without ARQ, our proposed protocol exhibits superior performance. It addresses the reliability issue of PUs and provides additional spectrum access opportunities for VnBSs, ensuring high-quality PU service and enhancing communication quality for VnBSs.

The remaining components of the paper are as follows: in Section 2, the system model and protocol are shown. Through the description in this section, the transmission process and the state transition relationship can be clearly understood. In Section 3, the system performances are analyzed. In order to analyze the system state, a one-dimensional Markov chain is also established in this section. Numerical simulations are presented in Section 4. Finally, the study is concluded in Section 5.

2. System Model and Protocol Description

Figure 2 depicts the bidirectional CR system supported by NOMA and SWIPT considered in this paper. PUs comprise a base station serving as the PT and a eVU functioning as the PR. In particular, cell–edge users are those located far from the base station, positioned at the edge of the network coverage area. Therefore, in this paper, a mobile vehicle with the above characteristics is regarded as eVU, and the influence of distance and multipath propagation effect on signal strength is characterized by channel fading. An eVU obtains traffic information through the base station to avoid traffic congestion and traffic accidents, so as to make early lane changes or other driving decisions. However, the direct link between the base station and eVU is hindered by factors such as distance and fading. A pair of close VnBSs in the cell act as relays to assist the PU in retransmission. By aiding services to the PU, VnBS₁ and VnBS₂ can complete mutual communication on the PU’s licensed spectrum. All participating nodes, including the primary system and the IoV system, are equipped with a single antenna and operate in half-duplex.

In order to complete the communication between the primary system and the IoV system and consider the power supply of the relay, the TS–based SWIPT technology is used. The transmission process is divided into two phases: EH and IP, with the IP phase further segmented into three phases, as depicted in Figure 3. In the EH phase, unlike other literature, both VnBSs are capable of harvesting energy from the RF signal transmitted by PT in order to facilitate the transmission of PU and their own information in the subsequent phase. In the IP phase, ARQ is incorporated into this protocol to enhance the transmission reliability of PU, allowing for up to 2 retransmissions by PU while prohibiting any retransmissions by VnBSs. In the IP-1 phase, PT broadcasts the primary signal. Based on the feedback from PR and VnBSs (ACK for successful decoding and NACK for unsuccessful decoding), the system switches to the appropriate state at the IP-2 phase. If PR decodes the primary information successfully, VnBSs exclusively enjoy the next two phases to complete the information exchange under interwave mode. If PR fails to decode, VnBSs will assist PU to complete information exchange under overlay mode. Since VnBS₁ wants to communicate with VnBS₂ first, VnBS₁ is used as the preferred relay. If the VnBS₁ succeeds in decoding the primary information in the IP-1 phase, the PU’s signal is retransmitted by VnBS₁; otherwise, it is retransmitted by VnBS₂. If none of the VnBSs decode the primary information successfully, PT retransmits by itself. According to the feedback of PR in the IP-2 phase, the system is transferred to the corresponding state in the IP-3 phase. If the first retransmission of PR is successful, the system continues to complete the transmission of IoV. If the first retransmission of PR is unsuccessful, the second retransmission is performed in the IP-3 phase in the order of retransmission.

Having established the system model as outlined above, we will proceed to present the channel model and subsequently provide a detailed description of the protocol based on it. The specific system states and the transition relationships between states will also be analyzed in Section C (IP phase), while the signal-to-interference and noise ratio (SINR) of the corresponding link is calculated for the analysis of the system performance in Section 3.

2.1. Channel Model

In the following, we assume that all the channels are independent and identically distributed Nakagami-m fading channels. The channel coefficients of link PT–PR, PT-–VnBS₁, PT–VnBS₂, VnBS₁–VnBS₂, VnBS₂–VnBS₁, VnBS₁–PR, and VnBS₂–PR are h_pr, h_p1, h_p2, h₁₂, h₂₁, h_1r, and h_2r, respectively, and h_ij ~ CN(0, Ω_ij), i ∈ {p, 1, 2}, j ∈ {r, 1, 2}, i ≠ j, where

is the average channel power, d_ij is the distance between nodes i and j, and the path fading coefficients are all given by α_ij = α₀. The PDF and CDF of

can be expressed as

()

where θ_ij = Ω_ij/m_ij denotes the scale parameter; m_ij represents the shape parameter; and Γ(·), Γ(·, ·) , and Υ(·, ·) show the complete, the upper incomplete, and the lower incomplete gamma functions, respectively. For any i ∈ {t, 1, 2}, j ∈ {r, 1, 2}, i ≠ j, we assume the shape parameter m_ij = m. In addition, each receiving node generates additive white Gaussian noise (AWGN), subjecting to CN(0, σ² = 1), and the average power of the signal is 1. Without loss of generality, we consider that

and

, i ∈ {1, 2}.

2.2. EH Phase

In this part, the TS strategy in the SWIPT technique is employed for EH. In the TS–based EH strategy, the transmission is divided into two parts by the TS factor α, where αT is used to complete the transmission in the EH phase and (1 − α)T is used to complete the transmission in the IP phase. In the EH phase, both VnBS₁ and VnBS₂ can harvest energy from the RF signal broadcast by PT for their own subsequent signal emission. It is assumed that VnBSs reset their energy after a transmission regardless of whether they participate in the transmission or not and collect new energy in the EH phase of the next transmission. The energy collected by VnBS₁ and VnBS₂ during the EH phase is denoted by

()

where i ∈ {1, 2}, P_p denotes the transmit power of PT, η represents the energy conversion efficiency, and α indicates the TS factor. The transmit power of VnBS₁ and VnBS₂ can be expressed as

()

where i ∈ {1, 2}, β = 3αη/(1 − α).

2.3. IP Phase

In order to accomplish the communication of the PU and the bidirectional communication between the VnBSs in a single transmission, the IP phase is divided into three parts. The PU fixedly occupies the IP-1 phase to broadcast the primary signal, and the system will turn to different states according to the feedback from PR and VnBSs. In the next two phases, VnBSs act as relays to complete the retransmission of the primary system and the communication of the IoV system at the same time, or the primary signal is retransmitted by PT. According to the SWIPT transmission structure in the proposed protocol shown in Figure 3, the description of the specific transmission process of the system and the division and composition of each state are shown in Table 1.

Table 1. Transmission state transition of the proposed protocol.

IP-1 (S₁: PT ⟶ PR, VnBS₁, and VnBS₂)			IP-2	IP-3
PR	VnBS₁	VnBS₂	IP-2	IP-3
ACK	ACK/NACK	ACK/NACK	S₂: VnBS₁ ⟶ VnBS₂	S₃: VnBS₂ ⟶ VnBS₁

NACK	ACK	ACK	S₄: VnBS₁ ⟶ VnBS₂ and PR	S₃
	ACK	ACK	S₄: VnBS₁ ⟶ VnBS₂ and PR	S₅: VnBS₂ ⟶ VnBS₁ and PR
	ACK	NACK	S₆: VnBS₁ ⟶ VnBS₂ and PR	S₃
	ACK	NACK	S₆: VnBS₁ ⟶ VnBS₂ and PR	S₇: PT ⟶ PR (retransmission)
	NACK	ACK	S₈: VnBS₂ ⟶ VnBS₁ and PR	S₉: VnBS₁ ⟶ VnBS₂
	NACK	ACK	S₈: VnBS₂ ⟶ VnBS₁ and PR	S₇
	NACK	NACK	S₁₀: PT ⟶ PR (retransmission)	S₉
	NACK	NACK	S₁₀: PT ⟶ PR (retransmission)	S₇

In the IP-1 phase, the system is in State S₁ (S₁: PT ⟶ PR, VnBS₁, and VnBS₂), and the EH phase is also included in the State S₁, and at this time, the PT broadcasts the primary signal x_p to PR, VnBS₁, and VnBS₂. The signal they received can be expressed as

()

where i ∈ {1, 2, r}. After transmitting the signal, PT will listen to the feedback from PR, VnBS₁, and VnBS₂. It will acknowledge successful decoding with a feedback ACK and indicate unsuccessful decoding with a feedback NACK. Currently, VnBSs acting as relays also listen to the feedback of the PR. The SINR at decoding is given by

()

where i ∈ {1, 2, r}, ρ_p = P_p/σ².

After the IP-1 phase, two cases of successful and unsuccessful PR decoding are generated. When PR fails to decode x_p, the retransmission conditions of the IP-2 and IP-3 phases can be divided into four cases according to the decoding conditions of VnBS₁ and VnBS₂. To sum up, there are five outcomes after the IP-1 phase as follows:

1.
PR decodes x_p successfully in IP-1 phase. VnBSs exchange their information in the next two phases. Then, the system first enters the State S₂ (S₂: VnBS₁ ⟶ VnBS₂) in IP-2 phase, and VnBS₁ sends the signal x₁₂ to VnBS₂, and the signal received by VnBS₂ is denoted by
()
where P₁ is the transmit power of VnBS₁ given by equation (4). The decoded SINR at VnBS₂ can be expressed as
()
where ρ₁ = P₁/σ². Then, the system will enter the State S₃ (S₃: VnBS₂ ⟶ VnBS₁), and after this state with probability one,VnBS₂ sends a signal x₂₁ to VnBS₁. The signal received by VnBS₁ is expressed as
()
where P₂ is the transmit power of VnBS₂, which is given by equation (4). The decoded SINR at VnBS₁ is denoted by
()
2.
PR fails to decode x_p in the IP-1 phase, but all the VnBSs decode x_p successfully. VnBSs utilize NOMA technology to transmit their own signal while assisting PU in retransmission. As VnBS₁ is the initiating party for communication, it is specified that if VnBS₁ successfully decodes x_p, it will transmit it first; otherwise, VnBS₂ will retransmit it. If VnBS₁ successfully decodes x_p, the system will transition to State S₄ (S₄: VnBS₁ ⟶ VnBS₂ and PR), and VnBS₁ will transmit the superimposed signal using NOMA. Due to the close proximity of VnBSs, a₂ > a₁, resulting in a higher power allocation to the PU’s signal, the retransmission signal received by the PR from VnBS₁ is denoted as
()
Since the PU’s signal is a strong power signal, PR will treat the VnBSs′ signal as interference and decode its own signal directly, and the decoding SINR is represented as
()
The signal received by VnBS₂ is expressed as
()
Since the signal of VnBS₂ is a weak power signal, it must use successive interference cancellation (SIC) to decode the signal of the PU first, and then decode its own signal after removing it. The SINR at VnBS₂ is given by
()

()
The system will enter State S₃ if this retransmission is successful; otherwise, it will enter State S₅ (S₅: VnBS₂ ⟶ VnBS₁ and PR).
Since the first retransmission of VnBS₁ fails in State S₄, the second retransmission is performed by VnBS₂ in State S₅. VnBS₂ uses NOMA to send a superimposed signal , where b₂ > b₁, and the signal received by the PR can be represented as
()
The decoding process is similar to the case of State S₄, except that the roles of VnBS₁ and VnBS₂ are swapped. PR can directly decode its own signal, and the decoding SINR is expressed as
()
The signal received by VnBS₁ is denoted by
()
SIC is used at VnBS₁ to decode the signal of PU first, and then decode its own signal after removing it, and the decoded SINR is denoted as
()

()
3.
PR fails to decode x_p in IP-1 phase, VnBS₁ decodes it successfully but VnBS₂ does not. The successful decoding of VnBS₁ triggers its retransmission in the IP-2 phase, placing the system in State S₆ (interpreted as State S₄). If this retransmission is successful, the system progresses to the IP-3 phase. However, if VnBS₂ fails to decode, it must be retransmitted by the PT itself and the system enters State S₇ (analyzed in the same way as state S₁).
4.
PR fails to decode x_p in IP-1 phase; VnBS₂ decodes it successfully while VnBS₁ fails to decode. The retransmission is initiated by VnBS₂ during the IP-2 phase, while simultaneously transmitting its own information in State S₈ (similar to the analysis of State S₅). The initial retransmission is successful, transitioning the system into State S₉ (similar to the analysis of tate S₂). However, the first retransmission fails and the system enters State S₇, which is retransmitted by PT itself.
5.
PR fails to decode x_p in the IP-1 phase and none of the VnBSs decodes successfully. In the IP-2 phase, PU autonomously performs the initial retransmission, while the system is in State S₁₀ (S₁₀: PT ⟶ PR and retransmission). The situation of the IP-3 phase is the same as that of IP-3 in 4, and the system will enter State S₉ if the first retransmission is successful; otherwise, it will enter State S₇.

There exist some states with the same transmission among the above transmission states shown in Table 1, so their analyses are omitted from the results. However, they still cannot be regarded as the same state, such as States S₆ and S₄, State S₈ and S₅, and States S₇, S₁₀, and S₁. Since being in different IP phases also means being in different retransmission states, their future state transfer results are distinct. On the other hand, they also cannot be considered as the same state for the sake of building a one-dimensional Markov chain correctly and simply.

Based on the above state transfer analysis and state space {S₁, S₂, ⋯, S₁₀}, we get the following state transfer diagram (Figure 4), which establishes the Markov chain of the protocol proposed, and the flowchart of its dynamic change process is shown in Figure 5.

The state transfer probabilities in Figure 4 can be expressed as

()

where

, and

denote the end-to-end OP of PT–PR, VnBS₁, and VnBS₂ under State S₁; VnBS₁–PR under State S₄; and VnBS₂–PR under State S₅, respectively, which will be calculated in the next section. The above statement leads to the following conservation equation:

()

where π_i, i ∈ {1, 2, ⋯, 10} denotes the steady-state probability that the system is in State S_i. Furthermore, according to

, the steady-state distribution of the system can be solved as follows:

()

3. Performance Analysis

In this section, the performance of the implemented protocol is analyzed. The end-to-end OP in different states will be examined based on the transmission process and system state described in the previous section. Furthermore, utilizing the steady-state probabilities obtained earlier, the OPs of both primary and secondary systems can be calculated. In addition, by analyzing the OPs of primary and SUs, the system throughput and EE will be computed.

3.1. OP Analysis

OP is a key performance metric for assessing the link reliability of wireless systems operating over fading channels. It is defined as the probability that the instantaneous achievable rate falls below a specified target rate at any given time. First, the end-to-end OP will be calculated in each state, and then the OPs of the primary and SUs will be calculated. The probability that PT–PR is interrupted in State S₁ is expressed as

()

Let R_p be the target rate for decoding the primary signal,

. Similarly, the OPs of PR–VnBS₁ and VnBS₂ are represented, respectively, as

()

The OP of VnBS₁–VnBS₂ in State S₂ is expressed as

()

Let R_s be the target rate for decoding the secondary signal,

. The following result can be computed based on the proof from equations (36) to (41) in the Appendix:

()

where K_ν(·) represents the modified Bessel function of the second kind of ν order. Similarly, the OP of VnBS₂–VnBS₁ in State S₃ can be calculated as follows:

()

The OP of VnBS₁–PR in State S₄ can be expressed as

()

where μ₁ = a₂ − a₁γ_p. However, VnBS₂ is required to decode the PU’s signal first and then decode its own signal. Therefore, successful decoding can only occur if both steps are successful, leading to its OP being denoted as

()

where τ = max{(γ_p/μ₁βρ_p), (γ_s/a₁βρ_p)}.

Similar to the analysis in State S₄, the OP of VnBS₂–PR in State S₅ is denoted by

()

The OP of VnBS₂–VnBS₁ is expressed as

()

Based on the end-to-end OP obtained as above and the steady-state probability of the system state obtained in the previous section, the OPs of the primary system and the IoV system can be calculated.

1.
The OP of the primary system is as follows:
()
2.
The OP of the IoV system is as follows:
()

3.2. System Throughput

In the context of delay-limited scenarios, the system throughput for the considered CRN is defined as the aggregate of average target rates achieved by successful transmissions of both primary and IoV systems over fading channels. Therefore, the system throughput can be expressed as

()

where

and

represent the OP of primary and IoV systems, respectively, which are given by equations (35) and (36).

3.3. EE

In accordance with the classical definition, the EE of the EH–based CR system is quantified as the ratio of the aggregate data transmitted to the total energy expended. Hence, the system’s EE can be mathematically represented as

()

where S_T is the system throughput given by equation (36).

4. Simulation Result

In this section, Monte Carlo simulation is used to verify the accuracy of the analysis results. In order to discuss the effects of system parameters on OP, system throughput, and EE, the performance analysis of the proposed protocol is based on equations (35)–(37). In addition, in order to further highlight the advantages of the introduction of ARQ to improve system performances, a comparative analysis is made with the spectrum sharing protocol without ARQ in [9]. Therefore, the experimental parameters are set with reference to [9], as shown in Table 2.

Table 2. Parameter table.

Parameter	Value
AWGN power	σ² = 1
Target rate	R_p = R_s = 0.5
The power allocation coefficient of the NOMA scheme	a₁ = b₁ = 0.2, a₂ = b₂ = 0.8
Energy conversion efficiency	η = 0.7
Time switching coefficient	α = 0.1
Distance between nodes	d₁₂ = d₂₁ = 0.51, d_pr = 1, d_p1 = d_p2 = 0.8
Path fading coefficient	α₀ = 4

The comparison curves of PU in the protocol proposed in this paper and the protocol described in [9] are presented in Figure 6. Similarly, Figure 7 illustrates the comparison curves of SU in the protocol proposed in this paper and the protocol mentioned in [9]. It can be observed from the Figures 6 and 7 that the analysis results are basically consistent with the simulation results, thus verifying the effectiveness and accuracy of the analysis results. The analysis of the comparison curves in Figure 6 reveals a notable trend: as the transmission SNR increases, the OP for both PU and SU decreases. Specifically, it is evident from the Figure 6 that under the proposed protocol, the OP of PU is significantly lower than that of PU under the protocol [9], which does not incorporate ARQ technology. This discrepancy underscores the distinct advantages offered by employing ARQ techniques in enhancing system performance and reliability. In addition, the protocol described in this paper enables the PU to engage in up to two retransmissions, thereby significantly enhancing transmission reliability.

The analysis of Figure 7 reveals that the OP of SU, denoted as VnBSs, is lower under the proposed protocol compared to the SU’s OP in the protocol presented in [9]. This discrepancy can be attributed to the fact that in [9], the SU has an opportunity to transmit its own information while acting as a relay for PU only upon successful decoding of PU information. In this paper, VnBSs are employed to transmit their own information in the phase subsequent to successful transmission by PU. In addition, the study introduces a two-way relay system, effectively equivalent to having two relays. Following successful assistance in PU retransmission by one of the relays, the other is granted exclusive access to the spectrum irrespective of its success in decoding PU information. This approach enhances overall system efficiency and reliability while optimizing spectrum utilization. It is also evident that the OP of VnBS₁ is marginally lower than that of VnBS₂ at low transmission SNR. This discrepancy arises from the protocol’s specification that VnBS₁, as the initiating party for communication, has greater access opportunities to the spectrum as the preferred relay compared to VnBS₂. However, as transmission SNR increases, this gap gradually diminishes, indicating a relatively minor impact.

Figures 8 and 9 depict the relationship curves between transmission SNR and system throughput, considering varying target rates of R_p and R_s. These rates correspond to the fading parameters m = 1 and m = 1.5 of the Nakagami-m fading channel model. Here, R_p and R_s specifically represent the targeted data rates for PU and IoV users, respectively. As the transmission SNR increases, the system throughput increases continuously and starts to level off at ρ_p = 30 dB. In accordance with equation (36), the gradual reduction in OP is directly associated with a continuous increase in system throughput. This observed trend tends to stabilize at a specific point, known as the saturation point, which signifies the maximum achievable throughput under the given set of parameters. However, as the target rate increases, the system throughput increases, and it is more obvious at high SNR. According to equation (36), the increase in the target rate corresponds to the increase in the system throughput. Thus, as expected, the system throughput is basically consistent with the simulation results. As shown in Figures 8 and 9, when m = 1, R_p = R_s is 0.4, 0.5, and 0.6, respectively, and the maximum achievable throughput is about 0.8089, 0.8986, and 0.9884, respectively; when m = 1.5, the maximum achievable throughput is about 0.8252, 0.9190, and 1.0128, respectively. It is apparent that the system throughput at fading parameter m = 1 is lower than that at m = 1.5. This discrepancy can be attributed to the characteristics of the Nakagami-m fading channel model, where a larger value of the fading parameter m indicates reduced fading and improved channel conditions. Therefore, in this context, the higher system throughput observed at m = 1.5 compared to m = 1 reflects the influence of these differing channel conditions on overall system performance.

Figure 10 provides a visual representation of the relationship between EE and transmit signal-to-noise ratio (SNR) for different target rates. The Figure 10 reveals that the system exhibits significant EE at low SNR levels and small target rates. However, as both the transmit SNR and target rate increase, there is a noticeable decrease in EE. Interestingly, it is also observed that an increase in the target rate leads to an enhancement in the maximum achievable EE by the system. This suggests that while higher SNR and target rates may lead to reduced EE, increasing the target rate can potentially improve overall system performance in terms of EE.

5. Conclusions

In this paper, we propose a bidirectional V2V communication scheme based on NOMA and SWIPT. Using CR technology, VnBSs can exchange information on the PU’s licensed spectrum alone or by assisting the PU. Second, the existing literature commonly increases the number of PU signal copies by allowing SU to transmit superposition signals containing both PU and its own information using NOMA. Subsequently, maximal ratio combining (MRC) or selection combining (SC) is employed at the receiver to enhance the reliability of PU transmission. However, this method results in low transmission quality for SU and introduces a certain degree of data redundancy, leading to unnecessary overhead. Hence, this study employs ARQ technology to significantly enhance the service quality of SU while ensuring successful PU transmission. Subsequently, we examine OP expressions for end-to-end, primary, and IoV systems, utilizing a one-dimensional Markov chain to analyze system states. In addition, system throughput and EE are investigated. Finally, the validity and accuracy of the analysis results are verified by Monte Carlo simulation, and the influence of some parameters on the system performance is revealed. The results show that with the increase in SNR, the system throughput tends to be saturated, that is, the maximum throughput of the system can be achieved. A comparative analysis shows that the proposed protocol outperforms protocols lacking ARQ functionality. Most importantly, in the proposed CR–NOMA network structure utilizing the SWIPT protocol, this paper innovatively combines ARQ technology with the concept that parked vehicles can be used as relays for cooperating cell–edge vehicle communication. This approach enhances traffic management and supports the deployment of C-IoV systems. In addition, the protocol architecture proposed in this paper is also applicable to the scenario of D2D communication in IoT.

However, in order to avoid the analysis being too complex and cumbersome, this paper simply regards the cell–edge vehicle as a vehicle far away from the base station and located in the edge area of the network coverage, regardless of its movement or not, as long as the above characteristics can be satisfied. Therefore, in future work, we will focus on mobile cell–edge vehicles and enable them to independently select the best relay. In this context, a relay refers to the available parked vehicles within the effective network coverage. The selection procedure will take into account the needs of edge vehicle users so as to improve the ability of traffic management of cellular vehicular networks and increase the application value of relay vehicles to be parked in cellular deployment. This approach aims to enhance traffic management capabilities in C-IoV and increase the practical value of parked relay vehicles within cellular deployments.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding

This research was funded in part by the National Natural Science Foundation of China under grant no. 62467004, in part by the Hongliu First Class Discipline Development Project of the Lanzhou University of Technology under grant no. 25-225305, and in part by the Erasmus+ Programme of European Commission under grant no. 573879-EPP-1-2016-1-FR-EPPKA2-CBHE-JP.

Appendix A: The CDF of the Product of Two Gamma Random Variables

To figure out equation (29), the CDF of the product of two gamma random variables is given as follows. Let Z = XY (X ≥ 0, Y ≥ 0) be the product of the sum of two gamma random variables, and their PDFs can be expressed as

()

The CDF of the variable Y is expressed as

()

The CDF of Z is denoted by

()

By substituting the CDF of Y, we can obtain

()

Using some algebraic operations and

, it can be finally represented as

()

where K_ν(·) denotes the modified Bessel function of order ν of the second kind.

Open Research

Data Availability Statement

The data supporting this article are from previously reported studies and datasets, which have been cited.

References

1 Khan A. A., Rehmani M. H., and Rachedi A., Cognitive-Radio-Based Internet of Things: Applications, Architectures, Spectrum Related Functionalities, and Future Research Directions, IEEE Wireless Communications. (2017) 24, no. 3, 17–25, https://doi.org/10.1109/mwc.2017.1600404, 2-s2.0-85028719579.
10.1109/MWC.2017.1600404
Web of Science® Google Scholar
2 Ercan A. O., Sunay M. O., and Akyildiz I. F., RF Energy Harvesting and Transfer for Spectrum Sharing Cellular IoT Communications in 5G Systems, IEEE Transactions on Mobile Computing. (2018) 17, no. 7, 1680–1694, https://doi.org/10.1109/tmc.2017.2740378, 2-s2.0-85028505336.
10.1109/TMC.2017.2740378
Web of Science® Google Scholar
3 Shen X. H., Yu H. T., Liu X. Y., Bin Q. A. K., Luhach A. K., and Saravanan V., The Optimized Energy-Efficient Sensible Edge Processing Model for the Internet of Vehicles in Smart Cities, Sustainable Energy Technologies and Assessments. (2021) 47, https://doi.org/10.1016/j.seta.2021.101477.
10.1016/j.seta.2021.101477
Web of Science® Google Scholar
4 Collotta M., Ferrero R., Giusto E. et al., A Fuzzy Control System for Energy-Efficient Wireless Devices in the Internet of Vehicles, International Journal of Intelligent Systems. (2021) 36, no. 4, 1595–1618, https://doi.org/10.1002/int.22353.
10.1002/int.22353
Web of Science® Google Scholar
5 Su L., Wang F., and Li C., Performance Analysis of CDRT Based Cognitive NOMA Vehicular Networks with Ambient Backscatter Relays, Physical Communication. (2024) 66, https://doi.org/10.1016/j.phycom.2024.102470.
10.1016/j.phycom.2024.102470
Web of Science® Google Scholar
6 Ahmed R., Chen Y., Hassan B., Du L. P., Hassan T., and Dias J., Hybrid Machine-Learning-Based Spectrum Sensing and Allocation with Adaptive Congestion-Aware Modeling in CR-Assisted IoV Networks, IEEE Internet of Things Journal. (2022) 9, no. 24, 25100–25116, https://doi.org/10.1109/jiot.2022.3195425.
10.1109/JIOT.2022.3195425
Web of Science® Google Scholar
7 Saraç S. and Aygölü Ü., ARQ-based Cooperative Spectrum Sharing Protocols for Cognitive Radio Networks, Wireless Networks. (2018) 25, no. 5, 2573–2585, https://doi.org/10.1007/s11276-018-1686-3, 2-s2.0-85042174964.
10.1007/s11276-018-1686-3
Google Scholar
8 Li S. P., Li H. L., Yang Q., and Gaber J., Hybrid Cooperative Spectrum Sharing Protocols Based on OMA and NOMA in Cognitive Radio Networks, Applied Sciences. (2022) 12, no. 24, https://doi.org/10.3390/app122412683.
10.1007/978-3-031-12762-5
Google Scholar
9 Luo L., Li Q., and Cheng J., Performance Analysis of Overlay Cognitive NOMA Systems with Imperfect Successive Interference Cancellation, IEEE Transactions on Communications. (2020) 68, no. 8, 4709–4722, https://doi.org/10.1109/tcomm.2020.2992471.
10.1109/TCOMM.2020.2992471
Web of Science® Google Scholar
10 Al-Obiedollah H., Salameh H. B., Gharaibeh A., Cumanan K., Ding Z. G., and Dobre O. A., Harvested Power Fairness-Based Multi-Carrier NOMA IoT Networks with SWIPT, IEEE Wireless Communications Letters. (2023) 12, no. 2, 381–385, https://doi.org/10.1109/lwc.2022.3227468.
10.1109/LWC.2022.3227468
Web of Science® Google Scholar
11 Chaudhary R. and Kumar N., SecGreen: Secrecy Ensured Power Optimization Scheme for Software-Defined Connected IoV, IEEE Transactions on Mobile Computing. (2023) 22, no. 4, 2370–2386, https://doi.org/10.1109/tmc.2021.3116954.
10.1109/TMC.2021.3116954
Web of Science® Google Scholar
12 Li S. P., Jia T. T., Yang H. L., Gao R. M., and Yang Q., SWIPT Cooperative Protocol for Improving the Communication Quality of Cell-Edge Users in NOMA Network and its Performance Analysis, Electronics. (2023) 12, no. 17, https://doi.org/10.3390/electronics12173583.
10.3390/electronics12173583
Web of Science® Google Scholar
13 Ibraheem S. M., Bedawy W., Saad W., and Shokair M., Full-Duplex Relaying Schemes for Cooperative NOMA Networks with SWIPT, Wireless Personal Communications. (2021) 124, no. 1, 727–739, https://doi.org/10.1007/s11277-021-09380-5.
10.1007/s11277-021-09380-5
Web of Science® Google Scholar
14 Tran H. Q., Two Energy Harvesting Protocols for SWIPT at UAVs in Cooperative Relaying Networks of IoT Systems, Wireless Personal Communications. (2021) 122, no. 4, 3719–3740, https://doi.org/10.1007/s11277-021-09108-5.
10.1007/s11277-021-09108-5
Web of Science® Google Scholar
15 Parihar A. S., Swami P., Bhatia V., and Ding Z. G., Performance Analysis of SWIPT Enabled Cooperative-NOMA in Heterogeneous Networks Using Carrier Sensing, IEEE Transactions on Vehicular Technology. (2021) 70, no. 10, 10646–10656, https://doi.org/10.1109/tvt.2021.3110806.
10.1109/TVT.2021.3110806
Web of Science® Google Scholar
16 Do T. N., Da Costa D. B., Duong T. Q., and An B., Improving the Performance of Cell-Edge Users in MISO-NOMA Systems Using TAS and SWIPT-Based Cooperative Transmissions, IEEE Transactions on Green Communications and Networking. (2018) 2, no. 1, 49–62, https://doi.org/10.1109/tgcn.2017.2777510, 2-s2.0-85050979271.
10.1109/TGCN.2017.2777510
Web of Science® Google Scholar
17 Özgün E. and Aygölü Ü., Energy Harvesting in ARQ-Based Cooperative Broadcast and NOMA Networks, Wireless Networks. (2024) 30, no. 5, pp4181–4194, https://doi.org/10.1007/s11276-021-02872-z.
10.1007/s11276-021-02872-z
Web of Science® Google Scholar
18 Singh C. K. and Upadhyay P. K., Overlay Cognitive IoT-Based Full-Duplex Relaying NOMA Systems with Hardware Imperfections, IEEE Internet of Things Journal. (2022) 9, no. 9, 6578–6596, https://doi.org/10.1109/jiot.2021.3111124.
10.1109/JIOT.2021.3111124
Web of Science® Google Scholar
19 Shukla A. K., Sharanya J., Yadav K., and Upadhyay P. K., Exploiting SWIPT-Enabled IoT-Based Cognitive Nonorthogonal Multiple Access with Coordinated Direct and Relay Transmission, IEEE Sensors Journal. (2022) 22, no. 19, 18988–18999, https://doi.org/10.1109/jsen.2022.3198627.
10.1109/JSEN.2022.3198627
Web of Science® Google Scholar
20 Yu Y. H., Yang Z., Wu Y., Hussein J. A., Jia W. K., and Dong Z. C., Outage Performance of NOMA in Cooperative Cognitive Radio Networks with SWIPT, IEEE Access. (2019) 7, 117308–117317, https://doi.org/10.1109/access.2019.2936321.
10.1109/ACCESS.2019.2936321
Web of Science® Google Scholar
21 Zhang Z., Ma Z., Xiao M., Ding Z., and Fan P., Full-Duplex Device-To-Device-Aided Cooperative Non-orthogonal Multiple Access, IEEE Transactions on Vehicular Technology. (2017) 66, no. 5, 4467–4471.
Web of Science® Google Scholar
22 Aslam S., Hasan N. U., Shahid A., Jang J. W., and Lee K. G., Device Centric Throughput and QoS Optimization for IoTsin a Smart Building Using CRN-Techniques, Sensors. (2016) 16, no. 10, https://doi.org/10.3390/s16101647, 2-s2.0-84990964670.
10.3390/s16101647
Web of Science® Google Scholar
23 Khuntia P., Hazra R., and Goswami P., A Bidirectional Relay-Assisted Underlay Device-To-Device Communication in Cellular Networks: An IoT Application for FinTech, IEEE Internet of Things Journal. (2023) 10, no. 3, 2174–2182, https://doi.org/10.1109/jiot.2021.3088539.
10.1109/JIOT.2021.3088539
Web of Science® Google Scholar
24 Gurjar D. S., Nguyen H. H., and Tuan H. D., Wireless Information and Power Transfer for IoT Applications in Overlay Cognitive Radio Networks, IEEE Internet of Things Journal. (2019) 6, no. 2, 3257–3270, https://doi.org/10.1109/jiot.2018.2882207, 2-s2.0-85057211018.
10.1109/JIOT.2018.2882207
Web of Science® Google Scholar
25 Gurjar D. S. and Das A., Wireless Information and Power Transfer in Three-phase Two-Way DF-Relay Networks over Nakagami-M Fading, 2019 IEEE VTS Asia Pacific Wireless Communications Symposium. (2019) 31–35.
Google Scholar
26 Gurjar D. S. and H H., Nguyen. Bidirectional Primary and Secondary Transmissions with Hybrid-SWIPT in Cognitive Radio Networks, 2018 IEEE 88th Vehicular Technology Conference (VTC-Fall), April 2018, 1–6.
Google Scholar
27 Zhao W., She R., and Bao H., Energy Efficiency Maximization for Two-Way Relay Assisted CR-NOMA System Based on SWIPT, IEEE Access. (2019) 7, 72062–72071, https://doi.org/10.1109/access.2019.2919965, 2-s2.0-85067352996.
10.1109/ACCESS.2019.2919965
Web of Science® Google Scholar
28 Mukherjee A., Acharya T., and Khandaker M. R. A., Outage Analysis for SWIPT-Enabled Two-Way Cognitive Cooperative Communications, IEEE Transactions on Vehicular Technology. (2018) 67, no. 9, 9032–9036, https://doi.org/10.1109/tvt.2018.2840140, 2-s2.0-85047618416.
10.1109/TVT.2018.2840140
Web of Science® Google Scholar
29 Ghosh S., Acharya T., and Maity S. P., On Outage Analysis in SWIPT Enabled Bidirectional D2D Communications Using Spectrum Sharing in Cellular Networks, IEEE Transactions on Vehicular Technology. (2020) 69, no. 9, 10167–10176, https://doi.org/10.1109/tvt.2020.3003228.
10.1109/TVT.2020.3003228
Web of Science® Google Scholar
30 Ghosh S., Acharya T., and Maity S. P., Outage Analysis in SWIPT Enabled Cooperative AF/DF Relay Assisted Two-Way Spectrum Sharing Communication, IEEE Transactions on Cognitive Communications and Networking. (2022) 8, no. 3, 1434–1443, https://doi.org/10.1109/tccn.2022.3171223.
10.1109/TCCN.2022.3171223
Web of Science® Google Scholar
31 Lavi N., Levy H., and Shavitt Y., Cellular Networks with Vehicular Relay Nodes: Capacity, Coverage, and Scalability, IEEE Transactions on Vehicular Technology. (2019) 68, no. 9, 9074–9087, https://doi.org/10.1109/tvt.2019.2927293.
10.1109/TVT.2019.2927293
Web of Science® Google Scholar

All articles

Exploiting SWIPT–Enabled ARQ–Based Bidirectional Cellular IoV Spectrum Sharing Protocol and Its Performance Analysis

Abstract

1. Introduction