Non‐orthogonal multiple access relaying with truncated ARQ
2016; Institution of Engineering and Technology; Volume: 11; Issue: 4 Linguagem: Inglês
10.1049/iet-com.2016.0650
ISSN1751-8636
AutoresZhiyuan Yu, Chao Zhai, Ju Liu,
Tópico(s)Full-Duplex Wireless Communications
ResumoIET CommunicationsVolume 11, Issue 4 p. 514-521 Research ArticleFree Access Non-orthogonal multiple access relaying with truncated ARQ Zhiyuan Yu, Zhiyuan Yu School of Information Science and Engineering, Shandong University, Jinan, People's Republic of China National Mobile Communications Research Laboratory, Southeast University, Nanjing, People's Republic of ChinaSearch for more papers by this authorChao Zhai, Chao Zhai School of Information Science and Engineering, Shandong University, Jinan, People's Republic of ChinaSearch for more papers by this authorJu Liu, Corresponding Author Ju Liu juliu@sdu.edu.cn School of Information Science and Engineering, Shandong University, Jinan, People's Republic of China National Mobile Communications Research Laboratory, Southeast University, Nanjing, People's Republic of ChinaSearch for more papers by this author Zhiyuan Yu, Zhiyuan Yu School of Information Science and Engineering, Shandong University, Jinan, People's Republic of China National Mobile Communications Research Laboratory, Southeast University, Nanjing, People's Republic of ChinaSearch for more papers by this authorChao Zhai, Chao Zhai School of Information Science and Engineering, Shandong University, Jinan, People's Republic of ChinaSearch for more papers by this authorJu Liu, Corresponding Author Ju Liu juliu@sdu.edu.cn School of Information Science and Engineering, Shandong University, Jinan, People's Republic of China National Mobile Communications Research Laboratory, Southeast University, Nanjing, People's Republic of ChinaSearch for more papers by this author First published: 01 March 2017 https://doi.org/10.1049/iet-com.2016.0650Citations: 7AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract In this study, the authors propose a non-orthogonal multiple access (NOMA) relaying protocol with truncated automatic repeat request (ARQ), where the source simultaneously transmits two signals to the relay and the destination using superposition coding technique in the power domain. The successive interference cancelation technique is adopted by the relay, who should first decode its own data and then cancel it for the decoding of destination data. If either the data intended for the relay or the destination is erroneously received by the relay, it will be retransmitted by the source using full power. After correctly decoding the source data intended for the destination, the relay will forward it and the retransmission is performed when the data is erroneously received by the destination. The authors derive the system throughput and investigate the impacts of several parameters, such as power allocation factor, target rate, and distance between terminals etc. In terms of throughput, the authors' proposed scheme can greatly outperform the conventional relaying scheme with truncated ARQ. 1 Introduction Recently, non-orthogonal multiple access (NOMA) has been attracting intensive research interests and is recognised as a promising technique in 5G network [1–3]. Different from the conventional orthogonal multiple access (OMA), the multi-user signals are superimposed in the power domain and the successive interference cancelation (SIC) technique is exploited by the receiver to retrieve its desired signal through decoding and cancelling the stronger signals. The NOMA can not only enhance the spectral efficiency [4], but also improve the capacity [5] and the service quality of cell-edge users [6]. Further, NOMA can guarantee the fairness of users through the optimal power allocation [7]. By its nature, NOMA has been integrated into several traditional scenarios to achieve better performance, such as the multi-user multiple-input multiple-output (MU-MIMO) system [8], the cognitive radio network [9], and the high-mobility network [10] etc. By applying NOMA in the cooperative relaying system, a suboptimal power allocation strategy was proposed in [5] to maximise the average rate. For the coordinated multipoint transmission, two base stations (BSs) can simultaneously transmit signals to their nearby users and one common cell-edge user using the Alamouti coding technique [6]. The fairness of downlink users was guaranteed by optimally allocating powers with the instantaneous or statistical channel state information available [7]. For the multiuser NOMA-MIMO system, Ding et al. studied the optimal power allocation based on users' quality of service (QoS) requirements and characterised the impacts of user pairing [8]. For the cognitive radio network, Liu et al. derived the outage probability and analysed the diversity order of NOMA with randomly deployed users [9]. For the high-mobility network, NOMA can maintain the cell throughput gain compared with OMA for various user speeds [10]. Kim and Lee proposed a coordinated direct and relay transmission protocol with NOMA and derived the ergodic sum capacity in [11]. For the multi-user network, the maximal mutual information was obtained through the power allocation at BS [12]. Automatic repeat request (ARQ) can efficiently overcome the channel fading [13] and guarantee the reliability of data transmission [14]. The transmitter will retransmit its data if it is erroneously received by the receiver. In practice, the truncated ARQ (TARQ) is often used to minimise the delay by limiting the number of retransmissions [15, 16]. Dai and Letaief proposed a selective cooperative diversity with TARQ, where the data is retransmitted using space-time coding [17]. Zhai et al. proposed uncoordinated TARQ schemes with the relays spatially randomly distributed, who compete for the retransmission based on the local information [18]. Recently, combined with the hybrid ARQ (HARQ), Saito et al. studied the system-level performance of the downlink NOMA in cellular network [1]. For the multiuser system, Choi showed that the NOMA with SIC outperforms the OMA scheme in the whole range of signal-to-noise ratio (SNR) for a lower data rate [19]. For the power allocation in the retransmission phase, Li et al. proposed to consider not only the channel quality indicator of paired user equipments, but also the HARQ combination gain [20]. In this paper, we propose a cooperative relaying NOMA protocol with TARQ to improve the transmission efficiency and reliability. The source generates a composite signal by linearly combining the data intended for the relay and the destination. Then, the source broadcasts this composite signal to the relay. If the data intended for the relay or the destination cannot be correctly decoded by the relay using the SIC technique, it will be retransmitted by the source using the total power. Otherwise, after correctly decoding the source data, the relay can forward it to the destination and the data is retransmitted in case the original transmission fails. Only one retransmission is allowed at either source or relay for both signals to guarantee the users' QoS requirements by introducing less delay. The outage event occurs if the data cannot be successfully decoded after the retransmission. For the superposition coding at source, a fraction of power is allocated to the relay data and the remaining power is used for the destination data. The relay should first decode its desired data and then cancel it to retrieve the data intended for the destination. Further, we analyse the system throughput by considering various transmission cases of the source data at both the relay and the destination. The impacts of various parameters to the system performance are revealed and the power allocation in the superposition coding is numerically determined through maximising the system throughput. The remainder of this paper is organised as follows: In Section 2, we describe the system model and illustrate the transmission process of our protocol. Section 3 analyses the system throughput. Performance results are presented in Section 4. Section 5 concludes this paper. 2 System model and protocol design We consider the cooperative relaying system with one source, one relay, and one destination. The source intends to transmit and to the relay and the destination, respectively, using the NOMA protocol [2], which is very necessary and helpful to improve the spectral efficiency. Here, the relay can be regarded as an user that needs to receive the source signal. For the superposition coding at the source, fraction of the total power is allocated to , while the remaining part is allocated to . There exists no direct link between source and destination due to the physical obstacles or long distances [11, 21]. The relay tries to decode both and using the SIC technique and then forwards to the destination. In the retransmission block, the source uses the total power to transmit or . In this paper, the following assumptions are made. The maximum number of retransmission for both signals at either source or relay is set as one, i.e. the data packets of and will be discarded if they are still erroneously decoded by the relay or the destination after one retransmission. This assumption is reasonable for the delay-limited data transmissions, such as audio or video signals [14]. Each terminal is equipped with a single omnidirectional antenna and the channel undergoes Rayleigh block fading [5, 7]. The channel fading is assumed to be independent over different links and across different time blocks. Between any two terminals i and j, the channel coefficient is denoted as and the large-scale path loss is denoted as with v being pathloss exponent. The small-scale power fading is , which is exponentially distributed with unit mean. The acknowledgement (ACK) and negative acknowledgement (NACK) frames are assumed to be error-free and the latency is negligible [13, 22]. 2.1 Transmission process As illustrated in the flow chart of Fig. 1, based on the decoding statuses of relay and destination, a NOMA protocol with TARQ is designed to realise the high-efficient and reliable data transmission. According to the decoding status of relay, we have the following observations: If the relay correctly decodes in the original transmission phase, the ACK frame is released to the source. The relay tries to decode after cancelling . If the relay fails to decode , the NACK frame is fed back to the source, who will retransmit to the relay in the following time block using the total power. If the relay correctly decodes but fails to decode , the NACK frame is sent to the source and the source only retransmits to the relay in the following time block. If the relay successfully decodes both and , then it forwards to the destination in the next time block. Figure 1Open in figure viewerPowerPoint Flow chart of NOMA relaying with truncated ARQ If the relay has successfully decoded , it will forward to the destination. According to the decoding status of destination, we have the following observations: If is correctly decoded by the destination, the ACK frame is released to the source via the relay. New data packets will be transmitted in the next time block. If the destination fails to decode in the original transmission phase, the NACK frame is released and then the relay retransmits to the destination. If the relay or the destination still fails to decode or after the retransmission, the NACK frame is reported to the source and new data packets will be transmitted by the source in the next time block. 2.2 Achievable rate at relay and destination In the broadcast phase, the source transmits and using the superposition coding. The received signal at the relay can be expressed as (1)where represents the source transmission power and denotes the additive white Gaussian noise at the relay. The signals are assumed to have unit power, i.e. . The relay first decodes by treating as noise. If is successfully decoded, it will be cancelled from the received composite signal to further decode . In the original transmission, the achievable rates at the relay for and are denoted as and , respectively, i.e. (2)where represents the small-scale power fading between source and relay in the original transmission block. If the relay fails to decode or , the source will retransmit the respective signal using its total power. The relay combines the received original and retransmitted signals in successive time blocks to decode or . The achievable rate at relay is denoted as and , respectively, for and , i.e. (3)where and represent the small-scale power fading when the source retransmits and , respectively. If the relay correctly decodes , it will forward to the destination. The achievable rate at the destination is denoted as and given as (4)where is the transmission power of the relay and is the small-scale power fading between relay and destination in the original transmission block. Otherwise, will be retransmitted from the relay to the destination. The destination combines the received signals in successive time blocks for the decoding of . The achievable rate is denoted as and expressed as (5)where denotes the small-scale power fading of the retransmission block. For the delay-limited scenario as considered in our work, the data transmission rate is fixed and the target rate of each user is determined by its QoS requirements [2]. If the achievable rate is larger than the target transmission rate, the signal can be successfully decoded by the relay or the destination, otherwise, the outage event occurs. 3 Throughput analysis The target transmission rates for and are denoted as and , respectively. The transmission is deemed to be successful when the achievable rate is greater than the target rate. According to the transmission process, there exist four cases as shown in Table 1, where Tx and Re represent 'Transmit' and 'Retransmit', respectively. The system throughput considering all the cases is expressed as (6)where , , , and represent the throughput of Case 1, Case 2, Case 3, and Case 4, respectively. The occurrence of different cases are depicted as following: Case 1: The relay correctly decodes and in the original transmission phase and there is no source retransmission, i.e. and . Case 2: The relay successfully decodes but fails to decode in the original transmission. Then, the source retransmits to the relay, which will be jointly decoded, i.e. and . Case 3: The relay fails to decode but successfully decodes it in the retransmission block. The relay can successfully decode after cancelling in the received composite signal, i.e. , , and . Case 4: If the relay erroneously decodes both and in the original transmission, then the source will retransmit both signals to the relay, i.e. , , and . Table 1. Transmission cases with truncated ARQ CASE SUBCASE LINK S→R R→D case 1 subcase 1a ✓ ✓ ✓ subcase 1b ✓ ✓ × ✓ subcase 1c ✓ ✓ × × case 2 subcase 2a ✓ × ✓ ✓ subcase 2b ✓ × ✓ × ✓ subcase 2c ✓ × ✓ × × subcase 2d ✓ × × case 3 subcase 3a × ✓ ✓ ✓ subcase 3b × ✓ ✓ × ✓ subcase 3c × ✓ ✓ × × case 4 subcase 4a × ✓ × ✓ ✓ subcase 4b × ✓ × ✓ × ✓ subcase 4c × ✓ × ✓ × × subcase 4d × ✓ × × ✓: Correctly decode, ×: Erroneously decode, N/A: Not available Let and denote the event of relay successfully decoding and in the original transmission, respectively. Let and denote the event of relay correctly decoding and , respectively, in the retransmission. 3.1 Case 1: relay successfully decodes both signals without retransmission In Case 1, the relay successfully decodes and in the original transmission. Then, the relay forwards to the destination. Considering whether the destination correctly decodes or not, we can divide Case 1 into three subcases, which can be denoted as Subcase 1a, 1b, and 1c: Subcase 1a: The destination successfully decodes without retransmission, i.e. , which is denoted as Event . Subcase 1b: The destination fails to decode in the original transmission, which is denoted as Event . In the following time block, is correctly decoded by the destination after the retransmission, which is denoted as Event , i.e. and . Subcase 1c: The destination still fails to decode after retransmission, which is denoted as Event , i.e. and . The throughput of Case 1 can be derived only when is satisfied, because the condition should be larger than zero [2]. Otherwise, Case 1 cannot occur and the throughput of Case 1 is zero. The throughput of Case 1 is expressed as (7)where , , and represent the throughput of Subcase 1a, 1b, and 1c, respectively. The throughput of Subcase 1a is derived as (8)where , , and . The throughput of Subcase 1b is derived as (9)The throughput of Subcase 1c is derived as (10)Substituting the results of (8)–(10) into (7), we can obtain the total throughput of Case 1. Take the derivative of with respect to , we can see that the throughput of Case 1 increases first with and then decreases with . If , the maximal throughput of Case 1 can be obtained. This phenomenon can be demonstrated in Fig. 3b. 3.2 Case 2: source retransmits In Case 2, the relay successfully decodes but fails to decode in the original transmission. Then, the source retransmits to the relay using the total power. If is correctly decoded, i.e. , the relay will forward it to the destination. Regarding to the decoding statuses of the destination, we have three subcases: Subcase 2a: The destination successfully decodes without retransmission, i.e. . Subcase 2b: The destination fails to decode in the original transmission. However, is correctly decoded by the destination after the retransmission, i.e. and . Subcase 2c: The destination still fails to decode after retransmission, i.e. and . Particularly, if the relay successfully decodes in the original transmission but fails to decode after the retransmission, i.e. , , and , the relay cannot forward to the destination, which is denoted as Subcase 2d. The throughput of Case 2 can be derived if is satisfied. Otherwise, Case 2 cannot happen and the throughput is zero. The throughput of Case 2 is expressed as (11)where , , , and represent the throughput of Subcase 2a, 2b, 2c, and 2d, respectively. The throughput of Subcase 2a can be derived as (12)where The throughput of Subcase 2b can be derived as (13)The throughput of Subcase 2c is derived as (14)The throughput of Subcase 2d is derived as (15)Substituting the results of (12)–(15) into (11), we can obtain the total throughput of Case 2. 3.3 Case 3: source retransmits In Case 3, the relay fails to decode in the original transmission but successfully decodes it in the retransmission block. If is correctly decoded by the relay without retransmission, the relay forwards to the destination. The destination will try to decode in the following time block. Based on the decoding statuses of the destination, there exist three subcases: Subcase 3a: The destination successfully decodes without retransmission, i.e. . Subcase 3b: The destination fails to decode in the original transmission. However, is correctly decoded by the destination after the retransmission, i.e. and . Subcase 3c: The destination still fails to decode after the retransmission, i.e. and . The throughput of Case 3 can be expressed as (16)where , , and represent the throughput of Subcase 3a, 3b, and 3c, respectively. The throughput of Subcase 3a is derived as (17)where the function is defined as (18)The throughput of Subcase 3b can be derived as (19) The throughput of Subcase 3c is derived as (20) In (17), (19), and (20), the integral is taken over different regions according to the value of , i.e. When , the integral region is . When , the integral is taken over . Substituting the results of (17), (19), (20) into (16), we can obtain the total throughput of Case 3. 3.4 Case 4: source retransmits both and In Case 4, the source will retransmit and when both signals are erroneously decoded in the original transmission. If and can be successfully decoded by the relay after retransmission, i.e. , , , and , the relay will forward to the destination. Based on the decoding statuses at the destination, Subcase 4a, 4b, or 4c will possibly occur in the following time block. Subcase 4a: The destination successfully decodes without retransmission, i.e. . Subcase 4b: The destination fails to decode in the original transmission. However, is correctly decoded by the destination after retransmission, i.e. and . Subcase 4c: The destination still fails to decode after retransmission, i.e. and . Particularly, if the relay can only successfully decodes but fails to decode after one retransmission, i.e. , no signal will be forwarded to the destination and new data packets will be transmitted by the source, which is denoted as Subcase 4d. The throughput of Case 4 can be expressed as (21)where , , , and represent the throughput of Subcase 4a, 4b, 4c, and 4d, respectively. The throughput of Subcase 4a can be derived as (22) The function is expressed as (23)The throughput of Subcase 4b is derived as (24) The throughput of Subcase 4c can be derived as (25) The throughput of Subcase 4d is derived as (26)In (22), (24)–(26), the integral is taken over different regions according to the value of , i.e. If , the integral is taken over , with If , the integral is taken over . Substituting the results of (22), (24)–(26) into (21), the throughput of Case 4 can be obtained. 4 Numerical and simulation results In this section, the simulation results are presented to validate our analysis. The impacts of various parameter settings to the system performance are revealed, such as power allocation, distance between terminals, target rates, and transmit SNR, etc. Unless stated otherwise, the system parameters are set as: and , the target rates , the pathloss exponent , and the transmission power of relay . We compare our scheme with the conventional relaying with truncated ARQ (CR-TARQ). In the CR-TARQ, the source will transmit and successively to the relay and the relay will forward to the destination after correctly decoding it. For the source-relay or the relay-destination link, the signal may be retransmitted when the original transmission fails. At either the relay or the destination, both the original and retransmitted signals will be combined for the decoding. Similarly to our proposed scheme, there exists no direct link between source and relay and only one retransmission is available. 4.1 System throughput Fig. 2a shows the system throughput versus the power allocation factor for different target rates R. The transmit SNR at source is set as 55 dB. When , our proposed scheme outperforms the CR-TARQ. If , our proposed scheme is inferior to CR-TARQ. It can be seen from Fig. 2b that the throughput of Case 1 and Case 2 dominates the total throughput when the operation condition of these two cases are satisfied. Otherwise, Case 3 and Case 4 govern the system throughput and the occurrence of retransmission will lower the throughput. Figure 2Open in figure viewerPowerPoint System throughput versus the power allocation factor a System throughput versus power allocation factor for different target rates R b System throughput versus power allocation factor for four cases with Fig. 3a shows the system throughput versus the target rate R for different . The transmit SNR of source is set as 50 dB. Given a fixed , the system throughput gets better first and then turns worse with the target rate R. From Fig. 3b, we can see that the throughput of Case 1 and Case 2 increases first and then decreases. The throughput of Case 4 increases rapidly with the increase of R, which dominates the total throughput and its curve gradually coincides with the total throughput. Moreover, given a target rate, our scheme outperforms the CR-TARQ by adjusting the power allocation factor, which can provide some guidelines for the network deployment. In Fig. 4, the power allocation factor for is fixed as and the transmit SNR of relay is set as 50 dB. With the increase of the source transmit SNR, the throughput improves quickly and then becomes smooth gradually. Moreover, our proposed scheme performes better than CR-TARQ in the higher SNR regime, because the higher transmit SNR will improve the success probability of data transmission and then achieves the higher spectral efficiency in the power domain of NOMA scheme compared with CR-TARQ. However, if the target rate gets larger, e.g. , the CR-TARQ outperforms our scheme in terms of throughput. Even the retransmission occurs, the data success probability will still decrease and then the system throughput becomes worse due to the additional transmission time. Figure 3Open in figure viewerPowerPoint System throughput versus the target rates a System throughput versus target rates for different b System throughput versus target rates for four cases with Figure 4Open in figure viewerPowerPoint System throughput versus source transmit SNR for different target rates R In Fig. 5, the transmit SNR at source and relay is set as 65 and 50 dB, respectively. The power factor is set as and the relay-destination distance is set as . The target rate for is set as . The system throughput is determined by the quality of both links: source-relay and relay-destination, which has mutual influence on each other. With the increase of source-relay distance, the throughput gets worse because the relay's outage probability increases. The throughput improves slightly for the proper source-relay distance, because the shorter distance between relay and destination is beneficial to successfully forward . When the source-relay distance continues to increase, it becomes difficult for the relay to correctly decode both and . Moreover, our proposed scheme performs better than CR-TARQ over a certain range of distance. Figure 5Open in figure viewerPowerPoint System throughput versus with fixed rate for different target rates From Figs. 2-4–5, we can see that our theoretical results coincide exactly with the simulation results, which can verify the tightness of our analysis. Moreover, our proposed scheme can greatly improve the throughput compared with CR-TARQ. 4.2 Maximal system throughput The throughput expressions are complicated w.r.t the power allocation factor . It is intractable to determine the optimal in closed-form. The maximal system throughput can be obtained through numerically searching in the interval of (0.01, 0.99). Fig. 6 shows the maximal throughput versus the source transmit SNR for different R. The transmit SNR of relay is fixed as 50 dB. The maximal throughput of our scheme increases faster in the lower SNR regime, but it changes slightly in the higher SNR regime. Our scheme outperforms CR-TARQ when SNR is large enough, because the higher transmit SNR will help to improve the success probability of data transmission and thus it is helpful to enhance the system throughput, which inspires us the importance of selecting the suitable transmit SNR for NOMA. Figure 6Open in figure viewerPowerPoint Maximal system throughput versus transmit SNR at source for different target rates R Fig. 7 shows the maximal throughput versus the target rate for different . The source transmit SNR is set as 65 dB. Both the target rates and impose great impacts to the system throughput. We can see that the maximal throughput gets better first and then turns worse with the increase of , because the higher target rate degrades the success probability of data transmission. Further, our scheme performs much better than CR-TARQ. Figure 7Open in figure viewerPowerPoint Maximal system throughput versus target rates for different In Fig. 8, the target rate is set as and the relay transmit SNR is fixed as 50 dB. The relay-destination distance varies as . For the fixed source transmit SNR, the maximal throughput increases first and then decreases with longer source-relay distance. It is because the shorter relay-destination distance can help improve the transmission quality of . However, if the source-relay distance continues to increase, the quality of source-relay link becomes much poorer, which is harmful for the signal decoding at the relay. Moreover, the high transmit SNR will improve the maximum system throughput compared with CR-TARQ. Figure 8Open in figure viewerPowerPoint Maximal system throughput versus for different source transmit SNR 5 Conclusion In this paper, we have proposed a cooperative relaying NOMA scheme by exploiting the truncated ARQ. The system throughput has been derived and the performance results are provided to reveal the impacts of various parameters to the system performance, which can provide some guidelines for the network deployment. The maximal system throughput can be determined through numerically searching the power allocation factor. Performance results show that our proposed scheme greatly outperforms the conventional relaying with truncated ARQ when the transmit SNR is high or the transmission rate is low. 6 Acknowledgments This work was supported by the National Natural Science Foundation of China (grant no. 61371188), the Research Fund for the Doctoral Program of Higher Education (grant no. 20130131110029), the Open Research Fund of the National Mobile Communications Research Laboratory (grant no. 2012D10), the Special Funds for Postdoctoral Innovative Projects of Shandong Province (grant no. 201401013), and the Fundamental Research Funds of Shandong University. References 1Saito Y., Benjebbour A., Kishiyama Y. et al.: ' System-level performance evaluation of downlink non-orthogonal multiple access (NOMA)'. Proc. IEEE 24th Annual Int. Symp. on Personal, Indoor and Mobile Radio Communications (PIMRC), London, UK, September 2013, pp. 611– 616 2Ding Z., Yang Z., Fan P. et al.: 'On the performance of non-orthogonal multiple access in 5G systems with randomly deployed users', IEEE Signal Process. Lett., 2014, 21, (12), pp. 1501– 1505 (doi: https://doi.org/10.1109/LSP.2014.2343971) 3Ding Z., Peng M., Poor H.V.: 'Cooperative non-orthogonal multiple access in 5G systems', IEEE Commun. Lett., 2015, 19, (8), pp. 1462– 1465 (doi: https://doi.org/10.1109/LCOMM.2015.2441064) 4Dai L., Wang B., Yuan Y. et al.: 'Non-orthogonal multiple access for 5G: solutions, challenges, opportunities, and future research trends', IEEE Commun. Mag., 2015, 53, (9), pp. 74– 81 (doi: https://doi.org/10.1109/MCOM.2015.7263349) 5Kim J.B., Lee I.H.: 'Capacity analysis of cooperative relaying systems using non-orthogonal multiple access', IEEE Commun. Lett., 2015, 19, (11), pp. 1949– 1952 (doi: https://doi.org/10.1109/LCOMM.2015.2472414) 6Choi J.: 'Non-orthogonal multiple access in downlink coordinated two-point systems', IEEE Commun. Lett., 2014, 18, (2), pp. 313– 316 (doi: https://doi.org/10.1109/LCOMM.2013.123113.132450) 7Timotheou S., Krikidis I.: 'Fairness for non-orthogonal multiple access in 5G systems', IEEE Signal Process. Lett., 2015, 22, (10), pp. 1647– 1651 (doi: https://doi.org/10.1109/LSP.2015.2417119) 8Ding Z., Adachi F., Poor H.V.: 'The application of MIMO to non-orthogonal multiple access', IEEE Trans. Wirel. Commun., 2016, 15, (1), pp. 537– 552 (doi: https://doi.org/10.1109/TWC.2015.2475746) 9Liu Y., Ding Z., Elkashlan M. et al.: 'Non-orthogonal multiple access in large-scale underlay cognitive radio networks', IEEE Trans. Veh. Tech., 2016, 65, (12), pp. 10152– 10157 (doi: https://doi.org/10.1109/TVT.2016.2524694) 10Benjebbour A., Saito Y., Kishiyama Y. et al.: ' Concept and practical considerations of non-orthogonal multiple access (NOMA) for future radio access'. Proc. Int. Symp. on Intelligent Signal Processing and Communications Systems (ISPACS), Naha, Japan, November 2013, pp. 770– 774 11Kim J.B., Lee I.H.: 'Non-orthogonal multiple access in coordinated direct and relay transmission', IEEE Commun. Lett., 2015, 19, (11), pp. 2037– 2040 (doi: https://doi.org/10.1109/LCOMM.2015.2474856) 12Choi J.: 'On the power allocation for a practical multiuser superposition scheme in NOMA systems', IEEE Commun. Lett., 2016, 20, (3), pp. 438– 441 (doi: https://doi.org/10.1109/LCOMM.2016.2518165) 13Krikidis I.: 'Distributed truncated ARQ protocol for cooperative diversity networks', IET Commun., 2007, 1, (6), pp. 1212– 1217 (doi: https://doi.org/10.1049/iet-com:20070034) 14Pack S., Shen X., Mark J.W.: 'Optimizing truncated ARQ scheme over wireless fading channels', IEEE Trans. Veh. Tech., 2008, 57, (2), pp. 1302– 1305 (doi: https://doi.org/10.1109/TVT.2007.907082) 15Malkamaki E., Leib H.: 'Performance of truncated type-II hybrid ARQ schemes with noisy feedback over block fading channels', IEEE Trans. Commun., 2000, 48, (9), pp. 1477– 1487 (doi: https://doi.org/10.1109/26.870015) 16Yang Y., Ma H., Aïssa S.: 'Cross-layer combining of adaptive modulation and truncated ARQ under cognitive radio resource requirements', IEEE Trans. Veh. Tech., 2012, 61, (9), pp. 4020– 4030 (doi: https://doi.org/10.1109/TVT.2012.2214410) 17Dai L., Letaief K.B.: 'Throughput maximization of ad-hoc wireless networks using adaptive cooperative diversity and truncated ARQ', IEEE Trans. Commun., 2008, 56, (11), pp. 1907– 1918 (doi: https://doi.org/10.1109/TCOMM.2008.041164) 18Zhai C., Zhang W., Mao G.: 'Uncoordinated cooperative communications with spatially random relays', IEEE Trans. Wirel. Commun., 2012, 11, (9), pp. 3126– 3135 (doi: https://doi.org/10.1109/TWC.2012.072512.111065) 19Choi J.: ' H-ARQ based non-orthogonal multiple access with successive interference cancellation'. Proc. IEEE Global Telecommunications Conf. (GLOBECOM), New Orleans, USA, November 2008, pp. 1– 5 20Li A., Benjebbour A., Chen X. et al.: ' Investigation on hybrid automatic repeat request (HARQ) design for NOMA with SU-MIMO'. Proc. IEEE 26th Annual Symp. on Personal, Indoor, and Mobile Radio Communications (PIMRC), Hong Kong, China, August 2015, pp. 590– 594 21Nasir A., Zhou X., Durrani S. et al.: 'Wireless-powered relays in cooperative communications: time-switching relaying protocols and throughput analysis', IEEE Trans. Commun., 2015, 63, (5), pp. 1607– 1622 (doi: https://doi.org/10.1109/TCOMM.2015.2415480) 22Liu Q., Zhou S., Giannakis G.B.: 'Cross-Layer combining of adaptive modulation and coding with truncated ARQ over wireless links', IEEE Trans. Wirel. Commun., 2004, 3, (5), pp. 1746– 1755 (doi: https://doi.org/10.1109/TWC.2004.833474) Citing Literature Volume11, Issue4March 2017Pages 514-521 FiguresReferencesRelatedInformation
Referência(s)