The Proceedings of Second International Conference on Electrical and Electronics Engineering, Clean Energy and Green Computing, Konya, Turkey, 2015
Proposed Optimization Technique for Maximization of Throughput under Using Different Multicarrier Systems in Cognitive Radio Networks Ahmed. A. Rosas*, Mona Shokair and Sami. A. El_dolil Faculty of Electronic Engineering, El-Menoufia University, Menouf, 32952, Egypt Email:
[email protected], mona.sabry@ el-eng.menofia.edu.eg and
[email protected]
ABSTRACT Beside the hard ware implementation problems in Cognitive Radio Networks (CRNs), some communication criteria’s are considering to generate CRN with improvement, reliability and motivation communication system. One of these criteria’s is how to maximize total throughput of the spectrum sharing CRNs to improve the Quality of Services (QOS) of Secondary Users (SUs), and adapt its communication parameters to communicate concurrent with primary system while assuring the QOS of primary users (PUs). In this paper, proposed optimization technique studies the power allocation strategies to maximize the total throughput of underlay CRN. This optimizing technique is based on Genetic Algorithm (GA) under the consideration of the maximum transmit power and interference constraints. Furthermore, this proposed technique will be applied on two different multicarrier modulation systems which are Orthogonal Frequency Division Multiplexing (OFDM) and Generalized Frequency Division Multiplexing (GFDM). Moreover, some OFDM problem will be solved such as cyclic prefix and subcarrier overlapping in CRNs. Finally, the performance of the proposed technique for maximization of throughput will be investigated by using different number of PUs, SUs, subcarriers M, noise levels and fully utilized system under maximum power and interference constraints while ensuring the QOS of PUs.
KEYWORDS Cognitive radio network, spectrum sharing, throughput, genetic algorithm, OFDM system, GFDM system
1 INTRODUCTION Recent years, Wireless communication has become an essential tool for the modern society. Over and above these years, it has witnessed a rabidly increased in the demand
ISBN: 978-1-941968-12-3 ©2015 SDIWC
on new wireless communication systems, applications and wireless services. In accordance with that the traditional approaches to spectrum, managements become more crowded by licensed systems, in the sense that more frequency bands are needed to be acclimated with the rapidly communication demands requirements. Based on the report of the Federal Communication Commission (FCC) [1] indicated that the radio spectrum is allocated by the licensed (primary) systems and these systems have utilization to their spectrum bands from 15 % reached to 85% under maximum usage. From this perspective, cognitive radio [2-4] is a new hopeful and promising technique. It is suggested to improve the spectrum utilization and efficiency. Cognitive radio based on very intelligent devices, network entities and software radio defined which have the ability be programmed and configured dynamically and can automatically detect channels behaviors in a wireless spectrum and change both transmission and reception parameters enabling more communications to run concurrently and also improve radio operating behavior. In CRNs, there are two main modes of operation. i) Opportunistic spectrum access (overlay) [5], which SUs in the cognitive radio system are allowed to access on the licensed spectrum of the primary system only if PUs are not allocated to this band (spectrum hole). ii) Spectrum sharing (underlay) [6, 7], which SUs are allowed to access on the licensed spectrum and share this band with PUs to achieve QOS requirements of PU. It will be done by making interference on the primary users
25
The Proceedings of Second International Conference on Electrical and Electronics Engineering, Clean Energy and Green Computing, Konya, Turkey, 2015
below a specified threshold level [1]. Actually, the main target of the most CRNs is how to maximize the total throughput of the system without effecting on the licensed systems under some constraints. Many searches study this target by selecting best modulation systems and intelligent power allocating techniques using optimization algorithms [8-9].Where these algorithms are used to select the optimal power values to maximize throughput under power and interference constraints to improve QOS of the SUs without effecting on QOS of PUs. Many researches supposed OFDM [10] as a one of the promising modulation systems with different power allocation strategies. In [11], different resource allocation algorithms are developed for the downlink of multiuser OFDM wireless communication systems and compared with the other classic techniques based on complexity. Different optimal and suboptimal power allocation schemes for OFDM-based on CRNs were investigated in [12]. Lagrangian optimization technique was proposed in [13]. Using Lagrangian and water filling algorithms to maximize the system capacity based on multiuser OFDM in the downlink CRN under interference constraints was discussed in [14]. In [15], Fair multiuser resource allocation algorithm based on greedy approaches for OFDM in CRN was proposed. Moreover, the minimization of cost function and complexity of these algorithms was made. Another efficient power allocation techniques based on barrier method added with subchannel allocation was suggested in [16, 17]. Actually, there are many problems with using OFDM system in CRNs like cyclic prefix (CP) over head [18], peak to average power ratio PAPR [19-22], out of band radiation (OOB) [23], synchronization and symbol error rate (SER). Many searches studied how to solve these problems but at the expense of the total system throughput. Therefore, a proposed technique was used
ISBN: 978-1-941968-12-3 ©2015 SDIWC
GFDM system [24] based on a filter bank approaches [25] to solve most of OFDM problems with improvement in the total throughput in CRN. Further, many optimization techniques were used a differential evaluation algorithms based on convex optimization [16] problems for a power allocation strategies in CRN. In this paper, power allocation strategies to maximize the total throughput of underlay CRN based on proposed technique under two constraints will be suggested. The first constraint is the maximum transmit power at SU access point and the other is the interference on the PU must be below the interference threshold level to save the QOS of the PU. Furthermore, this proposed technique will be studied under using two different multicarrier modulation systems, OFDM and GFDM. Then the improvement in total throughput under different modulating techniques will be investigated. Beside them, some OFDM problems will be solved in CRNs such as cyclic prefix. Moreover, the performance of the system with different number of PUs, SUs, subcarriers, and fully utilized system under maximum transmit power and interference constraints will be investigated. The rest of the paper is organized as the follows: Section 2, describes the proposed system model based on two modulation techniques. In Section 3, Analysis of the total throughput will be made. In Section 4, Optimization technique will be presented. Simulation results will be made in Section 5. Finally, Conclusions will be done in Section 6. 2 SYSTEM MODEL GFDM [26-27] is a flexible multicarrier modulation with pulse shaping based on filter bank approaches, which data spread across two dimensional block structures over time and frequency, but OFDM is a flexible in orthogonal multicarrier
26
The Proceedings of Second International Conference on Electrical and Electronics Engineering, Clean Energy and Green Computing, Konya, Turkey, 2015
modulation which data spread across one dimensional block structure(frequency) as illustrated in figure 1. In the OFDM, the total bandwidth is divided into narrowband subchannels, which each subchannel with a bandwidth much smaller than the coherence bandwidth of the channel. Where the high rate data stream is split into substreams of lower rate which are modulated into N OFDM symbols and transmitted simultaneously over N orthogonal subcarriers.
data transmitted on the m th subcarrier and v th subsymbol.
Figure 2. System model.
Figure1. OFDM and GFDM data structure.
Considering the downlink of multiuser GFDM or OFDM system coexists with the primary system, Both L PUS are served by Base Station (BS) in the licensed system and K SUS are served by Access Point (AP) in CRN arranged randomly in circuit with radius about 1Km and shared the same licensed spectrum band W as illustrated in figure 2. The total bandwidth is divided into M subchannels with band WS. Perfect channel state information will be assumed. In GFDM [28], all transmitted symbol can be expressed by, X [n] =∑
∑
(1)
Where binary data is encoded and QAM with modulation order j maps it into d data vector that contains MV element. dm,v is the
ISBN: 978-1-941968-12-3 ©2015 SDIWC
gm,v is pulse shape response with n sampling index. There are different types of pulse shape filter [29] using in transmitting side such as raised cosine, root raised cosine, Xia pulse, Gaussian pulse and dirichlet pulse. The choice of pulse shape filter has hardly affected on reduction of OOB, the spectral properties and symbol error rate. In the receiving side, the use of ZF receiver (for OOB reduction) or MF receiver (for interference cancelation) was made. Pulse shape filter determined the rolloff factor σ which determines the overlap of subcarriers in the frequency domain and takes the range from (0, 1). The pulse shaping is used at transmitter to reduce the OOB radiation on the cost of noise at receiving side. Where, the noise factor (NF) increased with the roll-off factor due to higher overlapping of subcarriers. 3 ANALYSIS OF THE THROUGHPUT Sum rate for all SUs over all subcarriers in GFDM can be achieved by the following statements [30] and select the parameter from Table 1. For OFDM modulation, will assume 1. The achievable rate over subchannels m employed by the SU i is given by,
27
The Proceedings of Second International Conference on Electrical and Electronics Engineering, Clean Energy and Green Computing, Konya, Turkey, 2015
Table 1. GFDM and OFDM parameters.
[
Ci,m=
], )
(∑
(2)
The achievable rate over all subchannels M employed by the SU i is denoted by, ∑
Ci =
[ ], )
(∑
(3)
The sum rate of all SUs can be formulated as follows, ∑
∑
[ ], )
(∑
(4)
s.t. ∑ Pi,m ∑
Parameter Number of PUS Number of SUS Mapping Number of subcarriers Number of subsymbols Number of cyclic prefix Number of cyclic suffix Subcarrier distance System band width Roll off factor
Variable
OFDM
GFDM
L
2, 4, 6.. 32
2, 4, 6... 32
K
8, 16 .. 64
8, 16.. 64
M
16 QAM 32
16 QAM 32
V
1… 9
9
NCP
V*8
8
NCS
0
0
WS
0.3 MH
0.3 MH
WT
10 MH
10MH
σ
--
0.1 …. 0.9
4 OPTIMIZATION TECHNIQUES ∑
, 0,
, Gi,p
,
(5) ,
(6) (7)
Where Gi,i is the channel power gain for data channel to SU i, Pi,m optimum power transmit for SU i over subchannel m and Г is a factor related to bit error rate over QAM mapper. Gp,i is the channel power gain for PU transmitter to SU i receiver and Gi,k is the channel power gain for SU i transmitter to SU k receiver. PP is the PU transmit power and Gi,p is the channel power gain for SU i transmitter to PU receiver. ρi represents the activity of the PU and δm is the subchannel activity. φm is the interference threshold at PU and N0 is the noise power of additive white Gaussian with zero mean and variance σ2 . Noise PS is the maximum power allocated by access point.
Genetic algorithm (GA) [26] is the proposed algorithm aimed to maximize the sum data rate (total throughput) for all SUs in down link CRNs. This occurred by selecting the optimum power level for each subcarrier to all SUs under both power and interference constraints. The operation of GA is shown in figure 3.
Figure 3. GA operation.
ISBN: 978-1-941968-12-3 ©2015 SDIWC
28
The Proceedings of Second International Conference on Electrical and Electronics Engineering, Clean Energy and Green Computing, Konya, Turkey, 2015
Depending on random solutions called chromosomes of all optimization variables are calculated. These values in the acceptable pounders are required to the optimization problem and this process called a population. Then the algorithm starts to calculate the objective function based on the population. By selection (parents) mutation and cross over trying to improvement the objective function by generate a new population from old one is calculated and repeating this step until reached to the stopping criteria. At stopping criteria, the optimum values for all variables will be generated based on some optimization parameters. Therefore, the genetic algorithm parameters are used in simulation preferred the population size about 20 for simplicity and process time. The maximum number of generation is equal to 100. Constraints depending on mutation, uniform selection and cross over functions are used with cross over fraction reached to about 0.8 and stopping criteria is the Function tolerance with value reached to 10-6.
Figure 4. Investigates the total throughput for all SUs in the downlink CRN with the total transmit power at AP using two different modulation techniques GFDM and OFDM. By increasing the total transmit power at AP, the throughput will be increased until it reached to saturation (maximum and constant level). Also this Fig. shows that the improvement in the total rate for all SUs using GFDM based on GA reached to 19.6% over OFDM under the same optimization technique at saturation. Under the effect of cyclic prefix at 5 dB, the total throughput reduced by 3.5% in GFDM but in OFDM, the reduction reached to 19%. This means that the improvement in GFDM using minimum number of cyclic prefix reached to approximately 15% rather than OFDM.
5 SIMULATION RESULTS
Figure 5. Sum rate for all SUs in CRN with maximum transmit power at AP for GFDM under different level of σ and £ .
Figure 4. Sum rate for all SU in CRN with maximum transmit power for GFDM and OFDM.
ISBN: 978-1-941968-12-3 ©2015 SDIWC
Figure 5. Studies the effect of roll off and noise factors on the total throughput in down link CRNs based on GFDM system. From this Fig., we conclude that by increasing the roll off factor, the overlaps between
29
The Proceedings of Second International Conference on Electrical and Electronics Engineering, Clean Energy and Green Computing, Konya, Turkey, 2015
subcarrier will be increased. Therefore, the noise level will be increased, and consequently the total throughput will be reduced. At transmit power equal to 5 dB, the roll off factor increases from 0.1 to 0.9 with incensement in noise factor from 0 dB to about 3 dB. Therefore throughput reduced from 2% to 17.9%.
Figure 7. Sum rate for all SUs in CRNs with maximum transmit power under different number of PUs using GFDM system.
Figure 6. Sum rate for all SUs in CRN with maximum transmit power under different number of PUs using OFDM system.
Figures 6 and 7 show the performance of the total throughput under different number of PUs using two different modulation techniques. The total throughput will be improved in the GFDM technique than using OFDM. Furthermore, by increasing the number of active PUs in the system, the total throughput will be reduced. In fact, the system reached to the fully utilized which means that all PUs becomes active and the total throughput reduced by about 58%.
ISBN: 978-1-941968-12-3 ©2015 SDIWC
Figure 8. Sum rate for all SUs in CRNs with maximum transmit power under different number of SUs using OFDM system.
30
The Proceedings of Second International Conference on Electrical and Electronics Engineering, Clean Energy and Green Computing, Konya, Turkey, 2015
Figure 9. Sum rate for all SUs in CRN with maximum transmit power under different number of SUs using GFDM system.
Figures 8 , 9. Show the performance under increasing the number of SUs. If increasing the number of SUs, the total throughput improves, but at the saturation level due to increasing the number of SUs, the saturation level will be reduced because the interference from SUs on each other on each subcarrier will be increased based on its number. Figure 10 and 11. Show that by increasing the total number of subcarriers for each SU, the total throughput will be improved. If the number of subcarrier increases to double, this improve the total throughput by about 58% in OFDM system and by about 67% in GFDM system. Figure 12. Shows that the total throughput under different noise levels which at low noise level, the total throughput will be high and by increasing the noise level, the total throughput will be reduced until the noise level will be high and the total throughput will be destroyed. Moreover, GFDM system improves the total throughput by approximately 20% over OFDM system.
ISBN: 978-1-941968-12-3 ©2015 SDIWC
Figure 10. Sum rate for all SUs in CRN with maximum transmit power under different number of subcarriers in OFDM system.
Figure 11. Sum rate for all SUs in CRN with maximum transmit power under different number of subcarriers in GFDM system.
31
The Proceedings of Second International Conference on Electrical and Electronics Engineering, Clean Energy and Green Computing, Konya, Turkey, 2015
[3] H. Sakran, M Shokair, O. Nasr, E.S. El-Rabaie and A.A. El-Azm, " Proposed relay selection scheme for physical layer security in cognitive radio networks," IET Communications 2012. [4] H. Venkataraman and G. Miro, “Cognitive Radio and Its Application for Next Generation Cellular and Wireless Networks”, Muntean.2012. [5] S. Huang, X. Liu, and Z. Ding, “Opportunistic spectrum access in cognitive radio networks,” in Proc. IEEE INFOCOM’08, pp. 1427–1435, 2008. [6] I. F. Akyildiz, W. Y. Lee, M. C. Vuran and S. Mohanty, “A Survey on Spectrum Management in Cognitive Radio Networks”, IEEE Communications Society, vol. 46, no. 6, pp.40 – 48, 2008.
Figure 12. Sum rate for all SUs in CRN with noise level.
6 CONCLUSIONS Proposed technique was suggested for Power allocation strategies to maximize the total throughput of underlay CRN. This proposed technique based on genetic algorithm under both interference and power constraints which is not clarified until now. Furthermore, this proposed technique based on two different multicarrier modulation systems OFDM and GFDM. Beside them, some OFDM problems in CRNs were solved. Moreover, the performance of the system under using different number of PUs, SUs, subcarriers, and fully utilized system under maximum transmit power and interference constraints was studied. REFERENCES [1] Federal Communications Commission, “Spectrum policy task force report, (ET Docket No. 02-135),” Tech. Rep., Nov. 2002. [2] S. Haykin, “Cognitive radio: brain-empowered wireless communications,” IEEE Journal of Selected Areas on Communications, vol. 23, no. 2, pp. 201220, Feb.2005.
ISBN: 978-1-941968-12-3 ©2015 SDIWC
[7] Li, Yang, and A. Nosratinia, “Spectrum Sharing with Distributed Relay Selection and Clustering”, IEEE Transactions 0n, communications, vol. 16, pp.53 – 62, 2013. [8] G. C. Onwubolu and B. V.Babu, “New Optimization Techniques in Engineering” SpringerVerlag Berlin Heidelberg, 2004. [9] A. M. Benaya, M. Shokair, E.S. El-Rabaie and M. F. Elkordy,” Optimal Power Allocation for SensingBased Spectrum Sharing in MIMO Cognitive Radio Network,” Wireless Personal Communications, 2015. [10] H. Rohling, “OFDM Concepts for Future Communication Systems” Springer-Verlag Berlin Heidelberg, pp.254, 2011. [11] S. Sadr, A. Anpalagan, K. Raahemifar, “Radio Resource Allocation Algorithms for the Downlink of Multiuser OFDM Communication Systems” IEEE Communications Surveys & TUTORIALS, VOL. 11, NO. 3, 2009. [12] S. Boyd and L. Vandenberghe, “Convex Optimization”. Cambridge University Press, 2004. [13] M.M. Saleh, “Adaptive Resource Allocation in Multiuser OFDM Systems” Final Report, Spring 2005. [14] T. Qin, C. Leung, “Fair Adaptive Resource Allocation for Multiuser OFDM Cognitive Radio Systems” IEEE, Communications and Networking, pp.115 – 119, 2007. [15] G. Bansal, Md. J. Hossain and V. K. Bhargava, “Optimal and Suboptimal Power Allocation Schemes for OFDM-based Cognitive Radio Systems” IEEE transactions on wireless communications, vol. 7, no. 11, 2008.
32
The Proceedings of Second International Conference on Electrical and Electronics Engineering, Clean Energy and Green Computing, Konya, Turkey, 2015
[16] S. Wang, M. Ge and C. Wang, “Efficient Resource Allocation for Cognitive Radio Networks with Cooperative Relays” IEEE Journal on Selected Areas in Communications vol. 31, no. 11, 2013. [17] P. Zhang, L. Yang and X. Liu, “Subcarrier and Power Allocation in OFDM-based Cognitive Radio Systems” MECS, Computer Network and Information Security, vol. 1, pp.24 – 30, 2010. [18] H. Kim, J. Kim, S. Yang, M. Hong, and Y. Shin, “An Effective MIMO OFDM System for IEEE 802.22 WRAN Channels,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 55, no. 8, pp. 821–825, Aug. 2008. [19] M. Shokair, “On Proposing Scheme for Reduction of PAPR in OFDM System under Equalization," Theoretical and Applied Information Technology Journal, pp. 91-97, 2010. [20] H. Sakran, M. Shokair, and A.A. El-Azm," A New Peak-to-Average Power Reduction Technique in the OFDM System Using μ-Law Compander," IEEE ISWCS 2008.
Systems with Generalized Frequency Division Multiplexing,” in Proceedings International Symposium on Intelligent Signal Processing and Communications Systems (ISPACS’13), Naha, Okinawa, Japan, Nov. 2013, pp. 651–655. [28] N. Michailow, I. Gaspar, S. Krone, M. Lentmaier, and G. Fettweis, “Generalized Frequency Division Multiplexing: Analysis of an Alternative Multi-Carrier Technique for Next Generation Cellular Systems,” in Proceedings 9th International Symposium on Wireless Communication Systems (ISWCS’12), Paris, France, Aug. 2012, pp. 171–175. [29] X. Xia, “A Family of Pulse-Shaping Filters with ISI-Free Matched and Unmatched Filter Properties,” IEEE Transactions on Communications, vol. 45, no. 10, pp. 1157–1158, 1997. [30] X. Yin and X. Cheng, “Propagation Channel Characterization, Parameter Estimation, and Modeling for Wireless Communications” 1 edition, Wiley-IEEE Press, pp. 350, 2015.
[21] H. Sakran, M. Shokair, and A.A. El-Azm," Combined Interleaving And Companding For PAPR Reduction In OFDM Systems," Progress In Electromagnetics Research C, Vol. 6, 67–78, 2009. [22] H. Sakran, M. Shokair, and A.A. El-Azm," An Efficient Technique For Reducing PAPR Of OFDM System In The Presence Of Nonlinear High Power Amplifier," Progress In Electromagnetics Research C, Vol. 2, 233–241, 2008. [23] J. Van, D. Beek and F. Berggren, “Out-of-Band Power Suppression in OFDM,” IEEE Communications Letters, vol. 12, no. 9, pp. 609–611, 2008. [24] G. Fettweis, M. Krondorf and S. Bittner, “GFDM – Generalized Frequency Division Multiplexing,” in Proceedings 69th IEEE Vehicular Technology Conference (VTC Spring’09), Barcelona, Spain, pp. 1–4, 2009. [25] B. Farhang-Boroujeny, “OFDM Versus Filter Bank Multicarrier,” Signal Processing Magazine, IEEE, vol. 28, no. 3, pp. 92–112, May 2011. [26] S. N. Sivanandam and S. N. Deepa, “Introduction to Genetic Algorithms”, SpringerVerlag Berlin Heidelberg, pp. 442, 2008 [27] N. Michailow and G. Fettweis, “Low Peak-toAverage Power Ratio for Next Generation Cellular
ISBN: 978-1-941968-12-3 ©2015 SDIWC
33
