Please use this identifier to cite or link to this item: http://hdl.handle.net/11527/13202
Title: Altına Sermeli Bilişsel Radyo Ağları İçin İşbirlikli Spektrum Paylaşım Protokolleri
Other Titles: Cooperative Spectrum Sharing Protocols For Underlay Cognitive Radio Networks
Authors: Aygölü, Hasan Ümit
Saraç, Sema
10129278
Elektronik ve Haberleşme Mühendisligi
Electronic and Communication Engineering
Keywords: Bilişsel Radyo Ağları
İşbirlikli Haberleşme
Otomatik Yineleme İsteme
Cognitive Radio
Cooperative Communicaiton
Automatic Repeat Request
Issue Date: 4-Nov-2016
Publisher: Fen Bilimleri Enstitüsü
Institute of Science and Technology
Abstract: Spektrum paylaşımlı bilişsel radyo ağlarında ikincil kullanıcıların (SU), birincil kullanıcının (PU) başarımını en az etkileyecek şekilde spektrumu kullanması önemlidir. Ancak bunu yaparken kendi başarımını da belli bir düzeyde tutması gerekmektedir. Bu tezde, otomatik yineleme isteme (ARQ) tabanlı PU'dan ve çift yönlü iletişim kurabilen SU'dan oluşan altına sermeli bilişsel radyo ağı için yeni iki farklı işbirlikli spektrum paylaşım protokolü önerilmiştir. Önerilen protokollerde SU bazı durumlarda kendi paketini iletmek için spektrumu kullanırken bazı durumlarda ise yalnızca PU'nun paketini iletmektedir. Böylece spektrumun paylaşımlı olarak kullanıldığı durumlarda kendi başarımını iyileştirirken, diğer durumlarda işbirliği yaparak spektrum paylaşımlı durumda PU'nun kötü etkilenen başarımını arttırmayı amaçlamaktadır. Bu tezde ilk olarak, telsiz iletişim ağları ve işbirlikli çeşitleme protokolleri ile ARQ protokolü anlatılmış ve bilişsel radyo ağ kavramı ve çeşitlerinden bahsedilmiştir. Sonrasında ele alınan ARQ tabanlı PU'dan ve iki yönlü iletişim kurabilen SU'dan oluşan altına sermeli bilişsel ağ modeli tanıtılmıştır. Sırasıyla çöz-ve-aktar (DF) tekniği ile kuvvetlendir-ve-aktar (AF) tekniği için önerilen spektrum paylaşım protokolleri açıklanmış ve her iki protokol için referans modeller oluşturulmuştur. PU'nun paketinin iletilebilmesi için farklı yineleme sayılarında bu protokoller incelenmiştir. Ele alınan protokollerde PU ve SU'nun servis kesilme olasılıkları kullanılarak verim analizleri yapılmış ve bilgisayar benzetimleriyle başarım değerlendirmeleri desteklenmiştir. Her iki teknik için de önerilen protokollerde yineleme sayısındaki artış PU'nun düşük SNR'lerde verimini az da olsa kötü etkilemesine rağmen yüksek SNR'lerde bir değişime neden olmamıştır. DF tekniği için önerilen protokolde PU'nun verimi AF için önerilen protokole göre düşük SNR değerlerinde daha iyi iken; SU'nun verimi sadece çok düşük SNR değerlerinde farklılık göstermiş; kalan değerlerde başarımının değişmediği gözlenmiştir. Ayrıca her iki protokolde de yüksek SNR değerlerinde PU'nun veriminden göz ardı edilebilecek bir kayıpla, SU'nun verimi aynı SNR değerlerinde belirli bir düzeyde sabit tutulmuştur. Her iki protokolün de birinde PU'nun başarımına öncelik veren diğerinde ise SU'nun başarımına öncelik veren farklı iki çalışmayla başarım karşılaştırılması yapılmıştır. Birincil kullanıcının başarımına öncelik veren çalışmaya göre önerilen protokollerde PU'nun başarımı düşük SNR'lerde daha iyi iken, yüksek SNR'lere gidildikçe biraz azalmıştır. Ancak yüksek SNR'lerde SU başarımı bu çalışmada sıfıra düşerken, önerilen protokollerin her ikisi için de belli bir değerde sabit kalmıştır. Önerilen protokollerde SU başarımı SU'nun başarımına öncelik veren çalışmaya göre yüksek SNR'lerde yaklaşık %50 kötüleşirken; aynı SNR değerlerinde PU başarımı bu çalışmaya kıyasla %70 daha iyileşmiştir.
Of late years, usage of wireless communication systems increases in military and public area. The increase in number of users and inefficient usage of the spectrum lead to the problem of spectrum scarcity. One of the most promising technologies to overcome this problem is the cognitive radio to sense and access the spectrum dynamically and to use the available spectrum efficiently. Cognitive radios are intelligent devices able to be aware of their surrounding environment to control and adjust their parameters according to their internal conditions. Cognitive radios optimize the communication in a sophisticated way for users by providing spectrum flexibility and spectrum management. The basis of cognitive radio networks consist of one primary and one secondary network. Primary user (PU) which belongs to primary network has license to access the spectrum band, while secondary user (SU) which belongs to secondary network has not. In earlier times, researches have been based on spectrum sensing by SU, whereas increasing number of researches have been carried out on spectrum sharing between PU and SU recently. In the literature, three fundamental paradigms for cognitive radio networks are explored: interweave, overlay and underlay. According to studies, the major part of the spectrum has spectrum holes in a certain time because of the inefficient usage. In interweave cognitive radio networks, SU opportunistically detects the spectrum holes in where PU does not operate and uses theses holes to access the spectrum without interfering at PU's receiver. When PU accesses that spectrum band, SU detects another spectrum hole and settle in after it concludes its communication. However, PU and SU are able to access spectrum at the same time in overlay cognitive radio networks. SU has knowledge of PU's information to mitigate or cancel interference at the SU's receiver; SU cooperates with PU and improves PU's performance by transmitting PU's packet to receivers with its own packet simultaneously. The underlay cognitive radio networks allow PU and SU to reach spectrum and transmit their packet concurrently. In contrast to overlay cognitive radio networks, SU can access the spectrum simultaneously with PU as long as the interference that it causes at PU's receiver is under an acceptable level. SU needs to know interference it causes at the PU's receiver. It is important that the SU's spectrum usage has minimum destructive effect on PU's transmission while SU has to keep its performance at a certain level. In this dissertation, first, basics on wireless communication networks and cooperative diversity techniques with automatic repeat request (ARQ) are reviewed, and the concept of cognitive radio network with its different paradigms are surveyed. Second, an underlay cognitive radio network model which consists of an ARQ based PU and a bidirectionally communicating SU is probed. Proposed spectrum sharing protocols for two different cooperative diversity strategies i.e., amplify-and-forward (AF), decode-and-forward (DF) are described. Reference schemes as cooperation-only scheme, underlay-only scheme and non-CR scheme are also presented to make performance comparisons for PU in both proposed protocols. In cooperation-only scheme, SU only uses spectrum to cooperate with PU without transmitting its own information to improve the PU's transmission. On the contrary, PU and SU use the spectrum concurrently where SU's transmission considers a certain interference threshold in underlay-only scheme to not degrade the PU's performance. SU stays silent during PU's transmission or vice versa. On the other hand, non-CR scheme is composed of only the PU network which indicates that no SU's activity occurs in the considered spectrum. PU transmits its own packet in the absence of SU nodes in surrounding environment. Finally, PU's and SU's throughputs are derived from outage probabilities and the throughput results are supported via computer simulation results for both proposed protocols. The performance comparisons between the proposed protocols and the existing protocols in the literature are also provided. In the considered underlay cognitive radio network, PU arises from a primary transmitter (PT) and a primary receiver (PR) whereas, SU has two nodes S1 and S2 which communicate with each other bidirectionally. Thanks to ARQ mechanisms, PR sends Acknowledgment (ACK) feedback for every successful packet which reaches to PR smoothly and sends Negative Acknowledgment (NACK) for every failed packet to increase the reliability of the packet transmission. Our cognitive radio network model operates on three different modes by switching between them. SU is permitted to reach spectrum either by transmitting its own packet in underlay mode or by transmitting PU's packet to enhance PU's transmission in cooperation mode. In without SU mode, PT transmits its own packet to PR while SU eavesdrops on PT's transmission and stays silent. In addition to SU's transmission structure, PU's packets can be transmitted under interference caused by SU's transmission in underlay mode or can be transmitted while SU stays silent. Apart from these, PU's packet can be transmitted over SU's nodes while PT stays silent in cooperation mode. Two different cooperative diversity methods are exploited in cooperation mode. In the proposed spectrum sharing protocol for DF one of the SU's nodes S1, S2 or both can decode received PT's packet and transmit the packet to PR in cooperative mode. In the proposed spectrum sharing protocol for AF, without any decision mechanism at S1 and S2, PU's packet is amplified and transmitted to PR. In that way, SU aims to help PU's transmission when it is affected badly from the channel and will improve its own performance when PU's transmission is successful. Both protocols including reference schemes operate on retransmission for PU's failed packet. If PU's packet from PT to PR is failed while SU listens PU's transmission, the unsuccessful packet is retransmitted either by one of the SU's nodes or by node PT. However, the new packet is transmitted from PT to PR in underlay mode, if PU's packet is successes. To observe the effects of permitted retransmission number on the successful packet rate of PU and improve PU's QoS, retransmission number is taken equal to 2 and 3 respectively for both proposed protocols, on the other hand retransmission number is taken equal to 2 for reference schemes to make comparison with the proposed protocols' throughputs. To sum up, PU's throughput in the proposed protocols is worse than the cooperation-only scheme and is better than the underlay-only scheme, as expected. PU in the proposed protocol for DF operates in underlay mode for attempting PU's and SU's transmission at the same time in very low SNR region, where SU has a good performance. SU's throughput starts decreasing at some level of SNR for which PU operates in cooperative mode by the assistance of SU; however, SU's throughput becomes stable with increasing SNR values. Similarly, PU in the proposed protocol for AF operates in cooperative mode in very low SNR region, so SU's has no performance. SU's performance that is as good as the proposed protocol for DF is observed with increasing SNR values. Because of this characteristic way of work, PU's throughput is decreased in high SNR region while, it is better than the non-CR scheme in low SNR region. Both of the proposed protocols stabilize the SU's throughput that is near 40% with a certain sacrifice about 15% in PU's throughput at high SNR region. PU's performance in the proposed protocol for DF is a little higher compared with that for AF in low SNR region while SU's throughput is different for both protocols only in very low SNR region. It is observed that the performances are equivalent for both protocols with increasing SNR values. Since the cooperation modes in DF protocol are higher than that of AF protocol, SU's performance is worse in DF protocol compared to that of AF protocol, at low SNR region. For both protocols, effects of the system parameter on PU's and S's performance are investigated. Increasing achievable rate Rs for SU lead to decrease the SU's performance while PU's performance is not affected. Similarly, the SU's performance gets worse if the distance ds between the nodes S1 and S2 are increased. SU's power Ps is limited for a certain value of the interference threshold which is taken as the difference between PU's power Pp and Ps to protect the PU's channel quality in the underlay mode. SU's performance is decreased; conversely PU's performance gets better with increasing values of the interference threshold. Both of the proposed protocols are compared with two protocols given in the literature. One of them prioritizes PU's performance while the other prioritizes SU's performance. PU's performance for the proposed protocols is better than that of reference protocol which prioritizes the PU's performance thanks to higher amount of cooperation mode in low SNR region. With the increasing SNR values, both proposed protocols operate in underlay mode where PU have some degradation in its performance and have a throughput near 70%. SU cooperates with PU in low SNR region where SU's performance is worse than that of the reference protocol which prioritizes the PU's performance. In relation to that, SU's throughput is maintained at a certain level in high SNR region where SU's throughput goes to zero in the reference protocol which prioritizes the PU's performance. SU's throughput for both proposed protocols is worse than that of reference protocol which prioritizes the SU's performance as expected. In low SNR region, PU's performance for both of the proposed protocols is worse than that of the reference protocol which prioritizes the SU's performance. However, PU's performance declines rapidly and goes to zero for the reference protocol which prioritizes the SU's performance with increasing SNR value where PU's throughput for both of the proposed protocols has throughput performance of about 70%.
Description: Tez (Yüksek Lisans) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 2016
Thesis (M.Sc.) -- İstanbul Technical University, Instıtute of Science and Technology, 2016
URI: http://hdl.handle.net/11527/13202
Appears in Collections:Elektronik Mühendisliği Lisansüstü Programı - Yüksek Lisans

Files in This Item:
File Description SizeFormat 
10129278.pdf1.74 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.