Exact throughput analyses of energy-harvesting cooperation scheme with best relay selections under I/Q imbalance

INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 15 |NUMBER: 4 |2017 |SPECIAL ISSUE Exact Throughput Analyses of Energy-Harvesting Cooperation Scheme with Best Relay Selections Under I/Q Imbalance Tan-Phuoc HUYNH1, Pham Ngoc SON2, Miroslav VOZNAK1 1Department of Telecommunications, Faculty of Electrical Engineering and Computer Science, VSB–Technical University of Ostrava, 17. listopadu 15/2172, 708 33 Ostrava, Czech Republic 2Department of Computer and Communication Engineering, Faculty of Electrical and Electronics Engineering, HCM City University of Technology and Education, 01 Vo Van Ngan, Ho Chi Minh City, Vietnam phuoc.h[email protected], [email protected], miroslav.v[email protected] DOI: 10.15598/aeee.v15i4.2302 Abstract. In this paper, we propose an energyharvesting cooperation scheme in which relays suffer In-phase and Quadrature-phase Imbalances (IQI) and harvest energy from a wireless transmit source. A best relay is selected based on end-to-end Signalto-Interference-plus-Noise Ratios (SINRs) in both Amplify-and-Forward (called an EHAF protocol) and Decode-and-Forward (called an EHDF protocol) cooperation methods. We analyze and evaluate the system performance in terms of exact closed-form throughputs over Rayleigh fading channels. Simulation and analysis results discover contributions as follows. Firstly, the throughput performance of the proposed protocols EHAF and EHDF is improved when compared with that of a non-selection cooperation scheme. Secondly, the EHDF protocol is more efficient than the EHAF protocol. Finally, the theoretical analyses are validated by performing Monte Carlo simulations. Keywords Amplify-and-forward, cooperative communication, decode-and-forward, energy harvesting, I/Q imbalance, opportunistic relay selection, outage probability and throughput. 1. Introduction In recent years, widening the range and rising the diversity capacity of wireless communication are effected by cooperative relaying. The cooperative communication supports the data transmission from the wireless source nodes to the destinations. There are two main phases which are broadcast phase and cooperation phase. The source nodes broadcast their data to relays in the first phase. Then, in the next phase, the received signal is forwarded to the destinations by the relays. The data is transferred from the sources to the destinations via the relays which use the following selected cooperative techniques: Amplify-and-Forward (AF) and Decodeand-Forward (DF) [1], [2] and [3]. There are many studied cases in cooperative networks under the impact of In-phase and Quadraturephase Im-balance (IQI) [4] and [5]. The IQI pertains to the phase and/or amplitude mismatch between the Inphase (I) and Quadrature (Q) signals at the Transmitter (TX) and Receiver (RX) sides. Most recent studies about IQI have focused on the performance analysis and baseband compensation for single hop communication systems. In [4], the authors researched a generalized performance analysis of AF dual-hop relaying, where IQI affects both the TX and RX front ends of the relay node. In recent time, there are many studied cases about energy harvesting in cooperative networks [6], [7], [8] and [9]. The researchers in [9] presented the throughput maximization based on the assumptions of both causal and non-causal knowledge of the harvested energy in the energy harvesting two-hop AF relaying network. In [9], the authors investigated harvesting energy of relays from source signals with a best relay selection protocol in which a best relay having the highest harvested energy will forward the received signal towards the destination. Most of the above researchers, the authors have not considered the energy-harvesting cooperation scheme c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 585 INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 15 |NUMBER: 4 |2017 |SPECIAL ISSUE with opportunistic relay selection under I/Q imbalance. In this paper, we consider a dual-hop DF and AF relaying networks with multiple wireless energy harvesting relay nodes under the impact of IQI. The main contributions of the paper are summarized as follows. Firstly, we propose an energy-harvesting cooperation scheme in which relays suffer In-phase and Quadraturephase Imbalances (IQI). A best relay based on maximum end-to-end Signal-to-Interference-plus-Noise Ratio (SINR) is selected in both Amplify-and-Forward (called an EHAF protocol) and Decode-and-Forward (called an EHDF protocol) cooperation methods. Secondly, exact closed-form throughputs over Rayleigh fading channels are derived and are confirmed by Monte Carlo simulations. Thirdly, the proposed EHDF protocol outperforms the proposed EHAF protocol. Finally, the throughput performance of the two protocols is also improved when the number of cooperative relays increases. This paper is organized as follows: Sec. 2. describes a dual-hop DF and AF relaying networks with multiple wireless energy harvesting relay nodes under impact of IQI; Sec. 3. analyzes and calculates the exact the throughput performances of the proposed EHDF and EHAF protocols; the simulation results are presented in Sec. 4. and Sec. 5. summarizes our conclusions. 2. System Model We consider a decode-and-forward and a amplify-andforward relaying model with a source node defined by S, a destination node defined by D and M energyharvesting relays denoted as Riwith i = 1,2,...,M as shown in Fig. 1. Fig. 1: A dual-hop decode-and-forward and amplify-andforward relaying model with M energy-harvesting relays under I/Q imbalance. In Fig. 1, each node is installed with a single antenna, and transmittance is forced in half-duplex mode where transmitting and receiving works can not happen concurrently. Assuming that there is no connection between the source and destination, the transmission signal from S to D is just via relays Ribecause of deep shadowing. In addition, the relays Risuffer In-phase and Quadrature-phase Imbalances (IQI). Operation principle of the proposed scheme is performed in two stages to transfer a single data. In the first stage, the data is transferred from the source S to the relays Ri, after that the relays harvest energy from the RF signals of the source S. Then the signal is moved to the destination in the second stage. The relays Riapply two methods to process the received signals: magnify and transmit to the destination D (called the EHAF protocol), and decode and transmit to the destination D (called the EHDF protocol). In this paper, we select a best relay based on end-to-end Signal-to-Interference-plus-Noise Ratios (SINRs). The mathematical expressions and throughput analyses of the two protocols EHDF and EHAF will be discussed in the next section. 3. Throughput Performance Analyses In this section, let h1iand h2idefine the Rayleigh channel factors of the S−Rilink and the Ri−D link, respectively. Moreover, let n1i∼CN(0, N1)and nD∼CN(0, N2)as the complicated Gaussian noises at the relays Riand D. Throughputs of the proposed protocols EHDF and EHAF are obtained based on [10] as follows: τX= (1 −PX out)(1−α)R/2,(1) where: PX out are outage probabilities of the protocols X, X∈ {EHAF,EHDF};αis a time-switching coefficient, 0< α < 1; and R is a target data rate and is related to a threshold SINR γ0as R= log2(1 + γ0). In the first stage, the signal is transmitted from the source S to the relays Ri. Then, the received baseband signals at the relays Riof the proposed protocols EHAF and EHDF after down shift and under effects of RX I/Q mismatch can be given as yX_SRi=K1(h1ix+n1i) + K2(h1ix+nD)∗,(2) where xis the transmit signal of the source node S with average transfer energy E|x|2=P(Ez is an expectation expression of z), and K1 ∆ = (1 + gejϕT T)/2,(3) K2 ∆ = (1 −gejϕT T)/2.(4) In Eq. (3) and Eq. (4), gTand ϕTcreate the TX magnitude and phase mismatch. The mirror datum introduced by the IQI is often assigned as (h1ix+n1i)∗terms in Eq. (2). From Eq. (2) and [10], the power is gotten from the collected energy c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 586 INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 15 |NUMBER: 4 |2017 |SPECIAL ISSUE in over the time (1−α)T/2for forwarding the processed signal to the destination D as PRi=|K1|2+|K2|22αηP |h1i|2 (1 −α).(5) In the second stage, the signal after is amplified (the EHAF protocol) or decoded (the EHDF protocol), it is transmitted from the relays Rito the destination D. 3.1. The EHAF Protocol In the EHAF protocol, the received signals at the relay Riare typically magnified at the baseband level with a magnification coefficient G, then converted up the Radio Frequency level (RF), and after that are transferred to the destination D. Under TX IQI at the relay Ri, the received baseband signal at the destination D is expressed as yEHAF _RiD=h2i(G1(G yEHAF _SRi) +G∗ 2(G yEHAF _SRi)∗) + nD,(6) where G1 ∆ = (1 + gejϕR R)/2,(7) G2 ∆ = (1 −gejϕR R)/2,(8) G=sPRi F(|h1i|2P+Ni).(9) In Eq. (7) and Eq. (8), gRand ϕRdenote the RX magnitude and phase mismatch. Substituting Eq. (2), Eq. (9) into Eq. (6), and after applying some manipulations, the end-to-end SINR is obtained as γEHAF _RiD=aPPRi|h1i|2|h2i|2 (bPPRi|h1i|2|h2i|2) + aN1PRi|h2i|2+ bN1PRi|h2i|2+N2P|h1i|2+N1N2) (10) where a=A2 F,b=B2 F,c=a+b, A∆ =K1G1+K∗ 2G∗ 2;B∆ =K1G2+K∗ 2G∗ 1, F∆ = (|K1|2+|K2|2)(|G1|2+|G2|2). Substituting the PRiin Eq. (5) into Eq. (10), we obtain the following result: γEHAF _RiD=2aαηP2(|K1|2+|K2|2)|h1i|4|h2i|2 (2bαηP |h1i|4|h2i|2) + 2aαηN1P|h1i|2|h2i|2+ 2bαηN1P|h1i|2|h2i|2+ N2P|h1i|2(1 −α) + N1N2(1 −α). (11) In the proposed EHAF protocol, the best relay Rb1is selected so that the end-to-end SINR γEHAF _Rb1Dis maximize. A selection criterion is expressed as Rb1= arg max i∈{1,2,...,M}γEHAF _RiD.(12) The outage probability of the EHAF protocol is obtained by a math expression as follows PEHAF out = Pr γEHAF _Rb1D< γ0 = Pr max i=1,2,...,M(γEHAF _RiD)< γ0 = M Q i=1 Pr γEHAF _RiD< γ0 = M Q i=1 Pr hw2i<ωw1i+ψ uw2 1i−vw1ii. (13) Applied the proposition in [10], the expression PEHAF out is given as: PEHAF out =  1−λ1 ∞ Z v/u e−(λ1x+λ2ωx+ψ (ux2−vx))dx   M (14) ≈h1−e−λ1v uµK1(µ)iM, (SINR approximation) where ω=γ0N2P(1 −α), ψ=γ0N1N2(1 −α), u= 2aαηP 2|K1|2+|K2|2−2bαηγ0P2, v= 2aαηP −2bαηγ0P. 3.2. The EHDF Protocol From Eq. (2), the SINR at the relays Rito decode the information signal xis obtained as follows yEHDF _SRi=|K1|2P|h1i|2 |K2|2P|h1i|2+ (|K1|2+|K2|2)N1 .(15) After successful decoding, the relays Riwill forward the decoded signal x to the destination D. The received signal at the destination D with the TX IQI at the relays Riis expressed as yEHDF _RiD=G1(h2ix+nD) + G2(h2ix+nD)∗.(16) Then, the SINR γEHDF _RiDat the destination D is obtained as γEHDF _RiD=|G1|2PRi|h2i|2 |G2|2PRi|h2i|2+ (|G1|2+|G2|2)N2 =2αηP |G1|2|h1i|2|h2i|2 2αηP |G2|2|h1i|2|h2i|2+ |G1|2+|G2|2N2(1 −α) . (17) c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 587 INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 15 |NUMBER: 4 |2017 |SPECIAL ISSUE Similar to the EHDF, the best relay Rb2is selected according a criterion as follows Rb2= arg max i∈{1,2,...,M}min γEHDF _SRi, γEHDF _RiD. (18) The outage probability of the EHDF protocol is obtained as PEHDF out = Pr max i∈{1,2,...,M}min γEHDF _SRi, γEHDF _RiD< γ0 = M Q i=1 Pr min γEHDF _SRi, γEHDF _RiD< γ0 = M Q i=1 1−Pr min(γEHDF _SRi, γEHDF _RiD)> γ0 = M Q i=1 Pr    1−Pr (γEHDF _SRi> γ0, γEHDF _RiD)> γ0 | {z } Φ    . (19) The Φterm is calculated as Φ = Pr w1i> m, w2i>o (p−q)w1i =(Pr hw1i> m, w2i>o (p−q)w1ii, p > q Pr [w1i> m], p ≤q =   λ1 ∞ R m e−(λ1x+λ2t x)dx , p > q e−λ1m, p ≤q (20) where m=(|K1|2+|K2|2)N1γ0 (|K1|2−|K2|2γ0)P; w1i=|h1i|2; o=|G1|2+|G2|2N1(1 −α)γ0 p=|G1|22αηP ; q=|G2|22αηP γ0. The probability Φin Eq. (20) contains the complex integrals, and solving of these integrals is not practical. However, we use numerical methods to find values of Φ. 4. Simulation Results In this section, the systems performance of the proposed protocols EHAF and EHDF is analyzed and evaluated using the exact theoretical analyses and the Monte Carlo simulations of the throughput. In the two dimensional plane, the coordinates of S, D, and Riare S(0, 0), D(1, 0) and Ri(x, y), respectively, satisfying 0 < x < 1. Therefore, d1=px2+y2and d2 =p(1 −x)2+y2. The IQI parameters are set to 20 log 10(gT) = 20 log 10(gR)=1.58 dB. In addition, the SINR on the x-axis is defined as γ=P Ni,{i∈1,2}. Figure 2 presents the throughputs of the proposed protocols EHDF and EHAF versus αwhen the symmetric network model is considered with d1 = d2 = 1, SINR = 5 dB, η= 0.9and 0< α < 1. It can be seen that the throughputs at the destination in both protocols EHDF and EHAF increase by time-switching coefficient αand then throughputs decreases when α increases. In addition, these throughputs are largest when values αbecome to an optimal value. For example, in the EHDF protocol, the optimal values are approximately to 0.27 and 0.17 when the number of relays, M, is set to 1 and 3, respectively. α 0 0.2 0.4 0.6 0.8 1 Throughput (bit/s/Hz) 10-3 10-2 10-1 100 EHDF - Simulation EHDF - Simulation Theory EHDF: M=3 EHDF: M=1 EHAF: M=3 EHAF: M=1 Fig. 2: The throughput at the destination in the protocols EHDF and EHAF versus when the IQI parameters: 20 log 10(gT) = 20 log 10(gR)=1.58 dB and ϕT= ϕR= 100,η= 0.9. In EHAF protocol, the optimal value α= 0.39 when M = 1 and α= 0.29 when M = 3. These optimal values can be obtained by the Golden Section Search (GSS) method in [11] with a minimal interval 10−3. The results show that the throughputs ascend by α and achieve the optimal limitation at the ideal value α. Furthermore, because of the effects of IQI noise on the system, the throughput will descend by α. From Fig. 2, the throughput performance of the EHDF protocol is greater than that of the EHAF protocol. In addition, due to the optimal relay selection approach, we have figured out that, the system model with 3 relays achieves the better throughput than the system model with 1 relay in both protocols. Figure 3 presents the throughput performances of the proposed protocols EHDF and EHAF versus SINR (dB) in which αis set to the optimal values αopt_AF and αopt_DF at each value of the SINR (dB), respectively. We have figured that the throughputs go up when the SINRs increase, and the throughput performance of the proposed EHDF protocol outperforms the EHAF protocol. Figure 4 illustrates the throughput performances of the proposed protocols EHDF and EHAF in the asymc 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 588 INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 15 |NUMBER: 4 |2017 |SPECIAL ISSUE SINR(dB) -5 0 5 10 15 20 Throughput (bit/s/Hz) 10-3 10-2 10-1 100 EHDF - Simulation EHAF - Simulation Theory EHDF: M=1 EHAF: M=1 EHAF: M=3 EHDF: M=3 Fig. 3: The throughput at the destination in the protocols EHDF and EHAF versus when the IQI parameters: 20 log 10(gT) = 20 log 10(gR)=1.58 dB and ϕT= ϕR= 100,η= 0.9. metric network scheme as a function of the IQI parameters gT=gR(dB) when ϕT=ϕR= 100,η= 0.9, the IQI parameters values are set to 0 dB and 30 dB, respectively. The parameters αare set to the optimal values αopt_AF and αopt_DF at each value of the gT. It can be seen that when the IQI parameters values increase, the throughput performances of both protocols EHDF and EHAF decrease. gT = gR(dB) 0 5 10 15 20 25 30 Throughput (bit/s/Hz) 10-2 10-1 100 EHDF - Simulation EHAF - Simulation Theory EHAF: M=1 EHAF: M=3 EHDF: M=1 EHDF: M=3 Fig. 4: The throughputs in the protocols EHDF and EHAF versus the IQI parameters gT=gR(dB) and ϕT=ϕR= 100,η= 0.9. Lastly, we can see that the simulation results fit well to the theoretical results. Hence, we can conclude that the derived formulas during analyzing are accurate. 5. Conclusion In this paper, we propose an energy-harvesting cooperation scheme in which relays suffer In-phase and Quadrature-phase Imbalances (IQI) and harvest energy from a wireless transmit source. One best relay is selected based on end-to-end SINRs in both Amplifyand-Forward (called an EHAF protocol) and Decodeand-Forward (called an EHDF protocol) cooperation methods. We have analyzed and evaluated the system performance in terms of exact closed-form throughputs over Rayleigh fading channels. Simulation and analysis results discover that the proposed EHDF protocol achieves higher throughput performance when comparing with the proposed EHAF protocol. In addition, the throughput performance of the two protocols also improved when the number of relays increase. References [1] LI, D. Amplify-and-Forward Relay Sharing for Both Primary and Cognitive Users. IEEE Transactions on Vehicular Technology. 2016, vol. 65, iss. 4, pp. 2796–2801. ISSN 1939-9359. DOI: 10.1109/TVT.2015.2418779. [2] ALVI, S. H. and S. WYNE. Error analysis of fixed-gain AF relaying with MRC over Nakagami-m fading channels. Radioengineering. 2016, vol. 25, iss. 1, pp. 106–113. ISSN 1210-2512. DOI: 10.13164/re.2016.0106. [3] ZHOU, Q. F., W. H. MOW, S. ZHANG and D. TOUMPAKARIS. Two-Way Decode-andForward for Low-Complexity Wireless Relaying: Selective Forwarding Versus One-Bit Soft Forwarding. IEEE Transactions on Wireless. 2016, vol. 15, iss. 3, pp. 1866–1880. ISSN 1536-1276. DOI: 10.1109/TWC.2015.2496949. [4] LI, J., M. MATTHAIOU and T. SVENSSON. I/Q imbalance in two-way AF relaying. In: Global Communications Conference. Austin: IEEE, 2014, pp. 2271–2285. ISBN 978-1-47993512-3. DOI: 10.1109/GLOCOM.2014.7037433. [5] QI, J., S. AISSA and M. S. ALOUINI. Impact of I/Q imbalance on the performance of two-way CSI-assisted AF relaying. In: IEEE Wireless Communications and Networking Conference (WCNC). Shanghai: IEEE, 2013, pp. 2507–2512. ISBN 978-1-4673-5939-9. DOI: 10.1109/WCNC.2013.6554955. [6] MINASIAN, A., S. SHAHBAZPANAHI and R. S. ADVE. Energy harvesting cooperative communication systems. IEEE Transactions on Wireless Communications. 2014, c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 589 INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 15 |NUMBER: 4 |2017 |SPECIAL ISSUE vol. 13, iss. 11, pp. 6118–131. ISSN 1536-1276. DOI: 10.1109/TWC.2014.2320977. [7] LI, T., P. FAN and K. B. LETAIEF. Outage probability of energy harvesting relay-aided cooperative networks over Rayleigh fading channel. IEEE Transactions on Vehicular Technology. 2016, vol. 65, iss. 2, pp. 972–978. ISSN 1939-9359. DOI: 10.1109/TVT.2015.2402274. [8] ZHAI, C and L. LIU. Cooperative wireless energy harvesting and information transfer in stochastic networks. EURASIP Journal on Wireless Communications and Networking. 2015, vol. 2015, iss. 1, pp. 44. ISSN 1687-1499. DOI: 10.1186/s13638-015-0288-3. [9] DO, N. T., V. N. Q. BAO and B. AN. A relay selection protocol for wireless energy harvesting relay networks. In: International Conference on Advanced Technologies for Communications (ATC). Ho Chi Minh City: IEEE, 2015, pp. 243–247. ISBN 978-1-4673-8374-5. DOI: 10.1109/ATC.2015.7388327. [10] NASIR, A. A., X. ZHOU, S. DURRANI and R. A. KENNEDY. Relaying Protocols for Wireless Energy Harvesting and Information Processing. IEEE Transactions on Wireless Communications. 2013, vol. 12, iss. 7, pp. 3622–3636. ISSN 15361276. DOI: 10.1109/TWC.2013.062413.122042. [11] CHONG, E. K. P. and S. H. ZAK. An Introduction to Optimization. 2nd ed. United States: Wiley, 2001. ISBN 978-0471654001. About Authors Tan Phuoc HUYNH was born in Ho Chi Minh City, Vietnam, in 1979. He received the M.Sc. from University of Information Technology in 2008. In 2009, he was lecturer at the School of Engineering, Eastern International University, Binh Duong City. Since 2015, he has been participating in Ph.D. program, that had been linked between Technical University of Ostrava, Czech Republic and Ton Duc Thang University, Ho Chi Minh City. His major research interests are wireless communication in 5G, energy harvesting, performance of cognitive radio and network security. Pham Ngoc SON was born in Ca Mau, Vietnam, in 1981. He received the B.E. degree (2005) and M. Eng. degree (2009) in Electronics and Telecommunications Engineering from Post and Telecommunication Institute of Technology, Ho Chi Minh City and Ho Chi Minh City University of Technology, Vietnam, respectively. In 2015, he received the Ph.D. degree in Electrical Engineering from University of Ulsan, South Korea. He is currently a Lecturer in the Faculty of Electrical and Electronics Engineering of Ho Chi Minh City University of Technology and Education (HCMUTE). His major research interests are cooperative communication, cognitive radio, physical layer security and energy harvesting. Miroslav VOZNAK received his Ph.D. in Telecommunications engineering in 2002 from the Faculty in Electrical Engineering and Computer Science, VSB– Technical University of Ostrava and was appointed as an Associate Professor after his habilitation in the same faculty in 2009. Since 2013, he has been leading a Department of Telecommunications in the VSB–Technical University of Ostrava in position of the department chair. He is an IEEE Senior member, actively enganged as a member in numerous conference programme committees and serving as a member of editorial boards in several journals such as Journal of Communications (US), Advances in Electrical and Electronic Engineering (CZ), Communications (TR), etc. He participated in more than fifteen national and four european projects. His research interests are focused generally on information and communications technology, particularly on Voice over IP, Quality of Experience, Network security, Wireless networks and last several years on Big Data analytics in mobile cellular networks as well. c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 590