fuzzy based call admission control in wideband cdma cellular networks full report
project topics Active In SP Posts: 2,492 Joined: Mar 2010 
26042010, 09:44 AM
Wideband codedivision multiple access.pdf (Size: 190.49 KB / Downloads: 68) Topic: fuzzy based call admission control in wideband cdma cellular networks ABSTRACT In this paper, a novel call admission control (CAC) scheme using fuzzy logic is proposed for the reverse link transmission in wideband code division multiple access (CDMA) cellular communications. The fuzzy CAC scheme first estimates the effective bandwidths of the call request from a mobile station (MS) and its mobility information, and then makes a decision to accept or reject the connection request based on the estimation and system resource availability. Numerical results are given to demonstrate the effectiveness of the proposed fuzzy CAC scheme in terms of new call blocking probability, handoff call dropping probability, outage probability, and resource utilization. PAPER SUBMITTED BY: M.Anand, DECE, J.David, DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING V.L.B.JANAKIAMMAL COLLEGE OF ENGG AND TECH KOVAIPUDUR Coimbatore 1 INTRODUCTION Wideband codedivision multiple access (CDMA) cellular systems for third generations (and beyond) wireless communications are expected to support multiple services with guaranteed quality of service (QoS). Sophisticated resource management techniques are needed to make efficient use of the available radio resources. Call admission control (CAC) is one of the resource management functions, which regulates network access to ensure QoS provisioning. It is the decisionmaking component of the network to guarantee the QoS requirements and, at the same time, to achieve system resource utilization as efficiently as possible. However, to design and efficient and practical CAC scheme a very challenging issue due to user mobility, limited radio spectrum, and multimedia traffic characteristics. In a cellular system, user mobility results in handoff calls. In order to use limited radio resources to accommodate and increasing demand for services, one approach is to deploy smaller size cells for more frequency reuse, a consequence of which is more frequent handoffs. From a user's point of view, it is better to be blocked at the beginning of a connection than to be dropped during the connection. As a result, handoff calls should be given higher priority than new calls by reserving some resources exclusively for handoff calls. Mobile station (MS) mobility information is required in order to determine the right amount of resources that should be reserved; Overreservation leads to low resource utilization, while underreservation results in a high handoff call dropping probability. Due to the random nature of user mobility and dynamic characteristics of multimedia traffic, QoS provisioning and high resource utilization are always conflicting goals. The CAC scheme should be designed to achieve the best compromise between the two goals. For example, to maximize resource utilization under the constraint of QoS satisfaction, the resource reservation should be a function of user mobility information, traffic load and characteristics, Qos requirements, etc. To address technical challenges in CAC for wireless communications, in this paper, we investigate CAC for a wideband CDMA cellular system using the fuzzy logic without making simplified assumptions on user mobility and teletraffic models. Traditionally, a CAC scheme should explore a set of measured parameters to make the decision of accepting or rejecting a requesting call. This type of control scheme makes little or no allowance for measurement uncertainties. However, in the wireless system, due to user mobility and varying channel condition, the measurements obtained are, in general, not accurate. Furthermore, it is also difficult to obtain the complete static's of input traffic. As a result, the decision has based on the imprecision and uncertain measurements. To this end, fuzzy logic provides and approximate but effective means of describing the behavior of the systems that are too complex and not easy to tackle mathematically. The fuzzy CAC scheme proposed here takes into account the MS mobility information is obtained via signal power measurements, without using a predetermine or simplified model. Both intracell and intercell effective bandwidths are used to determine the resource requirement of each call. When an MS requests for service, first the fuzzy CAC Scheme estimates its effective bandwidths in the target cell and neighboring cells and its mobility information, which facilitate the CAC decision and resource reservation for the call if admitted. Then, the fuzzy CAC scheme makes a decision to accept the call only if the QoS requirements of all the ongoing calls in the target cell and the neighboring cells can be guaranteed and the QoS requirement of the new call can ensured. Simulation results are presented to demonstrate the fuzzy CAC scheme can achieve low new call blocking and handoff call dropping probabilities, satisfy the outage probability requirement, and make efficient use of system resources. 2 SYSTEM MODEL a wideband CDMA cellular system with hexagonal cells of equal size. Each cell contains a centrally located base station (BS) . The same radio spectrum is reused in every cell. The MSs communicate with their home BSs via the air interface, and a number of BSs are connected to a mobile switching center (MSC) which is in turn connected to a backbone wire line network. Separate frequency bands are used for the forward and reverse links, so that each BS experiences interference only from the MSs. The focus is placed on the CAC process for the reverse link. In the forward link, each BS broad casts a unique pilot signal to the MSs. The pilot signals from different BSs are distinguished by different scrambling codes. The MS can detect the pilot signal from any BS when the strength of the pilot signal is above a certain level. Prior to transmission, an MS monitors the received pilot signal power. It is assumed that MSs, BSs, and MSCs are properly designed such that, while tracking the signal from the home BS, and MS searches for all the possible pilots and maintains a list of all pilots whose signal power levels are above a prescribed threshold. This list is transmitted to the MSC periodically through the home BS. The MSC uses the information to make a decision when handoff should start and to estimate the mobility information. In this paper, the mobility information is defined to be the probabilities that an MS will be active in other cells at future moments. When a call request arrives, the MSC uses the mobility information of all the MSs in the cell and in the neighboring cells, together there are sufficient resources available to accommodate the new call without degrading QoS of the calls already in service. MS under consideration as located at point M. For simplicity of presentation, we will focus on the mobility information and effective bandwidths of the MS related to its home BS (denoted as BS 0) and to its six firsttiers neighboring BSs (denoted as BS 1, BS2 .. .BS6). 3 DESIGN OF THE FUZZY CAC SHEME Power Levels (Eb/Io) s QoS Parameters Network Resource Estimator The call admission controller consists of four subsystems: The fuzzy effective bandwidth estimator estimates the intercell effective bandwidth of a call according to its intracell effective bandwidth, which is determined by its traffic characteristics, and the received pilot signal power levels measures at the MS. The fuzzy mobility information estimator estimates the MS mobility information according to the received pilot signal power levels measured at MS. The mobility information is used to determine the amount of bandwidth to be reserved in each of the neighboring cells for handoff calls. The network resource estimator estimates the available resource that the system can supply according to the measured (Eb/Io) s value of traffic class bandwidth reservation information. The fuzzy call admission processor makes the decision on whether to accept or reject the call request according to the MS effective bandwidth, mobility information and resource availability in the system. 3.1 FUZZY EFFECTIVE BANDWIDTH ESTIMATOR Measured pilot signal power Fuzzy sets in U(P0) and U(P1) Fuzzy sets in U(Bl) Effective Bandwidth Estimates B Fuzzifier Fuzzy Inference Engine Defuzzifier w w Fuzzy Rule Base The fuzzy effective bandwidth estimator shown in Fig. 3 is a knowledge system. It consists of a fuzzifier, a fuzzy rule base, a fuzzy inference engine, and a defuzzifier. The inputs to the estimator are the received pilot signal power from the home BS measured at the MS, Po and the received pilot signal power levels from the lth neighboring BS measured at the MS, Pl, l=1,2 6. The outputs of the estimator are the numerical value of the intercell effective bandwidths Bl, l=1,2...6. Within the estimator, the fuzzifier translates the input numerical measurements to the corresponding linguistic values of the fuzzy sets in the input universe of discourse given by U(Po)={ES(extremely small), VS(very small), S(small), SM (small to medium), M(medium),ML (medium to large),L(large),VL(very large), EL(extremely large)}, and U(Pl)={ES,VS,S,SM,M,ML,L,VL,EL} (L=1,2..6), respectively. Each fuzzy set is a set with continuous membership functions and can be represented as a set of ordered pairs of generic element, Po or P and its membership value as ES= {(P,j,ES(P)PeU}. The output of fuzzy inference engine are the linguistic variables Bl(l=1,2,..6), representing the intercell effective bandwidth in cell l. Each of the output linguistic variables is defined on the term set. U(B)={ES,VS,S,SM,M,ML,L,VL,EL} The membership function of both input and the output linguistic variables depend on the traffic characteristics, the cell size, transmit pilot signal power, path loss exponent, and channel shadowing statistics. The functions can be determined by using data collected from various sensors and/or field measurements. For exaple mth rule has the following form: If(Po is PFom) and (P1 is PF") and ... and (p6 is PF6m) Then (B1 is BF1m) and . and (B6 is BF6m), Where m = 1,2,...,M, and M is the total number of the fuzzy rules, (P0, P1,_,p6) e U(P0) x U(P) x ... x U(P) and (B^,..^) e U(B) x U(B) x ... x U(B) are linguistic variables, PFom, PFlm, and BFlm (Vl e {1,2, ... , 6}) are fuzzy sets in U(P0),U(P), and U(B), respectively. In the fuzzy inference engine, fuzzy logic principles are used to combine the fuzzy IfThen rules in the fuzzy rule base into a mapping from the fuzzy sets of the pilot signal powers to the fuzzy sets of the effective bandwidths. 3.2 FUZZY MOBILITY INFORMATION ESTIMATOR AND BANDWIDTH RESERVATION Measured pilot signal Fuzzy sets in power U(P0) and U(P1) Fuzzy sets in U(Bl) Mobility Information Mobility Estimates Pn Informtion prediction Fuzzifier Fuzzy Inference Defuzzifier proedictor Fuzzy Rule Base User mobility information can be used to assist user mobility management, to manage network resources, and to analyze handoff algorithms in wireless networks. Different from the previous research efforts on mobility information, which focused on statistics such as mobility model, user location tracking and trajectory prediction, channel holding time, cell boundary crossing rate, mean handover rate, and cell residence time, we are interested in the probabilities that an MS will be active in a neighboring cell at future moments. The mobility information with reasonable accuracy facilitates statistical multiplexing and plays and important role in efficient resource management of the wireless cellular system. In general, if an MS is closer to a BS, then the propagation path attenuation from the BS to the MS is smaller, and vice versa. Hence, if the BS transmits a pilot signal with constant power, then the received power of the pilot signal at the MS carries the information of the distance between the MS and the BS. Since the probability that the MS will be active in a particular cell at a future moment is a function of the current distance between the MS and the nearby BS, the probability can be estimated based on the realtime measurement of the received pilot signal power at the MS from the BS. Furthermore, the probability depends on the movement pattern of MS. Although the movement patterns of MSs are random in nature, the movement of an individual MS follows a relatively smooth trajectory most of the time. That is, the location of an MS at a future moment depends on its locations at the current moment and previous moments. As a result, it is possible to predict the mobility information based on the current and previous measurement data. As discussed proceeding, due to the random phenomenon of shadowing and inaccurate measurements caused by MAI, it is very difficult to accurately describe the relationship between the strength of the received pilot signal and the mobility information by a mathematical expression. To overcome the difficulty and adaptive fuzzy inference prediction system consisting of a fuzzy inference system and a recursive least square (RLS) predictor is employed to estimate and predict the MS mobility information. the functional block diagram of the fuzzy mobility information estimator. It consists of the fuzzy effective bandwidth estimator, and an RLS predictor. The fuzzy inference system takes the measurements of the received pilot signal power levels at the MS (P0, and Pls l = 1,2, ... ,6) as the input, and estimates the probability that the MS will be active in cell l at time tn, pl>n (l = 1,2, ...,6) based on the measured pilot signal strengths at time tn. 3.3 NETWORK RESOURCE ESTIMATOR In the network resource estimator, a fuzzy residual bandwidth estimator is explored to estimate the residual bandwidth of each cell according to the Eb/Io measurement of one specific traffic class in the cell. Each BS collects the Eb/Io measurement of this traffic class periodically, which is denoted by (Eb/Io)s. There exists a residual bandwidth if (Eb/Io)s> ( Eb/Io)r and the residual bandwidth increases with the value of [(Eb/Io)s( Eb/Io)r]. The relationship between the Eb/Io measurements and the residual bandwidth of the BS can be mapped to the fuzzy inference rules in the fuzzy residual bandwidth estimator with carefully define membership functions of (Eb/Io)s and Bs. With the information of The residual bandwidths at the BSs, The intarcell and intercell effective bandwidths of the new call, and The reserved bandwidths for the existing calls at nonhome BSs, the available bandwidths that the home BS, and its neighboring BSs are able to supply to the new call can be obtained. 3.4 Fuzzy Call Admission Processor When a new class j call requests for a connection in cell 0, the MSC first calculates its intracell bandwidth Bo, intercell bandwidth Bl, and residual bandwidth of the home cell and each neighboring cell Bs,l, and then checks the available bandwidth in both its home cell and neighboring cells, Ba,l, l=0,1,...6. If Ba>0>B0 and Ba>l>Bl for all le{1,2,....6}, the new call will be accepted; otherwise, it will be rejected. If the request come from an MS in service in a neighboring cell, the handoff call will be admitted as long as Bs o>Bo and Bsl>Bl for all le{1,2....6}. Due to the random nature of user movement and transmission rate of a multimedia traffic flow, reserving resources in the neighboring cell to keep a low handoff call dropping probability results in the waste of the resources from time. This issue cannot be solved by the above fixed rule CAC scheme since it does not adapt to the traffic load dynamics. To better make us of the system resources, a fuzzy CAC The processor (FCACP) can be employed to process the new call requests, instead of using the fixed rule control. The FCACP takes the effective bandwidths of a new call, the available bandwidth information that its home cell and its neighboring cells can supply, and the outage probability requirements as the input linguistic variables, and uses a fuzzy inference system to determine whether or not to accept the new call. The structure of the FCACP is similar to the fuzzy effective bandwidth estimator. The term sets for effective bandwidth Bl, available bandwidth information in its home cell Ba,0, available bandwidths information in its neighboring cells Ba>l(l=1,2,...6), and QoS indicator Q are defined U(Bl)={S(small), SM(small to medium), M(medium), ML(medium to large), L(large)}, U(Ba>0)={NE(not enough), E(enough),ME(more than enough)},u(Ba,l)={NE,E,ME}, AND U(Q)={S(satisfied),NS(not satisfied)}. The term set for output linguistic variables D of the FCACP is U(D)= {A(accept),WA(weakly accept),WR(weakly reject),R(reject)}. The output of the FCACP is a crisp value D s [0, 1].A call request will be accepted when the value of D is larger than a preset threshold value D0. 4 NUMERICAL RESULTS voice call arriving rate Fig. 5. Outage probability of voice calls versus the voice call arrival rate 0.6 NCBP of FCAC NCBP ofRPCAC NCBP of NPCAC HCDP of RPCAC HCDP ofNPCAC 0.5 0.4 0 CI 0.5 2.5 Voice call arriving rate Fig. 6. The new call blocking probability and handoff call droping probablity for voice calls 0.5 2 2.5 1 1.5 voice call arriving rate TABLE 1 Fuzzy Inference Rules of the Intercell Effective Bandwidth Estimator for a Voice Call P0 P1 B1 Degree P0 P1 B1 Degree M S ES 0.481 EL ES ES 0.956 M SM L 0.442 EL VS ES 1.000 M M L 0.250 EL S ES 0.854 TABLE 2 Fuzzy Inference Rules for Residual Bandwidth Estimator (Eb/Io)s Bs Degree (Eb/Io)s Bs Degree S3 B3 1.0000 S11 B11 0.1128 S1 0.1043 0.1094 S6 B0 0.0940 S33 B33 0.2650 TABLE 3 Fuzzy Inference Rules for the CAC Decision, where l = 1,2,3,.. .,6 Bl Ba,0 Ba,l Q D S E NE S WA M E NE S WR L E NE S R the outage probability of the system with respect to different call arrival rates for FCAC, RPCAC, and NPCAC, respectively. It can be seen that the FCAC and the NPCAC can always satisfy the outage probability requirements even under the heavy traffic load conditions, while the RPCAC cannot meet the requirement during heavy traffic load periods. Fig .6 shows the new call blocking probability (NCBP) and handoff call dropping probability (HCDP) for voice versus corresponding call arrival rate. It can be seen that the NPCAC has the highest NCBP and lowest HCDP, which is desirable. Fig .7 shows the mean admitted number of voice calls in the system as a function of the call arrival rate. The system resource utilization efficiency increases with the average number of active users. It can be seen that, while the mean number of active users for FCAC and RPCAC are similar, the mean number of active users for NPCAC is lower than those of the FCAC and RPCAC. For RPCAC, the call admission decision for one particular cell is made based on the total received power at that cell, without considering the power level in any other cell. When a call is accepted into a cell, it will not only increase the received interference levels in the neighboring cell. If the traffic load in the neighboring cell is not taken into account, the QoS requirements of the MSs in the target cell and its neighboring cells may not be satisfied. Even though the RPCAC can achieve a lower new call blocking probability and higher system resource utilization, the outage probability of the whole system is not satisfied during heavy traffic load periods. In addition the RPCAC does not differentiate new call and handoff call, resulting in the similar new call blocking and handoff call dropping probabilities. For the NPCAC, The MS'S intercell effective bandwidth is taken into account in the call admission decision procedure, but the product of a constant and its equivalent bandwidth represents the intercell effective bandwidth. This is not true in a CDMA cellular system since the intercell effective bandwidth of an MS is dependent on the MS location and other factors such as wireless channel conditions. It is not easy to choose a suitable value for the constant: If a large value is chosen, the resources in other cell may be wasted. If a small value is chosen, the resources in other cells may not be enough if the call is accepted in its home cell. The advantage of the NPCAC is that the QoS requirement of outage probability is satisfied with properly chosen parameters. The disadvantage is that it achieves a high new call blocking probability and low system resource utilization. For the FCAC, the intercell effective bandwidth of an MS is calculated according to the real time signal strength measurements, the handoff calls are given higher priority by adaptively reserving a proper amount resources according to the mobility information, and the new call admission decision is made by taking into account the new call's characteristics, system resource availability situation, and QoS provisioning of the system, which dynamically adapts to the current system status. As a result, the FCAC can achieve efficient resource utilization and, at the same time, meet the QoS requirement of outage probability. Table 1,2,3 are use as fuzzy inference rules of the intercell effective bandwidth estimator for a voice call, residual bandwidth estimator and CAC dicision form the home station and form the near by station. This tabulation is very useful in making the decisions during uncertainty and also very quick. 5 CONCLUSION A fuzzy CAC scheme for wideband CSMA cellular communications has been proposed to meet the challenges in CAC due to user mobility, limited radio spectrum, heterogeneous and dynamic nature of multimedia traffic, and QoS constraints. The fuzzy approach can overcome and avoid the requirements of complex mathematical relations among various design parameters. The resource requirement of each call is presented in terms of intracell and intercell effective bandwidths. The user mobility information is estimated and predicted based on the measurements of the pilot signal power levels received at the MS. The CAC decision is based on the resource availability where handoff calls are given with new calls via resource reservation. Simulation results show that the fuzzy CAC scheme can achieve QoS satisfaction in terms of the outage probability, and achieve lower new call blocking probability, lower handoff call dropping probability, and higher resource utilization efficiency, when compared to the previously proposed RPCAC and NPCAC schemes. 6 REFERENCES [1] C.Y.Huang and R.D.Yates, "Call Admission in power controlled CDMA system," Proc. IEEE Vehicular Technology Conf. [2] S.Sun and W.A.Krzymien, "Call Admission Policies and Capacity Analysis of a MultiService CDMA Personal Communication System with Continuous and Discontinuous Transmission," Proc. IEEE Vehicular Technology Conf. [3] A.Ghaffari, H.R. Mirkhani, and M. Najafi, "Stability Investigation of a Class of Fuzzy Logic Control Systems," Proc. IEEE Int'l Conf. [4] D.A. Levine, I.F. Akyildiz, and M. Naghshineh, " A Resource Estimation and Call Admission Algorithm for Wireless Multimedia Networks Using the Shadow Cluster Concept," IEEE/ACM Trans. [5] X. Shen, J.W. Mark, and J.Ye, "User Mobility Profile Prediction: An Adaptive Fuzzy Inference Approach," ACM/Wireless Networks, vol. 6, pp. 363  374, 2000. [6] M.M. Zonoozi and P. Dassanayake, "User Mobility Modeling and Characterization of Mobility Pattern," IEEE j, Selected Areas Comm., vol. 15, pp. 12391252, 1997. CONTENTS 1. INTRODUCTION 2. SYSTEM MODEL 3. DESIGN OF THE FUZZY CAC SCHEME 3.1. FUZZY EFFECTIVE BANDWIDTH ESTIMATOR 3.2. FUZZY MOBILITY INFORMATION ESTIMATOR AND BANDWIDTH RESERVATION 3.3. NETWORK RESOURCE ESTIMATOR 3.4. FUZZY CALL ADMISSION PROCESSOR 4. NUMERICAL RESULTS 5. CONCLUSION 6. REFERENCES Use Search at http://topicideas.net/search.php wisely To Get Information About Project Topic and Seminar ideas with report/source code along pdf and ppt presenaion



ruchi87singhal Active In SP Posts: 1 Joined: Jan 2012 
25012012, 12:49 PM
i need only call admission control in cdma not fussy logic.



