1. Introduction
Cognitive radio (CR) is regarded as a promising technology to provide high bandwidth to mobile users via heterogeneous wireless network architectures and dynamic spectrum access techniques, and the significant improvement of efficiency of spectrum utilization [1]-[5].
In CR networks, however, the CR users must immediately vacate the licensed spectrum bands they are using once the primary users return to access those bands. As a result, the interruptive transmissions in unlicensed users will lead to a discontinuous data service and intolerable delay. Challenges aforementioned can be addressed by using cognitive relaying, which enables distributed cognitive users to collaborate with each other and share their distinct spectrum bands. A seamless data transmission can therefore be realized by using cooperative relays, and a mutual benefit among CR users can be brought forward by communicating with each other to minimize the miss detection probability. Besides, more benefits can be brought by using cognitive relays, such as an expanded coverage, a better immunity against signal fading, a more system-wide power saving and an increased throughput of the whole system [6]-[9].
The diversity gain of cooperative networks can thus be improved by employing multiple relays. However, in the presence of multiple relays, a sophisticated resource allocation algorithm is required to perform so as to guarantee an orthogonal channel (in carrier frequencies, time slots or codes) being allocated to each relay and avoid inter-relay-interference [10], [11]. As the number of relays increases, the cost of bandwidth penalty may even deteriorate the benefits brought by cooperative relays. In consideration of the challenges aforementioned, relay selection can be regarded as one of the most attractive methods to solve the complicated interference-mitigation issue met in the multi-relay systems. Selecting the best relay to forward data is an ideal way to effectively balance the complexity and the spectral efficiency improvement due to the full diversity order obtained [12], [13]. Only two time slots are thus required to enable two orthogonal channels (regardless of the number of relays) to be necessarily reserved in both the source and the selected relay nodes [14].
Currently, relay selection has been treated in several existed literatures (see, e.g., [15]-[17] and the references therein). Most of those works deal with the case where the primary users are absent for the relay selection. By selecting the optimal relay from multiple candidates, only one relay is active in each time slot so as to significantly improve the spectral efficiency and obtain a full spatial diversity order simultaneously. By and large, relay selection techniques in cooperative communications systems can be classified into two modes, i.e., the opportunistic relay selection [18] and the partial relay selection [19]. In the former, the channel state information (CSI) of both the source-to-relay (S→R) and relay-to-destination (R→D) links should be considered by the central unit. In the latter, on the other hand, the CSI of only the S→R or R→D link is necessarily considered. In [20], the closed-form expressions for the lower bound and asymptotic symbol error rate of in Rayleigh fading relay channel with opportunistic relay selection are derived. The impact of relay selection with an outdated CSI on the capacity of cooperative communication systems has been studied in [21], [22], where the closed-form expressions for channel capacity of opportunistic relay selection with four classical adaptive transmission techniques are given out by [21], and that with partial relay selection has been studied in [22]. The opportunistic relay selection is proven to have an advantage over the partial relay selection if the SNR of links is relative high and a perfect CSI is assumed at the same time.
It is worth pointing out that the relay selection in cognitive networks got a considerable attention in recent years, with some important factors, such as mutual interference between primary users and secondary users, being taken into consideration in CR relay network [23]-[26]. The impact of interference threshold on the amplify-and-forward (AF) underlay cognitive networks has been studied in [27]. In [28], which focuses on exploiting the transmission opportunity in CR networks, cooperative communication appears to be a promising approach to improve the throughput of the secondary users by increasing the spatial and spectrum diversity orders. The author in [29] analyzes the outage probability of AF relaying cognitive networks, in which the best relay is selected based on either full CSI or partial CSI feedback. The BER of AF systems with partial relay selection in CR network is study in [30]. Furthmore, some efficient spectrum sensing schemes are also proposed in [31]-[33]. However, to the best of the our knowledge, the impact of spectrum sensing process and spectrum utilization efficiency of primary users on the overlay CR relay networks with partial relay selection has still not been studied considerably in prior works.
In this paper, partial relay selection in DF-based CR networks is studied. One of main contributions in this paper is to derive the exact closed-form expressions of some figures of merit, such as the outage probability, BER and the average channel capacity over i.i.d. Rayleigh fading channels, for the proposed relay selection mode. In particular, parameters aforementioned are proven to be dependent on spectrum sensing process as well as the spectrum utilization efficiency of primary users. Besides, a full diversity order is proven to be obtained at low SNR regime, regardless of the spectrum utilization efficiency of the primary users. At high SNR regime, on the other hand, the diversity order keeps unchanged, regardless of the number of potential relays. The validity of the proposed theoretical analysis is also proven by simulations.
The remainder of this paper is organized as follows. Section 2 introduces the system model of partial relay selection in cognitive radio relay networks. The probability density function (PDF) and cumulative distribution function (CDF) of received SNR at destination are introduced in section 3. The closed-form expressions of some critical parameters, including outage probability, BER, and average channel capacity, are derived in section 4. Section 5 gives out the numerical results. Finally, section 6 concludes this paper.
Notation: ℜ{x} and ℑ{x} are the real and imaginary part of x, respectively. γab represents the SNR of a→b link. fX(.) and FX(.) represent the PDF and CDF of the random variable (RV) X, respectively. Pr(.) denotes the probability.
2. System Model
In this section, a CR relay network, which consists of a secondary source terminal (S), M half-duplex DF potential relays (as denoted by the set of Ω={Ri, i=1,2,…M}), and a destination terminal (D) (please see Fig. 1), is considered. S communicates with D using a time division multiple access (TDMA) arrangement. Hence, two time slots are needed to complete one communication session. In the first time slot, S broadcasts the packets to the relays. In the second time slot, each potential relay decides whether to become a candidate for the best relay selection by examining two conditions, decoding of the received signal and availability of the SH. After the best relay selection, the selected relay forward a regenerated replica to D. At the end of each communication session, D combines the received signals using maximal ratio combining (MRC).
Fig. 1.A single-hop cooperating network with multi-relays in overlay cognitive radio relay networks
Without loss of generality, the wireless channels of S→Ri, Ri→D and S→D links are assumed to be i.i.d. Rayleigh distributed RVs. Therefore, we have the circularly symmetric complex Gaussian channel gain hab (a, b∈{S, Ri, D}) with mean zero and variance σ2 between node a and b, which is denoted by hab ~CN (0,σ2). In brief, we assume that all nodes transmit with unit power and the additive white Gaussian noise (AWGN) power is N0, and the effective SNR is given by From [34, Eq.237)], γab can be formulated as a chi-square (χ2) random variable with 2 degrees of freedom, and its PDF is given by where denotes the average SNR of a→b link. Accordingly, the average SNR of each S→Ri link can be represented as . The average SNR of Ri→D and S→D links can be denoted by respectively.
3. PDF and CDF of the Relay-Selection Channel
In this section, two conditions, i.e., the potential relay condition and the relay selection rule, will be analyzed, and afterwards, the PDF and CDF of the proposed relay selection channels are given out.
3.1 Potential Relay Condition
The relay may be a in the set of potential relay,Ω, only if it has successfully decoded the received packets and acquired the SH.
1) Successful Decoding: In order to decode successfully, the target rate r, measured in bits per second per Hertz, is lower than the mutual information, the successful decoding probability can be written as
where γSRi and denote the SNR and the average SNR of S→Ri link, respectively, and γth=22r−1 is the threshold SNR.
2) Successful Acquisition of SH: When the primary users is absent with probability, Pr[H0PU], and present with probability, Pr[H1PU], the opportunity of SH is obtained by Ri with probability Pai. The probability of Ri successful acquisition of SH can be rewritten as
where the first term denotes, the probability of the secondary users obtain the SH when the primary user is absent with false alarm probability, Pf,i; the second term denotes, the probability of the secondary users obtain the SH when the primary user is present with detection probability, Pde,i.
Therefore, if the Ri have successful spectrum acquisition probability Pai and successful decoding probability Pdi, the relay become successful potential relay with probability Qi, Pr[Ri∈Ω]=PdiPdi.
Since all the S→Ri links is i.i.d. Rayleigh channels, Therefore, Pd1=Pd2=…=PdM=Pd. Likewise, Pa1=Pa2=…=PaM=Pa. Hence, Q1=Q2=…QM=Q. In this case, Pr[|Ω|=m], becomes
where |Ω| is the cardinality of Ω.
3.2 Relay Selection Rule
In this subsection, for the Rayleigh fading on the Ri→D links is i.i.d., the average SNR in each the Ri→D links is given by , and is the same as the average SNR of S→D link. In other words, Two scenarios, i.e., |Ω|=0 and |Ω|≥1, will be analyzed separately as follows.
1) |Ω|=0:When the set of Ω is null, φ (with no elements content the condition of potential relay), the secondary users only have the S→D link. Hence, the effective SNR at the destination can be given as γd=γSD. In this case, the PDF of γd becomes
where is the average SNR of S→D link.
Evidently, (4) leads to
2) |Ω|≥1: The secondary user selects the optimal relay according to the following rule, i.e.,
From Appendix I (18), with the partial relay selection rule (6), the conditional PDF of γRkD can be derived as
In the presence of cooperative relays, the spatial diversity order can be improved by utilizing the relay-forward links together with the S→D link. Some combining method, e.g., MRC, can be employed to the destination to optimize the effective SNR as γd=γRkD+γSD.
From Appendix I (19), in this case, the conditional PDF of γd for a given |Ω| can be derived as
Evidently, (8) leads to (the detail please see Appendix II)
Hence, the uncondition CDF of the received SNR is derived as
4. Performance Analysis
In this section, the closed-form expressions for the outage probability, BER and channel capacity will be derived.
4.1 Outage Probability Analysis
For a pre-set threshold γth, from (10), the outage probability of the partial relay selection scheme can be derived as
4.2 Bit Error Ratio Analysis
Using [34, Eq.230)], [35, Eq.4)], the BER of the proposed partial relay selection can be derived as
1)|Ω|=0:
2)|Ω|≥1: with aid of [36, Eq31], the result becomes
Hence, the unconditional BER is derived as
and η are specified modulation constants determined by modulation format (e.g., for phase shift keying (PSK) modulation, we have η=2[37]).
4.3 Channel Capacity Analysis
From Appendix III, the closed-form expression for the channel capacity of the proposed partial relay selection can be given as
where B stands for the signal bandwidth.
5. Numerical Results
In this section, we study the performance of some of the derived closed-form expressions through numerical evaluation as well as using Monte Carlo simulation. Binary phase shift keying (BPSK) modulation is considered in this paper, and this implies η=2. Let us start with some figure of merit, including the outage probability, BER and the average channel capacity over i.i.d. Rayleigh fading channels which implies A normalized system bandwidth is assumed.
The outage probability as a function of γ for different number of potential relays, M=0, 1, 2, 3, 4, 5, 6, is depicted in Fig.2. At low average SNR, since with larger number of relays, the destination has more opportunity to correct decode the received data. Hence, the spatial diversity order is increasing with the number of potential relays. At high average SNR, since the relay can decoded the received data packets correctly regardless of the number of potential relays, two copies are transmitted to the destination. Hence, the diversity order keeps two with opportunity, and the slope of curves keeping same as in the high average SNR values are shown in Fig. 2. However, the number of candidate relays which acquire SH and decode the received packets successfully is increased by increasing the number of relays. The more optimum relays can be chose from the candidate relays set with larger number of elements. Hence, the outage probability performance is improved by the larger number of relays.
Fig. 2.Outage probability versus the average SNR of the links for different values of M using simulations and analytical results
By keeping M=5 unchanged, for different Pr[H1PU], a smaller Pr[H1PU] implies a larger probability of acquisition SH, and the outage probability is therefore a monotonically increasing function of Pr[H1PU], as shown in Fig. 3. Note that Pr[H1PU]=0 implies the primary user is absent all the time. If the relays can decode successfully, it can be the candidate relays. For all Pr[H1PU]>0 Hscenarios, the primary user is presented occasionally.
Fig. 3.Outage probability versus the average SNR of the links for different values of Pr[H1PU] using simulations and analytical results
BER as a function of average SNR is illustrated in Fig. 4. For a specific average SNR of links, spectrum sensing process and spectrum utilization efficiency of primary users, more potential relays implies a better BER performance due to an improved relay link. However, similar to the properties exhibited in outage probability performance, the slope of curves is the same when average SNR goes beyond a certain threshold.
Fig. 4.BER versus the average SNR of the links for different values of M using simulations and analytical results
The effect of the probability of the primary users is present Pr[H1PU] on BER performance of the partial relay selection is also illustrated in Fig. 5. Like in the outage probability performance, BER is also a monotonically increasing function of Pr[H1PU]. At low SNR-regime, the diversity order is impacted by the number of relays, because the cooperative networks with more relays have more opportunities to successfully delivery the data. However, at high SNR-regime, the data can be successfully decoded and forwarded, with the diversity order being uncorrelated to the number of relays. Hence, the slopes for the BER-curves in terms of SNR exhibit some changes in Fig. 5.
Fig. 5.BER versus the average SNR of the links for different values of Pr[H1PU] using simulations and analytical results
The average channel capacity as a function of average SNR for the proposed relay selection scheme is described in Fig. 6.The channel capacity is a monotonically increasing function of the number of available relays. Note that M=0 implies the secondary users only have the S→D links. For all M>0 scenarios, if the relay can decode successfully, it could be the candidate relays, and the secondary users will have the opportunity to use the relay. Hence, the capacity can be improved by the diversity order. At high average SNR, since the diversity order remain unchanged. Hence, with large number of M, further increasing M may cause the average channel capacity increase slightly. When the noise significantly impacts the performance. Therefore, it appears some gap between analytic and simulation results for capacities at low SNR-regime, as shown in Fig. 6. In fact, if linear coordinate instead of logarithmic coordinate was utilized, the gap would appear to be very little.
Fig. 6.Average channel capacity versus the average SNR of the links for different values of M using simulations and analytical results
6. Conclusion
The performance of DF cooperative cognitive networks was studied, with the exact closed-form expressions, including the outage probability, the BER and the channel capacity, being derived. The validity of the proposed theoretical close-form expression on the critical figures of merit, including the outage probability, the BER and the average channel capacity, was proven via simulations, and the theoretical analysis matches the corresponding numerical results well. It was also shown in the numerical results that some other parameters, including the number of relays, spectrum sensing process, and the spectrum utilization efficiency of primary users impact the system performance greatly in the presence of multiple potential relays. Simulation results proved that the full spatial diversity order can be achieved at low SNR by partial relay selection.
References
- Z. Zhang, K. Long, and J. Wang, "Self-Organization Paradigms and Optimization Approaches for Cognitive Radio Technologies: A Survey," IEEE Wireless Commun., vol. 20, no. 2, pp. 36-42, Apr. 2013.
- I. F. Akyildiz, W. Y. Lee, M. C. Vuran, and S. Mohanty, "Next generation/dynamic spectrum access/cognitive radio wireless networks: A survey," Computer Networks, vol. 50, no. 13, pp. 2127-2159, May 2006. https://doi.org/10.1016/j.comnet.2006.05.001
- Z. Zhang, K. Long, J. Wang, and F. Dressler, "On swarm intelligence inspired self-organized networking: its bionic mechanisms, designing principles and optimization approaches," IEEE Commun. Surv. Tut, vol. 16, no. 1, pp.513-537, 2014. https://doi.org/10.1109/SURV.2013.062613.00014
- A. A. Alkheir and M. Ibnkahla, "Performance analysis of cognitive radio relay networks using decode and forward selection relaying over Rayleigh fading channels," in Proc. of IEEE Global Telecommn. Conf. (GLOBECOM), pp. 1-5, 2011.
- Y.-C. Liang, K.-C. Chen, G. Y. Li, and P. Manhonen, "Cognitive radio networking and communications: an overview," IEEE Trans. Veh. Technol., vol. 60, no. 7, pp. 3386-3407, Sept. 2011. https://doi.org/10.1109/TVT.2011.2158673
- Z. Zhang, W. Zhang, and C. Tellambura, "OFDMA uplink frequency offset estimation via cooperative relaying," IEEE Trans. Wireless Commun., vol. 8, no. 9, pp. 4450-4456, Sept. 2009. https://doi.org/10.1109/TWC.2009.081525
- R. U. Nabar, H. Bolcskei, and F. W. Kneubuhler, "Fading relay channels: performance limits and space-time signal design," IEEE J. Select. Areas Commun., vol. 22, no. 6, pp. 1099-1109, Aug. 2004. https://doi.org/10.1109/JSAC.2004.830922
- Z. Zhang, W. Huangfu, K. Long, X. Zhang, X. Zhang, and B. Zhong, "On the designing principles and optimization approaches of bio-inspired self-organized network: a survey," Science China Information Sciences, vol. 56, no. 7, pp. 1-28, Jul. 2013.
- J. Jia, J. Zhang, and Q. Zhang, "Cooperative relay for cognitive radio network," IEEE INFOCOM, 2009, pp. 2304-2312.
- Z. Zhang, W. Zhang, and C. Tellambura, "Cooperative OFDM channel estimation in the presence of frequency offsets," IEEE Trans. Veh. Technol., vol. 58, no. 7, pp. 3447-3459, Sept. 2009. https://doi.org/10.1109/TVT.2009.2016345
- D. Gunduz, A. Yener, A. Goldsmith, and H. V. Poor, "The multi-way relay channel," in Proc. of IEEE Int. Symp. on Infor.Theory (ISIT), Seoul, Korea, pp. 339-343, 2009.
- S. S. Ikki and M. H. Ahmed, "Performance analysis of adaptive decode-and-forward cooperative diversity networks with best-relay selection," IEEE Trans. Commun., vol. 58, no. 1, pp. 68-72, Jan. 2010. https://doi.org/10.1109/TCOMM.2010.01.080080
- Z. Zhang, C. Tellambura, and R. Schober, "Improved OFDMA uplink transmission via cooperative relaying in the presence of frequency offsets-part I: ergodic information rate analysis," Euro. Trans. Telecommun., vol. 21, no. 3, pp. 224-240, Apr. 2010.
- A. Muller and J. Speidel, "Relay selection in dual-hop transmission systems: Selection strategies and performance results," in Proc. of IEEE Int. Conf. Communications (ICC), 2008, pp. 4998-5003.
- I. Ahmed, A. Nasri, D. S. Michalopoulos, R. Schober, and R. K. Mallik, "Relay subset selection and fair power allocation for best and partial relay selection in generic noise and interference," IEEE Trans. Wireless Commun., vol. 11, no. 5, pp. 1828-1839, May 2012. https://doi.org/10.1109/TWC.2012.031212.111113
- Y. Li, Z. Zhang, X. Zhang X, B. ZHONG, and K. Long, "Best relay selection in decode and forward cooperative cognitive radio relay networks over rayleigh fading channels," in Proc. of IET Inter. Conf. on Inform. & Commun.Tech. (IETICT), Apr. 2013.
- Z. Zhang, C. Tellambura, and R. Schober, "Improved OFDMA uplink transmission via cooperative relaying in the presence of frequency offsets-part II: Outage information rate analysis," Euro. Trans. Telecommun., vol. 21, no. 3, pp. 241-250, Apr. 2010.
- M. Torabi, D. Haccoun, and J.-F. Frigon, "Impact of outdated relay selection on the capacity of AF opportunistic relaying systems with adaptive transmission over non-identically distributed links," IEEE Trans. Wireless Commun., vol. 10, no. 11, pp. 3626-3631, Nov. 2011. https://doi.org/10.1109/TWC.2011.092711.110136
- H. Chen, J. Liu, Z. Dong, Y. Zhou, and W. Guo, "Exact capacity analysis of partial relay selection under outdated CSI over rayleigh fading channels," IEEE Trans. Veh. Technol., vol. 60, no. 8, pp. 4014-4018, Oct. 2011. https://doi.org/10.1109/TVT.2011.2167166
- P. L. Yeoh, M. Elkashlan, Z. Chen, and I. B. Collings, "SER of multiple amplify-and-forward relays with selection diversity," IEEE Trans. Commun., vol. 59, no. 8, pp. 2078-2083, Aug. 2011. https://doi.org/10.1109/TCOMM.2011.053111.090279
- M. Torabi and D. Haccoun, "Capacity of amplify-and-forward selective relaying with adaptive transmission under outdated channel information," IEEE Trans. Veh. Technol., vol. 60, no. 5, pp. 2416-2422, June 2011. https://doi.org/10.1109/TVT.2011.2139232
- B. Zhong, Z. Zhang, X. Zhang, Y. Li, and K. Long, "Impact of partial relay selection on the capacity of communications systems with outdated CSI and adaptive transmission techniques," EURASIP J. Wirel. Comm., pp. 1-13, Feb. 2013.
- J. Lee, H. Wang, J. G. Andrews, D. Hong, "Outage probability of cognitive relay networks with interference constraints," IEEE Trans. Wireless Commun., vol. 10, no. 2, pp. 390-395, Feb. 2011. https://doi.org/10.1109/TWC.2010.120310.090852
- Y. Zou, J. Zhu, B. Zheng, and Y. Yao, "An adaptive cooperation diversity scheme with best-relay selection in cognitive radio networks," IEEE Trans. Signal Processing, vol. 58, no. 10, pp. 5438-5445, Oct. 2010. https://doi.org/10.1109/TSP.2010.2053708
- Y. Zou, Y. Yao, and B. Zheng, "Cognitive transmissions with multiple relays in cognitive radio networks," IEEE Trans. Wireless Commun., vol. 10, no. 2, pp. 648-659, Feb. 2011. https://doi.org/10.1109/TWC.2010.120610.100830
- Y. Zou, Y. Yao, and B. Zheng, "Cooperative relay techniques for cognitive radio systems: Spectrum sensing and secondary user transmissions," IEEE Commun. Magazine, vol. 50, no. 4, pp. 98-103, Apr. 2012.
- S. I. Hussian, M.-S. Alouini, M. Hasna, and K. Qarage, "Partial relay selection in underlay cognitive networks with fixed gain relays," in Proc. of IEEE Vehicular Technology Conf. (VTC), May 2012.
- Q. Zhang, J. Jia, and J. Zhang, "Cooperative relay to improve diversity in cognitive radio networks," IEEE Commun. Mag., vol. 47, no. 2, pp. 111-117, Feb. 2009.
- D. Li, "Outage probability of cognitive radio networks with relay selection," IET Commun., vol. 5, no. 18, pp. 2730-2735, Dec. 2011. https://doi.org/10.1049/iet-com.2011.0216
- K. B. Fredj and S. Aissa, "Performance of amplify-and-forward systems with partial relay selection under spectrum-sharing constraints," IEEE Trans. Wireless Commun., vol. 11, no. 2, pp. 500-504, Feb. 2012. https://doi.org/10.1109/TWC.2011.120911.102240
- R. Tandra, S.M.Mishra, and A. Sahai, "What is a spectrum hole and what does it take to recognize one?" in Proc. of IEEE, vol. 97, no. 5, pp. 824-848, May 2009. https://doi.org/10.1109/JPROC.2009.2015710
- Q. Wu, G. Ding, J. Wang, and Y. Yao, "Spatial-temporal opportunity detection for spectrum-heterogeneous cognitive radio networks: Two-dimensional sensing," IEEE Trans. Wireless Commun., vol. 12, no. 2, pp. 516-526, Feb. 2013. https://doi.org/10.1109/TWC.2012.122212.111638
- G. Ding, J.Wang, Q.Wu, F. Song, and Y. Chen, "Spectrum sensing in opportunity-heterogeneous cognitive sensor networks: How to cooperate?," IEEE Sensors Journal, vol. 13, no. 11, pp. 4247-4255, Nov. 2013. https://doi.org/10.1109/JSEN.2013.2264161
- J. G. Proakis and M. Salehi, Digital Communications. New York: McGraw-Hill, 2008.
- Y. Zhao, R. Adve, and T. J. Lim, "Symbol error rate of selection amplify-and-forward relay systems," IEEE Commun. Lett., vol. 10, no. 11, pp. 757-759, 2006. https://doi.org/10.1109/LCOMM.2006.060774
- I. S. Gradshteyn and I. M. Ryzhik, Table of Integrals, Series, and Products, 7th ed. Academic Press, Inc., 2007.
- D. S. Michalopoulos, H. A. Suraweera, G. K. Karagiannidis, and R. Schober, "Amplify-and-forward relay selection with outdated channel estimates," IEEE Trans. Commun., vol. 60, no. 5, pp. 1278-1290, May 2012. https://doi.org/10.1109/TCOMM.2012.032012.110430