Fuzzy Logic System for Opportunistic Spectrum Access using Cognitive Radio

Transcription

1 IJCSI Internationa Journa of Computer Science Issues, Vo. 10, Issue 1, No 1, January 2013 ISSN (Print): ISSN (Onine): uzzy Logic System for Opportunistic Spectrum Access using Cognitive Radio Kaniezhi. R 1, Danie Nesa Kumar. C 2 and Prakash. A 3 1 Research Schoar, Department of Computer Science, Periyar University, Saem, TamiNadu , India 2 Asst Professor, Dept Of Computer Appications, Hindusthan coege of arts and science, Coimbatore, India 3 Asst Professor, Dept Of Computer Appications, Hindusthan coege of arts and science, Coimbatore. Abstract Opportunistic spectrum access approach is enabed by cognitive radios which are abe to sense the unused spectrum and adapt their operating characteristics to the rea-time environment. However, a naive spectrum access for secondary users can make spectrum utiization inefficient and increase interference to adjacent users. In this paper, we propose a nove approach using uzzy Logic System (LS) to contro the spectrum access and spectrum for secondary user is seected by possibiity of the spectrum access instead of considering the antecedents. Keywords: Antecedent, LS, Interference, Spectrum access, spectrum utiization. 1. Introduction In recent studies, the spectrum aocated by the traditiona approach shows that the spectrum aocated to the primary user is under-utiized and the demand for accessing the imited spectrum is growing increasingy. Spectrum is no onger sufficienty avaiabe, because it has been assigned to primary users that own the privieges to their assigned spectrum. The concept of opportunistic spectrum access is used in order to efficienty utiize the spectrum which is underutiized. The opportunistic spectrum access improves the spectrum utiization by cognitive radio adopting the secondary user to use the unused spectrum of the primary user. This opportunistic spectrum access, avoids the spectrum scarcity. The idea of cognitive radio was first presented officiay in an artice by Joseph Mitoa III and Gerad Q. Maguire, Jr in Reguatory bodies in various countries found that most of the radio frequency spectrum was inefficienty utiized. or eampe, ceuar network bands are overoaded in most parts of the word, but amateur radio and paging frequencies are not. This can be eradicated using the dynamic spectrum access. The key enabing technoogy of dynamic spectrum access techniques is cognitive radio (CR) technoogy, which provides the capabiity to share the wireess channe with icensed users in an opportunistic manner. rom this definition, two main characteristics of the cognitive radio can be defined as foows: Cognitive capabiity: It refers to the abiity of the radio technoogy to capture or sense the information from its radio environment. Through this capabiity, the portions of the spectrum that are unused at a specific time or ocation can be identified. Consequenty, the best spectrum and appropriate operating parameters can be seected. Reconfigurabiity: The cognitive capabiity provides spectrum awareness whereas reconfigurabiity enabes the radio to be dynamicay programmed according to the radio environment. More specificay, the cognitive radio technoogy wi enabe the users to (1) determine which portions of the spectrum is avaiabe and detect the presence of icensed users when a user operates in a icensed band (spectrum sensing), (2) seect the best avaiabe channe (spectrum management), (3) coordinate access to this channe with other users (spectrum sharing), and (4) vacate the channe when a icensed user is detected (spectrum mobiity). Copyright (c) 2013 Internationa Journa of Computer Science Issues. A Rights Reserved.

2 IJCSI Internationa Journa of Computer Science Issues, Vo. 10, Issue 1, No 1, January 2013 ISSN (Print): ISSN (Onine): The paper is organized as foows; Section 2 and 3 provide the reated work and fuzzy ogic system for its impementation. In section 4 and 5, opportunistic spectrum access by uzzy ogic system to improve the spectrum efficiency and performance anaysis. inay, concusions are presented in Section Reated Work In the research iterature on the opportunistic spectrum access, different methods are appied using fuzzy ogic to access spectrum efficienty. In [7], a nove approach using uzzy ogic system provides the possibiity of accessing spectrum band for secondary users and the user with the greatest possibiity has to be assigned the avaiabe spectrum band. or enhancing the performance of Cognitive radio fuzzy ogic based scheme is deveoped by Aniesh Dey et a., [8], where efficient spectrum utiization depends upon the ink oss, hod time and interference temperature. With fuzzy controer, Cognitive radio opportunisticay adjust its transmit power in response to the changes of the interference eve to the primary user is discussed in [9]. A uzzy ogic based agorithm [10] is used to reduce the spectrum handoff and improves the probabiity of PU s occupancy at a certain channe. In [11], uzzy rues are used to optimize the bandwidth aocation based on three antecedents as: arriva rate of both icensed and unicensed users and the avaiabiity of unoccupied number of channes within the system. A decentraized method has been deveoped using fuzzy based techniques for both channe estimation and channe seection in [12]. The efficient decision making in the cognitive radio by fuzzy ogic is aso discussed by Matinmikko et a [13], which presented an overview of appication of fuzzy ogic to teecommunications. Our work is different from the above said works avaiabe on Cognitive radio based on fuzzy ogic system. This paper presents a nove approach using uzzy ogic system to utiize the avaiabe spectrum by the secondary users without interference with the primary user. The secondary users access the spectrum based on the highest possibiity of the secondary users. 3. uzzy Logic 3.1 uzzy sets uzzy sets theory is an eceent mathematica too to hande the uncertainty arising due to vagueness. uzzy sets may be viewed as an etension and generaization of the basic concepts of crisp sets. An important property of fuzzy set is that it aows partia membership. uzzy set is the description of fuzzy events and conceptions. The fuzzy events means some events that there are no strict boundary and cannot be characterize easiy by eact measurement. uzzy ogic theory can mode the vagueness of the rea word. A uzzy Logic System (LS) is unique in that it is abe to simutaneousy hande numerica data and inguistic knowedge. It is a noninear mapping of an input data (feature) vector into a scaar output, i.e., it maps numbers into numbers. L provides a simpe way to arrive at a definite concusion based upon vague, ambiguous, imprecise, noisy, or missing input information. L's approach to contro probems mimics how a person woud make decisions, ony much faster. igure 1 shows the structure of a fuzzy ogic system. ig. 1. The Structure of the uzzy Logic System Since there is a need to fuzzify the fuzzy resuts we generate through a fuzzy system anaysis i.e., we may eventuay find a need to convert the fuzzy resuts to crisp resuts. Here, we may want to transform a fuzzy partition or pattern into a crisp partition or pattern; in contro we may want to give a singe-vaued input instead of a fuzzy input command. The dufuzzification has the resut of reducing a fuzzy set to a crisp singe-vaued quantity, or to a crisp set. Generay, uzzy Logic and uzzy decision making is divided into three consecutive phases namey uzzification, uzzy reasoning and Dufuzzification [17]. 1. uzzification: The input variabes are fuzzified using predefined membership functions (M). Unike in binary ogic where ony 0 and 1 are accepted, aso numbers between 0 and 1 are used in fuzzy ogic. This is accompished with M to which the input variabes are compared. The output of fuzzification is a set of fuzzy numbers. Copyright (c) 2013 Internationa Journa of Computer Science Issues. A Rights Reserved.

3 IJCSI Internationa Journa of Computer Science Issues, Vo. 10, Issue 1, No 1, January 2013 ISSN (Print): ISSN (Onine): uzzy reasoning: uzzy numbers are fed into the predefined rue base that presents the reations of the input and output variabes with I THEN Causes. The output of the fuzzy reasoning is a fuzzy variabe that is composed of the THEN causes. y ʹ cos( ) M p C T 1 ( ʹ1 ) G 1 M p T ( 1 ʹ1 ) where M is the number of rues in the LS. 1 (2) 3. Dufuzzification: The output of the fuzzy reasoning is changed into a non-fuzzy number that represents the actua output of the system. 3.2 Membership unctions Consider a fuzzy ogic system with a rue base of M rues, R and et the th rue be denoted by. Let each rue have p antecedents and one consequent (as is we known, a rue with q consequents can be decomposed into rues, each having the same antecedents and one different consequent), i.e., it is of the genera form R : I u 1 is 1 and u2 is 2 and and u p p is, THEN v is G. u,k 1,...p where k and v are the input and output inguistic variabes, respectivey. Whie appying a singeton fuzzification, when an input X ' ' ' ' ' 1, 2, 3,... p is appied, the degree of firing corresponding to the th rue is given by n i. (i ) n ( i ) (1) Where denotes a T-norm, n represents the number of eements, i ( ) s are the eements and i is its membership function. There are many kinds of dufuzzification methods, but we have chosen the centre of sets method for iustrative purpose. It computes a crisp output for the LS by first C computing the centroid, G of every consequent set G and, then computing weighted average of these centroids. The weight corresponding to the th rue consequent centroid is the degree of firing associated with the th rue, p ' T ( 1 ) 1, so that We use trapezoida membership functions (Ms) to represent near, ow, far, high, very ow and very high, and triange Ms to represent moderate, ow, medium and high. Ms are shown in ig. 2, 3, 4. Since we have 3 antecedents and fuzzy subsets, we need setup rues for this LS. ig. 2. Membership unction (M) used to represent the antecedent1 ig. 3. Membership unction (M) used to represent the antecedent2 Copyright (c) 2013 Internationa Journa of Computer Science Issues. A Rights Reserved.

4 IJCSI Internationa Journa of Computer Science Issues, Vo. 10, Issue 1, No 1, January 2013 ISSN (Print): ISSN (Onine): In our approach, we combine the three antecedents to aocate spectrum opportunisticay inorder to find out the optima soutions using the fuzzy ogic system. Mobiity of the secondary user pays a vita roe in the proposed work. Wireess systems aso differ in the amount of mobiity that they have to aow for the users. Spectrum Mobiity is defined as the process when a cognitive radio user echanges its frequency of operation. The movement of the secondary user eads to a shift of the received frequency, caed the Dopper shift. ig. 4. Membership unction (M) used to represent the antecedent3 Linguistic variabes are used to represent the spectrum utiization efficiency; distance and degree of mobiity are divided into three eves: ow, moderate, and high. We use 3 eves to represent the distance, i.e., near, moderate. The consequence, i.e., the possibiity that the secondary user is chosen to access the spectrum is divided into five eves which are very ow, ow, medium, high and very high. 4. uzzy Logic for Opportunistic Spectrum Access We design the fuzzy ogic for opportunistic spectrum access using cognitive radio. In this paper, we are seecting the best suitabe secondary users to access the avaiabe users without any interference with the primary users. This is coected based on the foowing three antecedents i.e., descriptors. They are Ant 1 : Spectrum Utiization Efficiency Ant 2 : Degree of Mobiity Ant 3 : Distance of Secondary user to the PU. uzzy ogic is used because it is a muti-vaued ogic and many input parameters can be considered to take the decision. Generay, the secondary user with the furthest distance to the primary user or the secondary user with maimum spectrum utiization efficiency can be chosen to access spectrum under the constraint that no interference is created for the primary user. We appy different avaiabe spectrum inorder to find out the spectrum efficiency which is the main purpose of the opportunistic spectrum access strategy. Hence, we cacuate the spectrum efficiency as the ratio of average busy spectrum over tota avaiabe spectrum owned by secondary users. Since we chose a singe consequent for each rue to form a rue base, we averaged the centroids of a the responses for each rue and used this average in pace of the rue consequent centroid. Doing this eads to rues that have the foowing form: R ' : If Degree of mobiity ( 1 1 ) is ; and its distance between primary user and the secondary users ( 2 ) is 2 ; and the spectrum utiization efficiency of the secondary user ( 3 3 ) is, then the Possibiity (y) c choosing the avaiabe spectrum is avg, where = c 1,2,..27 and avg is defined as foows: c avg 5 i w i c i 1 5 w i i 1 (3) wi is the number of choosing inguistic abe i for which the consequence of rue and c i is the centroid of the i th consequence set (i: 1; 2;...; 5; : 1; 2;...; 27). Tabe 2 i provides c for each rue. or every input ( 1,2, 3 ), the output y( 1,2, 3 ) of the designed LS is computed as Copyright (c) 2013 Internationa Journa of Computer Science Issues. A Rights Reserved.

5 IJCSI Internationa Journa of Computer Science Issues, Vo. 10, Issue 1, No 1, January 2013 ISSN (Print): ISSN (Onine): c avg 1 ( 1 ) 2 ( 2 ) 3 ( 3 ) y( 1, 2, 3 ) 27 1 ( 1 ) 2 ( 2 ) 3 ( 3 ) (4) which gives the possibiity that a secondary user is seected to access the avaiabe spectrum. In igure 5, shows that the Spectrum user (SU4) having the highest possibiity of accessing the spectrum than the other users. Even though the other secondary users have the furthest distance from primary user to the secondary users, highest spectrum utiization and highest mobiity but we prefer SU4 to access the spectrum because it has the highest possibiity i.e., ig. 6. Mean Arriva VS Ca bocking rate igure 6 shows the ca bocking of the service provider using the uzzy ogic system. As the ca arriva rate increases the bocking rate gets decreased. Traffic rate increases aong with the ca bocking rate. ig. 5. Possibiity of choosing spectrum for opportunistic spectrum access Thus, the secondary user wi seect the spectrum for accessing based on the highest possibiity rather than the highest spectrum utiization and the furthest distance from the primary user. 4. Performance Anaysis In this section, we present simuation resuts on the performance of our proposed work based on uzzy ogic System. In the proposed work, we are choosing the avaiabe channe with the high possibiity and high spectrum utiization efficiency. ig. 7. Mean Arriva VS Interference using LS When interference increases spectrum utiization wi decrease. The igure 7 shows that there is a decrease in the interference which provides an increase spectrum utiization using LS. This eads to efficient spectrum utiization. Copyright (c) 2013 Internationa Journa of Computer Science Issues. A Rights Reserved.

6 IJCSI Internationa Journa of Computer Science Issues, Vo. 10, Issue 1, No 1, January 2013 ISSN (Print): ISSN (Onine): Concusions The proposed approach using a uzzy Logic System detects the effective spectrum access for secondary users via cognitive radio. The secondary users are seected on the basis of spectrum utiization, degree of mobiity and distance from secondary users to the primary user. Our designed LS is used to contro the spectrum assignment and access procedures in order to prevent mutipe users from coiding in overapping spectrum portions. Therefore, the secondary user with the highest possibiity is guaranteed to access the spectrum. Acknowedgments ig. 8. Distance of Secondary Users to the Primary users using LS As iustrated in the igure 8, the distance from primary user to the secondary users is shown. The distance between primary user and the secondary users heps us for cacuating the possibiity for accessing the spectrum opportunisticay. ig. 9. Mean Arriva VS Channe Utiization using LS The Spectrum Efficiency (Channe Utiization) is defined as the ratio of average busy channes over tota channes owned by service providers. The igure 9 shows that there is an increase in channe utiization with decrease in the ca arriva rate. The irst Author etends her gratitude to UGC as this work was supported in part by Basic Scientist Research (BSR) Non SAP Scheme, under grant reference no /2008(bsr) UGC XI Pan. References [1] H.J. Zimmermann, uzzy Set Theory and its Appications, ISBN , ourth Edition, Springer Internationa Edition, [2] Timothy J. Ross, uzzy Logic with Engineering Appications, ISBN , Second Edition, Wiey Student Edition, [3] James J. Buckey, Esfandiar Esami, An Introduction to uzzy Logic and uzzy Sets, ISBN , Physica Verag, [4] Andreas. Moisch, Wireess Communications, ISBN: , Second Edition, John Wiey & Sons Ltd, [5] R. Kaniezhi, Dr. C. Chandrasekar, S. Nithya Rekha, Dynamic Spectrum Sharing for Heterogeneous Wireess Network via Cognitive Radio, Proceedings of the Internationa Conference on Pattern Recognition, Informatics and Medica Engineering, March 21-23, 2012, pp [6] R. Kaniezhi, Dr. C. Chandrasekar, Mutipe Service providers sharing Spectrum using Cognitive Radio in Wireess Communication Networks, Internationa Journa of Scientific & Engineering Research (IJSER), Voume 3, Issue 3, eb 2012, pp.1-6. [7] R. Kaniezhi, Dr. C. Chandrasekar, Spectrum Sharing in a Long Term Spectrum Strategy via Cognitive Radio for Heterogeneous Wireess Networks, Internationa Journa on Computer Science and Engineering (IJCSE), Voume 4, No.6, June 2012, pp [8] R. Kaniezhi, Dr. C. Chandrasekar, S. Nithya Rekha, Performance Evauation of QoS Parameters in Spectrum Sharing using SBAC Agorithm, IEEE Internationa Conference on Advances in engineering, Science and Management (IEEE-ICAESM 2012), Nagapattinam, March 30-31, 2012, pp [9] R. Kaniezhi, Dr. C. Chandrasekar, An Efficient Spectrum Utiization via Cognitive Radio using uzzy Logic System Copyright (c) 2013 Internationa Journa of Computer Science Issues. A Rights Reserved.

7 IJCSI Internationa Journa of Computer Science Issues, Vo. 10, Issue 1, No 1, January 2013 ISSN (Print): ISSN (Onine): for Heterogeneous Wireess Networks, INCOSET 2012, Dec 13-14, [10] R. Kaniezhi, Dr. C. Chandrasekar, Comparing Spectrum Utiization using uzzy Logic System for Heterogeneous Wireess Networks via Cognitive Radio, Internationa Journa of Scientific & Engineering Research, Voume 3, Issue 7,Juy 2012, pp [11] Jerry M. Mende, uzzy Logic Systems for engineering: A Tutoria, IEEE Proc., March 1995, vo. 83, no. 2, pp [12] Ia Sharma, G. Singh, A Nove Approach for Spectrum Access Using uzzy Logic in Cognitive Radio, I.J. Information Technoogy and Computer Science, 8, 2012, pp [13] Aniesh Dey, Susovan Biswas, Saradindu Panda, and Santanu Monda, A New uzzy Rue Based Spectrum Utiization and Spectrum Management for Cognitive Radio, Nationa Conference on Eectronics, Communication and Signa Processing, NCECS-2011, 19th September 2011, pp [14] Hong-Sam T. Le and Qiian Liang, An Efficient Power Contro Scheme for Cognitive Radios, IEEE conference on Wireess Communications and Networking, 2007, pp [15] Tang Wanbin, Peng Dong, SPECTRUM HANDO IN COGNITIVE RADIO WITH UZZY LOGIC CONTROL, JOURNAL O ELECTRONICS (CHINA), Vo.27 No.5, September 2010, pp [16] Prabhjot Kaur, Moin Uddin, Arun Khosa, Adaptive Bandwidth Aocation Scheme for Cognitive Radios, Internationa Journa of Advancements in Computing Technoogy, Voume 2, Number 2, June, 2010, pp [17] Aa A-uqaha, Bia Khan, Ammar Rayes, Mohsen Guizani, Osama Awwad, Ghassen Ben Brahim, Opportunistic Channe Seection Strategy for Better QoS in Cooperative Networks with Cognitive Radio Capabiities, IEEE Journa on Seected Areas in Communications, vo. 26, no. 1, January 2008, pp [18] Marja Matinmikko, Tapio Rauma, Miia Mustonen,Ikka Harjua, HeiS Arvanko and Aarne Mammea, Appication of fuzzy ogic to cognitive radiosystems, IEICE Trans. Communication, vo. E92B, no.12, December 2009, pp R. Kaniezhi is a member of the IEEE and she received her B.Sc Degree from University of Madras in She received her MCA and M.Phi degrees from Periyar University and Annamaai University in 2001 and 2007, respectivey. Her research interests incude mobie computing, Cognitive Radio and wireess communication. C. Danie Nesa Kumar competed his MCA in Bishop Heber Coege (Autonomous), Bharathidasan University, and Trichy. He is currenty working as an Assistant Professor in Hindusthan Coege of Arts & Science. His Research area is Networking. A. Prakash competed his M.Sc (CT), MCA, M.Phi., Ph.D. Currenty he is working as an Assistant professor in Department of Computer Science & Appications, Hindustan Coege of arts and Science, Bharathiar University, Coimbatore. His Area of Interest is Data Mining. Copyright (c) 2013 Internationa Journa of Computer Science Issues. A Rights Reserved.