TELK OMNIKA T elecommunication, Computing, Electr onics and Contr ol V ol. 19, No. 1, February 2021, pp. 192 198 ISSN: 1693-6930, accredited First Grade by K emenristekdikti, No: 21/E/KPT/2018 DOI: 10.12928/TELK OMNIKA.v19i1.16245 r 192 Enabling full-duplex in multiple access technique f or 5G wir eless netw orks o v er Rician fading channels Chi-Bao Le 1 , Dinh-Thuan Do 2 1 F aculty of Electronics T echnology , Industrial Uni v ersity of Ho Chi Minh City (IUH), V ietnam 2 W ireless Communications Research Group, F aculty of Electrical and Electronics Engineering, T on Duc Thang Uni v ersity , Ho Chi Minh City , V ietnam Article Inf o Article history: Recei v ed Apr 4, 2020 Re vised Jul 6, 2020 Accepted Sep 24, 2020 K eyw ords: Full-duple x Non-orthogonal multiple access Outage probability Unmanned aerial v ehicle ABSTRA CT No w adays, unmanned aerial v ehicle (U A V) relays’ assisted internet of things (IoT) systems pro vide f acility in order to o v ercome the lar ge scale f ading between source and sink. The full-duple x scheme enables wireless netw ork to pro vide higher spectrum ef ficient technology . This paper analyses performance of tw o users which are serv ed by ne w emer ging non-orthogonal multiple access (NOMA) technique. Exact outage probability of such tw o users are deri v ed and check ed via Monte-Carlo simulation. These analytical res ults pro vide guideline to design U A V in real application. This paper pro vides a comprehensi v e study to e xamine impact of interference, fix ed po wer allocation f actors to system performance. This is an open access article under the CC BY -SA license . Corresponding A uthor: Dinh-Thuan Do W ireless Communications Research Group F aculty of Electrical & Electronics Engineering T on Duc Thang Uni v ersity Ho Chi Minh City , V ietnam Email: dodinhthuan@tdtu.edu.vn 1. INTR ODUCTION State of the art, wireless netw ork pro vides ability to serv e massi v e connections and such require- ment satisfied by the application of non-orthogonal multiple access (NOMA) in fifth generation (5G) netw orks. NOMA has dra wn wide attention due to its potential to impro v e spectral ef ficienc y [1] and more reliable im- pro v ement with relaying sc h e me [2]. Dif ferent from con v entional orthogonal mult iple access (OMA), NOMA benefits from relaying design, f ar user can be serv ed by the base station under help of the relay . NOMA enables v arious applications to serv e mult iple users to be serv ed at the same time and frequenc y by superimposing mul- tiple users in the po wer domain at the transmitter and using successi v e interference cancellation (SIC) at the recei v er [3, 4]. In NOMA system, de vice-to-de vice transmission mode is acti v ated based on the users di vided into dif ferent kinds of cate gories according to their channel conditions, i.e., the near user and the f ar user [5]. In the conte xt of NOMA, cooperati v e mode is mainly di vided into tw o cases. These cases e xhibit implementation of NOMA to imrpo v e performance of users locat ed edge of cell [6–9]. In cooperati v e NOMA, the near users with strong channel conditions needs to act as relays. The relay pro vides impro v ed performance of the f ar users who ha v e poor channel conditions [9, 10]. Ho we v er . Ho we v er , half-duple x (HD) mode is studied in the coop e rati v e NOMA relay in the early w orks [7–11]. The authors in [6] considered full-duple x (FD) relay into cooperati v e NOMA. The main adv antages are reducing delay caused by the dedicated relay J ournal homepage: http://journal.uad.ac.id/inde x.php/TELK OMNIKA Evaluation Warning : The document was created with Spire.PDF for Python.
TELK OMNIKA T elecommun Comput El Control r 193 and enhancing end to end transmission quality . The authors in [12] in v estig ated adv antage achie v ed by FD mode, which pro v ed enhancing performance g ain. While the authors in [13] maximized ener gy ef ficienc y for full-duple x cooperati v e NOMA with po wer allocation. Ho we v er , open problem still e xists related to f ading model. This paper fulfills a g ap in [13–22], in which U A V -based relay is not considered. 2. SYSTEM MODEL This paper considers a tw o-user NOMA architecture, where UE-1 directly e xchanges data with the base station (BS) as depicted in Figure 1, while UE-2 recei v es signals from the BS via unmanned aerial v ehicle (U A V) relay . The link BS to UE-2 is supported by a dedicated relay . Note that each node is equipped single antenna e xcept for relay which requires tw o antenna to pro vide ability of FD. The probabilistic LoS and non- LoS (NLoS) model for U A V is adopted to indicate the lar ge scale f ading. W e use such model for the channel between the U A V and terrestrial user due to impact of the density of b uildings and the distance between the U A V and users. The probability of user which has benefit of a LoS link is e xpressed as [24] P LoS ;k = 1 1+ pe q ( k p ) ; k 2 f 1 ; 2 g (1) in which we denote p and q are constant v alues depending on the surrounding en vironment (sub-urban, urban, dense-urban). Therefore, k in (1) can be e xpressed as k = arcsin H d k ; (2) in which H denotes the height of U A V , d k = p r 2 k + H 2 is the dis tance between user k and the U A V , and r k is the distance between users and U A V . Ob viously , the probability of NLoS is P N LoS ;k = 1 P LoS ;k .   S i g n a l   l i n k   I n t e r f e r e n c e   l i n k 0 h 1 h 2 h 1 g B S B S U E - 2 U E - 1 R 0 g Figure 1. Enabling FD mode in relaying netw ork The recei v ed signal at the relay is e xpressed by y R = h 0 2 X k =1 q d S ' k P S s k + p P R g 0 s LI 0 + n R ; (3) where , 0 1 denotes as le v el of self-interference (LI), ' k ; k = 1 ; 2 are po wer allocation f actors to tw o NOMA users who need recei v e signal s k , P S is transmit po wer of the BS, s LI is self-interference signal due to FD mode, d S = p r 2 S + H 2 is the distance between B S and U A V , and is the path loss e xponent from the B S to U A V . The signal to interference plus noise (SINR) to detect signal s 2 is gi v en by s 2 R = ' 2 d S j h 0 j 2 ' 1 d S j h 0 j 2 +  j g 0 j 2 + 1 ; (4) Enabling full-duple x in multiple access tec hnique for 5G wir eless networks... (Chi-Bao Le) Evaluation Warning : The document was created with Spire.PDF for Python.
194 r ISSN: 1693-6930 where = P S / N 0 = P R / N 0 is the transmit signal-to-noise radio (SNR). Emplo ying SIC, interference s 2 is deleted to detect s 1 corresponding SINR as s 1 R = ' 1 d S j h 0 j 2  j g 0 j 2 + 1 : (5) Then, the recei v ed signal at user UE-1 is gi v en as y U E 1 = h 1 2 X k =1 q d 1 ' k P S s k + q P L 1 d 1 P R g 1 s LI 1 + n U E 1 (6) where d 1 is the distance between BS to UE-1, P L k = ( P LoS ;k + P N LoS ;k ) , k 2 f 1 ; 2 g and denotes the additional attenuation f actor of NLoS transmission. The SINR to detect s 2 and then s 1 at UE-1 are respecti v ely e xpressed by s 2 U E 1 = d 1 ' 2 j h 1 j 2 d 1 ' 1 j h 1 j 2 + P L 1 d 1 j g 1 j 2 + 1 ; (7a) s 1 U E 1 = ' 1 d 1 j h 1 j 2 P L 1 d 1 j g 1 j 2 + 1 : (7b) The recei v ed signal and SNR at user UE-2 is formulated respecti v ely as ~ y U E 2 = h 2 q P L 2 d 2 P R ~ s 2 + n U E 2 : (8) ~ s 2 U E 2 = P L 2 d 2 j h 2 j 2 : (9) 3. OUT A GE PR OB ABILITY AN AL YSIS Case 1: 0 < < 1 , the outage probability without interference link of UE-1 is gi v en by O P U E 1 =1 Pr s 2 U E 1 > " 2 \ s 1 U E 1 > " 1 =1 Pr j h 1 j 2 > j g 1 j 2 + 1  ; (10) where = P L 1 d 1 , = max " 2 ( ' 2 ' 1 " 2 ) ; " 1 ' 1 , " 1 = 2 2 R 1 1 with R 1 is denoted as the tar get rate at UE-1 to detect s 1 and " 2 = 2 2 R 2 1 with R 2 being the tar get rate at UE-2 to detect s 2 . Thus, the probability distrib ution function (PDF ) of the unordered squared channel g ain X , X 2 f h 0 ; h 1 ; h 2 ; g 0 ; g 1 g , is formulated by a non-central chi-square distrib ution with tw o de grees of freedom as [25] f j X j 2 ( x ) = (1 + K X ) e K X ( 1+ K X ) x X X I 0 0 @ 2 s K X (1 + K X ) x X 1 A ; (11) where I 0 ( : ) is the zeroth-order modified Bessel function of the first kind, K X = j X j 2 2 2 is the Rician f actor and X = E n j X j 2 o = 1 is the normalized a v erage f ading po wer . The corresponding cumulati v e distrib ution function (CDF) is kno wn as F j X j 2 ( x ) = 1 Q 0 @ p 2 K X ; s 2 (1 + K X ) x X 1 A ; (12) where Q ( ; ) = R 1 xe 2 + x 2 2 I 0 ( ax ) denotes the MarcumQ-function of first order . By using result included in [23, Eq. (8.445)] with I 0 ( z ) = P 1 r =0 z 2 r r !( r +1)2 2 r , the O P U E 1 is calculated as TELK OMNIKA T elecommun Comput El Control, V ol. 19, No. 1, February 2021 : 192 198 Evaluation Warning : The document was created with Spire.PDF for Python.
TELK OMNIKA T elecommun Comput El Control r 195 O P U E 1 =1 Pr j h 1 j 2 > j g 1 j 2 + 1  =1 1 Z 0 f j g 1 j 2 ( x ) 1 Z ( x +1) f j h 1 j 2 ( y ) dxdy =1 1 X q =0 1 X b =0 K b g 1 K q h 1 b +1 g 1 q +1 h 1 e ( K h 1 + K g 1 ) b ! q ! ( b + 1) ( q + 1) 1 Z 0 x b e ( 1+ K g 1 ) g 1 x 1 Z ( x +1) y q e ( 1+ K h 1 ) h 1 y dxdy ; (13) where g 1 = ( 1+ K g 1 ) g 1 , h 1 = ( 1+ K h 1 ) h 1 and ( ) is the Gamma function [23, Eq. (8.310)]. W ith the help of [23, Eq. (3.324.1)], [23, Eq. (1.111)], [23, Eq. (3.324.3)] we can further simplify the abo v e as O P U E 1 = 1 1 X q =0 1 X b =0 q X j =0 j X p =0 j p q ! ( b + p )! p j K b g 1 K q h 1 b +1 g 1 j h 1 b ! q ! j ! ( b + 1) ( q + 1) ( g 1 + h 1 ) b + p +1 e h 1 ( K h 1 + K g 1 ) + ( 1+ K h 1 ) h 1 : (14) Case 2: = 0 , the outage probability without interference link of UE-1 is gi v en by O P U E 1 =1 Pr j h 1 j 2 > = F j h 1 j 2 =1 Q 0 @ p 2 K h 1 ; s 2 (1 + K h 1 ) h 1 1 A : (15) In particular , the outage probability with impact of interference at UE-2 is gi v en by O P U E 2 =1 Pr ( s 2 R > " 2 \ s 1 R > " 1 ) Pr ~ s 2 U E 2 > " 2 =1 Pr j h 0 j 2 > ~  j g 0 j 2 + 1  | {z } 1 Pr j h 2 j 2 > " 2 P L 2 d 2 | {z } 2 ; (16) where ~ = max " 2 d S ( ' 2 " 2 ' 1 ) ; " 1 ' 1 d S . Similarly with solving (13), it can be achie v ed 1 as 1 = 1 X r =0 1 X a =0 r X q =0 q X w =0 q w ( a + w )! w w ~ q K a g 0 K r h 0 a +1 g 0 q h 0 a ! q ! ( a + 1) ( r + 1) ( g 0 + h 0 ) a + w +1 e h 0 ( K h 0 + K g 0 ) + ~ ( 1+ K h 0 ) h 0 ; (17) where g 0 = ( 1+ K g 0 ) g 0 and h 0 = ( 1+ K h 0 ) h 0 . Ne xt, 2 is calculated as 2 = Pr j h 2 j 2 > " 2 P L 2 d 2 = 1 Z " 2 P L 2 d 2 f j h 2 j 2 ( x ) dx = 1 X c =0 K c h 2 c h 2 (1 + K h 2 ) e K h 2 c ! ( c + 1) h 2 1 Z " 2 P L 2 d 2 x c e h 2 x dx; (18) Enabling full-duple x in multiple access tec hnique for 5G wir eless networks... (Chi-Bao Le) Evaluation Warning : The document was created with Spire.PDF for Python.
196 r ISSN: 1693-6930 where h 2 = ( 1+ K h 2 ) h 2 . Based on [[23], Eq. (3.351.2)], 2 is gi v en by 2 = 1 X c =0 K c h 2 c h 2 (1 + K h 2 ) e K h 2 c ! ( c + 1) h 2 c +1 h 2 c + 1 ; h 2 " 2 P L 2 d 2 ; (19) where ( :; : ) is the upper incomplete Gamma function [[23], Eq. (8.350.2)]. Substituting (18) and (16) into (15), O P U E 2 is gi v en by O P U E 2 =1 1 X r =0 1 X a =0 r X q =0 q X w =0 1 X c =0 q w ( a + w )! w w ~ q K a g 0 K r h 0 a +1 g 0 q h 0 a ! q ! ( a + 1) ( r + 1) ( g 0 + h 0 ) a + w +1 K c h 2 c h 2 (1 + K h 2 ) e K h 2 c ! ( c + 1) h 2 c +1 h 2 c + 1 ; h 2 " 2 P L 2 d 2 e h 0 ( K h 0 + K g 0 ) + ~ ( 1+ K h 0 ) h 0 : (20) 4. NUMERICAL RESUL TS T o perform simulations, we set K = K h 0 = K h 1 = K h 2 = K g 0 = K g 1 = 2 and = h 0 = h 1 = h 2 = g 0 = g 1 = 1 ; po wer allocation f actors are ' 1 = 0 : 2 and ' 2 = 0 : 8 ; tar get rates are R 1 = 1 and R 2 = 0 : 5 ; coef ficient related to SI from FD is = 0 : 01 . P ath loss e xponent is = 2 , the height of U A V H = 30 m , additional attenuation f actor is = 20 (dB), en vironment parameter is p = 4 : 8860 , en vironment parameter is q = 0 : 4290 . The times of Monte Carlo simulation 10 6 , d 1 = 0 : 7 . In Figure 2, outage performance of user UE-2 is better than that of UE-1 at numerous case of Rician f ading parameters. It is v aluable as well- matching between Monte-Carlo and analytical simulations. At higher SNR, impro v ed outage performance can be seen. As illustration in Figure 3, it is e xistence of optimal outage performance of user UE-1 as v arying a 2 from 0.5 to 1. It can be further seen that lo wer SI leads to better outage performance at tw o users. While Figure 4 indicates that im pro v ement outage performance happens at higher v alue of K related to Rician f ading channel. -10 0 10 20 30 40 50 10 -3 10 -2 10 -1 10 0 K = 1 K = 4 Figure 2. Outage probability v ersus SNR TELK OMNIKA T elecommun Comput El Control, V ol. 19, No. 1, February 2021 : 192 198 Evaluation Warning : The document was created with Spire.PDF for Python.
TELK OMNIKA T elecommun Comput El Control r 197 0.5 0.6 0.7 0.8 0.9 1 10 -3 10 -2 10 -1 10 0  = 15, 25 (dB)  = 25 (dB)  = 15 (dB) Figure 3. Impact of po wer allocation f actor a 1 on outage performance, with K = 2 1 2 3 4 5 6 7 8 9 10 10 -3 10 -2 10 -1 10 0 Figure 4. Impact of K on outage performance, with = 15 (dB) 5. CONCLUSION In this paper , we ha v e discussed ho w impacts of Rician f ading and full-duple x mode ha v e been im- plemented in the relaying netw ork. W e ha v e e xplored se v eral e xact e xpressions of outage probability to sho w performance of each us er . This findings benefit to deplo yment of full-duple x to pro vide higher spectrum ef fi- cienc y for cellular netw ork in practice. REFERENCES [1] S. M. R. Islam, N. A v azo v , O. A. Dobre, and K. S. Kw ak, ”Po wer -domain non-orthogonal multiple access (NOMA) in 5G systems: Potentials and challenges, IEEE Commun. Surv e ys T uts. , v ol. 19, no. 2, pp. 721-742, 2017. [2] P . M. Nam, D. T . Do, T . T . Nguyen, P . T . T in, ”Ener gy harv esting assiste d cogniti v e radio: random location-based transcei v ers scheme and performance analysis, T elecommunication Systems , v ol. 67, no. 1, pp. 123-132, 2018. Enabling full-duple x in multiple access tec hnique for 5G wir eless networks... (Chi-Bao Le) Evaluation Warning : The document was created with Spire.PDF for Python.
198 r ISSN: 1693-6930 [3] D. T . Do and A. T . Le, ”NOMA based cogniti v e relaying: T ranscei v er hardw are impairments, relay selection policies and outage performance comparison, Computer Communications , v ol. 146, pp. 144-154, 2019. [4] D. T . Do, M. V aezi and T . L. Nguyen, ”W ireless Po wered Cooperati v e Relaying using NOMA with Imperfect CSI, in Proc. of IEEE Globecom W orkshops (GC Wkshps) , Ab u Dhabi, U AE, pp. 1-6, 2018. [5] Dinh-Thuan Do and M. S. V an Nguyen, ”De vice-to-de vice transmission modes in NOMA netw ork with and without W ireless Po wer T ransfer , Computer Communications , v ol. 139, pp. 67-77, May 2019. [6] C. Zhong and Z. Zhang, ”Non-orthogonal multiple access with cooperati v e full-duple x relaying, IEEE Commun. Lett. , v ol. 20, no. 12, pp. 2478-2481, 2016. [7] J. B. Kim and I. H. Lee, ”Non-orthogonal multiple access in coordinated direct and relay transmission, IEEE Com- mun. Lett. , v ol. 19, no. 11, pp. 2037-2040, 2015. [8] W . Duan, M. W en, Z. Xiong, and M. H. Lee, ”T w o-stage po wer allocation for dual-hop relaying systems with non- orthogonal multiple access, IEEE Access , v ol. 5, pp. 2254-2261, 2017. [9] Z. Ding, M. Peng, and H. V . Poor , ”Cooperati v e non-orthogonal multiple access in 5G systems, IEEE Commun. Lett. , v ol. 19, no. 8, pp. 1462-1465, 2015. [10] R. Sun, Y . W ang, X. W ang, and Y . Zhang, ”T ranscei v er design for cooperati v e non-orthogonal multiple access systems with wireless ener gy transfer , IET Commun. , v ol. 10, no. 15, pp. 1947-1955, 2016. [11] D. Niyato and E. Hossain, ”A g ame-theoretic approach to competiti v e spectrum sharing in cogniti v e radio netw orks, Proc. IEEE W ireless Commun. Netw . Conf. , pp. 16-20, 2007. [12] Z. Zhang, Z. Ma, M. Xiao, Z. Ding, and P . F an, ”Full-duple x de vice-to-de vice aided cooperati v e non-orthogonal multiple access, IEEE T rans. V eh. T echnol. , v ol. 66, no. 5, pp. 4467-4471, 2017. [13] Z. W ei, X. Zhu, S. Sun, J. W ang, and L. Hanzo, ”Ener gy-ef ficient full-duple x cooperati v e non-orthogonal multiple access, IEEE T rans. V eh. T echnol. , v ol. 67, no. 10, pp. 10123-10128, 2018. [14] D-T . Do et al. ”W ireless po wer transfer enabled NOMA relay systems: tw o SIC modes and performance e v aluation, TELK OMNIKA T elecommunication Computing Electronics and Control , v ol. 17, no.6, pp. 2697-2703, 2019. [15] Dinh-Thuan Do, Chi-Bao Le and A.-T . Le, ”Cooperati v e underlay cogniti v e radio assisted NOMA: secondary net- w ork impro v ement and outage performance, TELK OMNIKA T elecommunication Computing Electronics and Con- trol , v ol. 17, no. 5, pp. 2147-2154, 2019. [16] Dinh-Thuan Do and T .-T . Thi Nguyen, ”Exact Outage Performance Analysis of Amplify-and F orw ard-A w are Coop- erati v e NOMA, TELK OMNIKA T elecommunication Computing Electronics and Control , v ol. 16, no. 5, pp. 1966- 1973, 2018. [17] Dinh-Thuan Do and C.-B. Le, ”Exploiting Outage Performance of W ireless Po wered NOMA, TELK OMNIKA T elecommunication Computing Electronics and Control , v ol. 16, no. 5, 1907-1917. [18] T . L. Nguyen and Dinh-Thuan Do, ”Exploiting Impacts of Intercell Interference on SWIPT -assisted Non-orthogonal Multiple Access, W ireless Communications and Mobile Computing , v ol. 2018, no. 17, 2018. [19] D. T . D o, M.-S. V . Nguyen, ”Outage probability and er godic capacity analysis of uplink NOMA cellular netw ork with and without interference from D2D pair , Ph ysical Communication , v ol. 37, 2019. [20] Dinh-Thuan Do, A. Le and B. M. Lee, ”NOMA in Cooperati v e Underlay Cogniti v e Radi o Netw orks Under Imperfect SIC, IEEE Access , v ol. 8, pp. 86180-86195, 2020. [21] D. T . Do, A. T . Le and B. M. Lee, ”On Performance Analysis of Underlay Cogniti v e Radio-A w are Hybrid OMA/NOMA Netw orks with Imperfect CSI, Electronics , v ol. 8, no. 7, pp. 819-839, 2019. [22] D. T . Do, A. T . Le, C. B. Le and B. M. Lee, ”On Exact Outage and Throughput Performance of Cogniti v e Radio based Non-Orthogonal Multiple A ccess Netw orks W ith and W ithout D2D Link, Sensors , v ol. 19 no. 15, pp. 3314-3330, 2019. [23] I. S. Gradshte yn and I. M. Ryzhik, T able of Inte gr als, Series and Pr oducts , 6th ed. Ne w Y ork, NY , USA: Academic Press , 2000. [24] Aitong Han, T iejun Lv and Xue wei Zhang, ”Outage Performance of NOMA-based U A V -Assisted Communication with Imperfect SIC, IEEE W ireless Communications and Netw orking Conference (WCNC) , April 2019. [25] P . K. Sharma and D. I. Kim, ”U A V -Enabled Do wnlink W ireless System with Non-Orthogonal Multiple Access, IEEE Globecom W orkshops (GC Wkshps) , Sing apore, pp. 1-6, 2017. TELK OMNIKA T elecommun Comput El Control, V ol. 19, No. 1, February 2021 : 192 198 Evaluation Warning : The document was created with Spire.PDF for Python.