Spectrum Sensing For Cognitive Radio In Vehicular Ad Hoc Network

A spectrum resource is one of the main components of wireless networks,continuously fixed allocations of this limited resource lead to depletion of the spectrum band.So,the efficient usage share-capability of this resource will continue as the main direction of the topics the attract the researchers.New emerging application or services or extension of the present one,such as Vehicular Ad Hoc Network,require in most cases more band to use.Therefore,the inefficient usage and lack of spectrum have called a new paradigm in a wireless communication system where the available wireless spectrum bands are exploited opportunistically.In such a paradigm,reliable communication is provided wherever and whenever is needed radio spectrum is used more efficiently and that granted future emerge services the required spectrum.In the cycle of cognitive radio,a cognitive radio monitor's spectrum band catches their information then detects the available spectrum spaces.As an essential stage in cognitive radio,spectrum sensing aims to decide whether a particular part of the spectrum is being occupied or not-that is,differentiating between the presence and absence of the PU.The cognitive users(CUs),called also secondary users(SUs),in cognitive radio systems,are needed to perform transmissions such that they do not hinder the primary user(PU)transmission.This mainly involves recognition of the PU with utilizing one of diverse spectrum sensing techniques.So that can protect the right of primary user in case that it uses own spectrum,and at the same time use the band in the absence of PU.Currently,cognitive radio network(CRN)has been envisaged as a promising methodology to tackle the dilemma of spectrum scarcity by enabling the opportunistic access of the unoccupied frequency bands without interfering with the primary users.The Federal communication committee(FCC)stated that cognitive users should assure utilizing the spectrum with limited or no interference to the primary user.Therefore,spectrum sensing is a mandatory function of cognitive radio systems as it enables the cognitive user to abide by this stipulation.The sensing phase for spectrum can be achieved in multidimensional.it could be in opportunity on the dimension of frequency,on account of that not all the bands are used simultaneously i.e.might be,some bands available for usage opportunistically.Also,the opportunity in the dimension of time is one of the potential scopes that cognitive system users could use the primary network resources,because of that the band is not used continuously.Hence,there are times where it would be available to usage opportunistically.At a given time,the spectrum band can be available in some parts of the geographical area and reserved in other.Taking advantage of path loss in space,these measurements could be obviated by looking at a level of interference.When there is no interference that indicates no primary user transmission in that local area,so the spectrum at the location is available to be used by cognitive users.The spreading code and frequency hopping(FH)sequences used by the primary users open a new outlook to explorer more efficient usage for the rare spectrum resource.Over a wideband,the spectrum band might be used at a certain time through frequency hopping or spread spectrum.This doesn't mean there is no availability over this band.Hence,simultaneous usage of the band without interfering with license users could be potential in code domain with orthogonality of code respecting to codes that primary users are being used.In code domain,besides spectrum usage detecting,also determining the used codes by primary users is required as a necessary mission to be achieved by the cognitive system,and possibly the parameters of multipath as well need to be known.The direction of the beam for the primary user and its location can be exploited,with getting knowledge for the location or direction of PUs,opportunities of the spectrum can be created in the angle dimension.For instance,when a PUs is transmitting in a particular direction,the CUs can transmit,too,in other directions without causing interference on PUs.Each cognitive user in the cognitive system can sense the state of the primary user spectrum band and make its individual and autonomous decision.However,the destructive effects of wireless channels including shadowing,fading and path loss might render single SU spectrum sensing as unreliable.Cooperative spectrum sensing(CSS)is used to obviate this problem.The main idea of CSS is to make a global decision about the availability of the PU by combining the sensing results originated by multiple cognitive usersIn cooperative spectrum sensing(CSS),that many cognitive users are involved in the cooperative procedures,the possible correlation between cognitive users degrades the sensing performance as it diminishes the gain of spatial diversity.Since spectrum utilization results that are observed by closely situated cognitive users may be highly correlated as a result of the inherent spatial correlation in the received primary users'signals due to multipath and mainly because of correlated shadow fading,the local decisions of cognitive users,in this case,are correlated.Even though the independence postulate on the local detectors'observations simplifies the performance analysis of cooperative spectrum sensing,this presumption is not feasible in the case where in fact the proximity of the cognitive users results in correlated observations.To make sure trusted detection,the correlation among various cognitive users should also be taken into consideration.Thus,it was revealed that having a small number of cognitive users over a large distance might be more efficient than a large number of closely located cognitive users in correlated shadowing scenario.Therefore,it is advantageous to select uncorrelated cognitive users for cooperation process among cognitive users to enhance the cooperative gain with a smaller number of secondary users.So,the selection of suitable uncorrelated secondary users(SUs)requires the evaluation of the correlation measurement between cognitive users.Under log-normal shadowing model,the observations of two closely located cognitive users might be considered as a correlated because of their proximity,when it exceeds a certain threshold,that can be computed as an exponential function that depended mainly in the distance between the users.As long as the distance between the cognitive users getting smaller the correlation coefficient is approaching the unity,in this case,cognitive users experience more shadowing effects called spatially correlated shadowing.It was shown that spatial correlation due to shadowing degrade the performance of detection(P_d)and compromise the doable cooperative gain.In addition to that,the correlation influences the performance of cooperative spectrum sensing in term of sensing time.Let's say there are four cognitive radio users(CRUs)in the cognitive network,which are in a correlated state with each other,due to their correlation the,the sensing measurement will be the same.So,taking the measurement of single cognitive user among them or taking the measurements of all four CRUs is same in perspective of probability of detection,but fusion center(FC)will consume less time on deciding the existence or absence of PU by taking the measurement of only a single secondary user instead of all four CRUs.The work in the thesis,illustrates that increasing in the power to noise ratio(SNR)has apparent enhancement influence in the spectrum sensing performance that leads to more resource usage efficiency,due to less interference between cognitive users and primary users as a consequence for enhancing the detection probability in the cognitive system.On the contrary,the decreasing in SNR to very low-level leads degradation in the probability of detection(P_d)toward the boundary of false alarm probability(P_f).The thesis shows,that the number of secondary users that participate in cooperative spectrum sensing process has a significant impact on the performance of sensing,increasing number of cognitive users will produce increasing rates in both detection and false alarm probabilities,but the rate of increase of P_f is less than P_d and this the advantage is guaranteed by the cumulative SNR of users.The work reveals the effect of SNR on performance for the three schemes,noncooperative,soft fusion,and hard fusion,and as a general conclusion that increasing the level of SNR leads increasing in the performance of spectrum sensing is true for all decision-making schemes and for all cases either static or VANET,and it appears with the distribution of users in the works that soft decision making is overcome hard fusion,and the performance of VANET case is worse than for static case due to bad signal propagation conditions as a result for the mobility.Influence of the number of cognitive users that participate in the cooperative sensing on performance for all decision-making schemes is explorer,the main result is in general,that the increase of the amount of cognitive user enhances spectrum sensing performance is valid for all decision-making schemes and for static and vanet cases.Also,with an increase of CUs number soft decision scheme performance still better than for hard decision and the performance of the static case better than for VANET/mobile case.So,with taking advantage of our findings either for signal to noise ratio effect in the performance of sensing,number of cognitive users that participate in the cooperative,and the performance of the different decision-making schemes,in this thesis we implementing the process that fulfills better detection performance for the situation of the sensed spectrum during sensing period needs by employing the appropriate steps to get such a reliable decision beside gain more resources by reduction the overhead needs for the process.These steps could be summarizing as;Firstly,utilizing cooperative spectrum Sensing(CSS)scheme,that show its improvement in the detection comparing to non-cooperative scheme,which is necessary to obtain the right decisions for the condition of channel either idle or occupied by the primary user,and that also helps to overcome the matters of hidden terminal,receiver uncertainty,noise uncertainty,and other issues that are related to uncooperative schemes,taking in the consideration that the energy detection is the detection approach in this work.Then,selecting a subset of cognitive users,among all cognitive users that involved in the cooperation process,based on their higher signal to noise ratio to perform the cooperative spectrum sensing process.Those nodes that can receive higher primary signal are more candidate to get more reliable information about the sensed channel whether it's under usage state or not.Whereas the nodes that have a very low signal to noise ratio have a negative impact on the performance detection.And it's noted that the high level of the received signal for the primary user in one of secondary user has more positive impact for enhancing detection in the soft fusion scheme compared to the hard scheme.Reduction in a number of cognitive users that perform the cooperative sensing due to the selection stage,also gain the system more resource to use for users'traffic due to minimizing the overhead that is consumed in the reporting phase.The work is shown the result in VANET and static cases and with a different scheme of cooperative spectrum sensingNext,We implement cooperative spectrum sensing in vehicles(CSSV)selection approach by first employing the general conclusion of our works regarding the influence of signal to noise and the more appropriate fusion combination,second by apply stability for the mobile nodes,and at the last,for the selected nodes spatial correlation process will be done to get more reliable and robust spectrum sensing and mitigate the shadow effect.So,with the proposed algorithm,the nodes that have very low signal to noise ratio are neglected to participate in spectrum sensing process due to that they decrease the performance of detection that will be more reliable by electing nodes that have high SNR.Then,the cognitive users,vehicles,that have a lower speed than a predefine threshold speed will be chosen as they give more reliable and accurate sensing result comparing to those vehicles with very high speed,that exceed predefined threshold speed.The final phase for implementing CSSV is considering the shadow correlation effect by measurement the distance between those nodes that pass SNR and stability phases,with predefine uncorrelated distance.So only users that separated by a distance more than the uncorrelated distance they will be treated as an uncorrelated node.Therefore,they can participate in the cooperative process.The proposed approach CSSV shows its better performance compared to the conventional approach that can give better performance of detection with a lower number of involved cognitive users in the cooperative process.Thus,the spectrum can be used more efficiently with less interference to primary users too.