AN ADAPTIVE BANDWIDTH ALLOCATION FOR WIRELESS COMMUNICATION IN OFDM NETWORKS

Transcription

1 AN ADAPTIVE BANDWIDTH ALLOCATION FOR WIRELESS COMMUNICATION IN OFDM NETWORKS K. ARUNA KUMARI, JYOTHIRMAI NALLAGATLA Sr. Assistant Professor, Department of ECE, PVP Siddhartha Institute of technology, Vijayawada. PG Student, Department of ECE, PVP Siddhartha Institute of technology, Vijayawada 677 ABSTRACT In wireless/mobile networks various kinds of encoding schemes were used for transmission of data over a bandwidth. The desired quality and generated traffic varies with the requirement with this bandwidth. From the designing point of view these requirements demands for an alternative resource planning, especially for bandwidth is a scare resource, the system may need to block incoming user if all of the bandwidth has been used to provide highest quality of service to existing users. However this bandwidth resource planning may be unacceptable for larger application. A degradable approach to multiple users can be made on bandwidth allocation to reduce the blocking probability without degrading the quality of service to existing users. Keywords: Bandwidth allocation, Mobile Networks, Quality of Service, 1. INTRODUCTION Cellular wireless technology today has become the prevalent technology for wireless networking. Not only mobile phones but also other types of devices such as laptops and Personal Digital Assistant(PDA) can connect to Internet via cellular infrastructure These mobile devices are often capable of running multimedia applications(e.g., video, images).therefore cellular networks need to provide quality of service(qos)guarantee to different types of data traffic in mobile environment. A call admission control(cac)scheme aims at maintaining the delivered QOS to the different calls(or users) at the target level by limiting the number of ongoing calls in the system. One major challenge in designing a CAC arises due to the fact that the cellular network has to service two major types of calls: new calls and handoff calls. The QOs performances related to these two types of calls are generally measured by new call blocking probability and handoff call dropping probability. In general, users are more sensitive to dropping of an ongoing and handed over call than blocking a new call. Therefore, a CAC scheme needs to prioritize handoff calls over new calls by minimizing handoff-dropping probability. Again, bandwidth adaptation and scheduling are necessary mechanisms for achieving high utilization of the wireless resources(e.g., channel bandwidth)while satisfying the QOs requirements for the users. These two techniques are closely related to call admission control, and in fact these three mechanism jointly determine the call-level and the packet-level QOS for the different types of traffic in the cellular wireless air interface. For example, upon arrival of a new call or handoff call, bandwidth adaptation can be performed to degrade the channel allocations for some calls (still maintaining the QOS requirements)so that the new call can be admitted. Scheduling mechanisms impact the packet-level system dynamics(e.g., queuing behaviour),and therefore, packet-level QOS. The packet-level dynamics can be exploited for designing efficient call admission control methods. The call admission control(cac) and the adaptive channel adaptation (ACA) mechanisms are generally treated as the network layer(above layer-2) functionalities in the wireless transmission protocol stack(figure 1.1).The scheduling and the adaptive modulation and modulation and coding(amc) are layer-2 and layer-1(i.e., physical functionalities, respectively. Fig.1. MIMO-OFDM Relay network 1.1QOS IN WIRELESS COMMUNICATION Guaranteeing the QoS requirements is a challenging task with wireless communication one of the key elements in providing QoS in an effective resources allocation policy, which not only ensures meeting QoS of newly arriving calls, if accepted but also not deteriorating the existing on-going services. These enhancements will enable a better mobile user experience and will make more efficient use of the wireless channel. As the performance of system with given physical resources(e.g., the available bandwidth of radio spectrum)depends

2 678 heavily on resource management schemes including multiple access techniques, the call admission control policies and the congestion control schemes, to make efficient resource management schemes have to be devised. Many real-time applications can use different encoding schemes according to their desired quality and generate traffic with different bandwidth requirements. For example, generic video telephony may require more than 40KBPS,but low-motion video telephony requiring about 25 KBPS may be acceptable. From the standpoint of a system administrator, this property provides an alternative for resource planning, especially for bandwidth allocation in wires networks. In wireless networks where the bandwidth is a scarce resource, the system may need to block incoming users if all of the bandwidth has been used up to provide the highest QoS to existing users. However, if these users can be degraded to a lower QoS level, if it is possible to reduce the blocking probability with out degrading the QoS of existing users to an unacceptable level. Various approaches and algorithms adopting this idea have been proposed. A graceful degradation mechanism is proposed in to increase bandwidth utilization by adaptively adjusting bandwidth allocation according to user-specified loss profies. Thus, a system could free some could free some bandwidth foe new users by lowering the QoS levels of existing users. 2. SYSTEM MODEL AND PROBLEM STATEMENT We consider the wireless cellular networks in which bandwidth of ongoing calls can be adjusted adaptively according to the states of the network. The total bandwidth in a cell is constant and is denoted by c. We assume that the bandwidth for a{call is chosen from a set of discrete values b={b1,b2..bm} where bi<bi+1 for i=1..n- 1.The minimum and the maximum amount of bandwidth allocated to a call is b1 and bn respectively. The bandwidth requirement for a mobile is denoted by breq(bn>=breq>=b1).if the amount bandwidth allocated to a mobile is less than breq, then degradation in call quality occurs, however, the call is not dropped. The desired quality and generated traffic varies with requirements with this bandwidth. If the users can be degraded to a lower QoS level, it is possible to reduce the blocking probability without degrading this bandwidth resource e QoS of existing users to an Unacceptable. Various approaches and algorithms adopting this idea have been proposed a grace full degradation mechanism is proposed to increase bandwidth allocation by adjusting bandwidth utilization and attempted to achieve fairness with a generic algorithm. In terms of bandwidth utilization or service provider s revenue can be improved significantly but allowing QoS degradation. 3. PROPOSED METHOD 3.1 QOS IN WIRELESS COMMUNICATION In this paper, we exploit the adaptive bandwidth allocation for QOS provisioning in wireless/mobile networks. An analytical model for a wireless/mobile network with multilevel degradable QOS is provided. This model includes two very important QOS merits degradation on ratio and upgrade/degrade frequency-both of which are necessary for QOS provisioning. Moreover, our analytical model includes user mobility to enable the study of its impact on user perceived QOS. Our work not only provides an analytical frame work for predictive or adaptive bandwidth allocation algorithms, but also helps decide the operation region based on some desired criteria. It should be noted that our scheme can be applied to various wireless architectures. For a orthogonal frequency division multiple access(ofdm)system, the multicode OFDM can be used for service degrade/upgrade for a Time Division Multiple Access(TDMA)system(e.g., Bluetooth),service degrade/upgrade can be achieved by an adequate assignment of timeslots(i.e., polling policy)users so as to improve the overall system performance. For example, we may be able to achieve high bandwidth utilization and maintain a small blocking and/or forced termination probability. In a system with degradable QOS, a user may receive different levels of QOS during the entire duration of his connection, depending on the loads of cells he traverses. Even if a user receives the highest level of QOS when he is admitted to a cell, the QOS may still be degraded when some other base stations on his path decide to degrade to degrade his QOS in order to accept more users. From the users perspectives, this may raise two important questions:1)how long does his connection stay at each individual QOS level?2)hoe often does the received QOS switch between these levels? Even though these two questions are interrelated, the first question does not necessarily imply the second, or vice versa. Therefore, two performance metrics associated with these questions, degradation ratio and upgrade/degrade. 3.2 QUALITY OF SERVICE Since there are K different QOS levels, we can define the system state,ns as ns=(n1,n2..n1-)_where ni is the number of users in the ith QOS level in acell.such a system can be modelled as a Markov chain and the transition probabilities can be obtained accordingly. In our model, the transition probabilities depend on the admission control(i.e., the value of our model, the transition probabilities depend on the admission control(i.e., the value of our model, the transition probabilities depend on the admission control(i.e., the value of Nthresh)policy shortage of bandwidth, allocating only Wmin to an incoming user minimizes the need to degrade the QOS levels of existing users and,hence,the result in smaller transmission delay and higher throughput On the other hand, fairness is an important issue when e consider bandwidth reallocation in a system supporting multilevel QOS. We may evenly

3 679 degrade the QOS of existing users to accommodate a new users or to degrade as few users as possible so as to minimize the change of current bandwidth constellation. Therefore, one can make a trade off between fairness and transmission time delay. In wireless mobile multimedia networks adaptive bandwidth allocation (ABA)is necessary to maximize the utilization of radio channels while keeping the quality-of-service(qos) of a multimedia call at the acceptable level.aba can minimize the number of blocked new calls and the number of dropped handoff calls by adjusting the allocated bandwidth of ongoing calls and allowing the incoming calls to be serviced with out degrading QOS of the ongoing calls below the acceptable level BANDWIDTH ADAPTION ALGORTIHM We consider a fairness-based bandwidth adaption algorithm, which works in such a way that the allocated bandwidth to the ongoing calls will not differ from each other by more than one step. The bandwidth of an ongoing call is also allowed to be degraded below bandwidth requirement breq to minimize new call blocking and handoff call dropping probabilities. Let Walk and balk denote the expected bandwidth for an incoming call and the bandwidth vector of ongoing calls, respectively. When a call arrives, the cell performs admission control by checking whether the total number of ongoing calls is less than the threshold t. If this condition is satisfied or if the incoming call is an and off call, the cell tries to allocate maximum bandwidth to the incoming call; otherwise the incoming new call is blocked. However, if the available bandwidth is not enough to allocate maximum bandwidth to an incoming call, the adaptation algorithm is invoked. The adaptation algorithm will randomly select an ongoing call with the current maximum bandwidth (i.e., max (balk))and de-grade allocated bandwidth of that call one step. At this point,expected bandwidth for incoming call increases one step. This operation is iteratively performed until the expected bandwidth for an incoming call is equal. For call thinning scheme, line 1 of this algorithm would be changed to admit the user If((incoming is a new call)and (number of ongoing calls <k) { If(available bandwidth>or=bmax) Then assign bmax to incoming call Else { ballocated=0 for(t=1,t<n,t++) While(ballocated<bmin and nt>0) { Randomly degrade one of nt connections by amount of bdegrade Bdegrade=min(bmin,br-bmin) Ballocated=ballocated+bdegrade } } } Else reject incoming call QOS MEASURES We consider three QOS measures, namely, degradation ratio, through put, degradation bandwidth. These QOS measures can be calculated from total bandwidth occupies in the channel. In wireless channel bandwidth is scarce resource. Channel can accommodate only few users. If a new user wants to enter in the network due to lack of available bandwidth the request cannot be accepted. In this model by degrading the QOS of existing users a new request can be accepted. We are measuring these quality of service measures based on the above algorithm. Degradation Ratio(DR).The fraction of time a user receives degraded QOS since we are considering multi level QOS systems DR is defined as If a user receives level-i QOS for Ti seconds. Throughput: It gives the amount of number of packets arrived at receiver. It measures the efficiency of the system. Throughput=(number of bits received/number of bits send)*100 Degraded bandwidth (Db): It measure the amount of bandwidth degraded from existing users. If number of users enters in the channel are increased then bandwidth degraded is increase. The degradation is stops for a user if he reaches the minimum bandwidth (b1).

4 RESULT ANALYSIS Observation 1 Here We observe the graphical analysis and the results are verified with benchmarks. Fig 2. Comparison of Conventional and proposed methods Observation 2 Here the Fig.3. Shows that use the Number of relays in a network to share the bandwidth in anetwork. Fig 3.Effect of number of relays in Conventional and Proposed method Observation 3 Here the Fig.4.Shows that the minimum energy distribution in a OFDM Network when sharing the bandwidth to the users with reference to relays. Fig 4.Energy Distribution in channel Observation 4 Here the Fig.5. Shows that the OFDM Network effects from different types of fading and non fading condition. Fig 5. Fading and non fading in OFDM network

5 681 Observation 5 Here Fig.6.shows that comparison analysis of fixed bandwidth allocation scheme and adaptive bandwidth allocation scheme with reference of Throughput. Fig 6. Throughput plot for communication system with user arrival/removal Observation 6 Here the Fig.7. shows that comparison of degradation ration in Fixed bandwidth allocation scheme and Adaptive bandwidth allocation scheme. Fig 6. Degradation ratio plot for communication system with user arrival/removal Observation 7 Here the Fig.7. shows that comparison of degraded bandwidth in Fixed bandwidth allocation scheme and Adaptive bandwidth allocation scheme. Fig 7. Degraded bandwidth with 1st group of communication system Observation 8 Here the Fig.8. shows that comparison of propagation delay in Fixed bandwidth allocation scheme and Adaptive bandwidth allocation scheme. Fig.8. Propagation delay for the communication system.

6 682 CONCLUSIONS In this paper, we derived an analytical model for a wireless network which uses adaptive bandwidth allocation to provide users multilevel QOS. Four performance metrics Throughput, transmission time delay, degraded bandwidth, degradation ratio are observed. The performance plots obtained gives that with increase in load with respect to time the through put level falls down because of increase in compression level which could be controlled by adaptive band width allocation method. With increase in demand for transmitting data over a constraint bandwidth new algorithm is to be implemented to overcome the resource constraints observed in wireless communication although the various techniques where proposed in last for proper resource allocation they get constrained once the cell capacity is reached. Hence they require a advanced algorithm for proper utilization and scheduling of resources to handle more number of user with the constrained band width. In this project work an approach is made to overcome the resource constraint by degrading the QOS provided to each user the system is implemented following OFDM] Architecture with cellular communications where each cell constitute of users communicating simultaneously. The preformance is evaluated over OFDM architecture by adding or removing different group of users to evaluate the algorithm efficiency. The metrics used to evaluate the performance are throughput, propagation delay, degradation ratio allocated bandwidth degradation is observed to outperform the existing fixed bandwidth allocation technique with more number of users it is seen that the delay for proposed adaptive bandwidth allocation method get reduced by 40% compared to fixed bandwidth allocation method. REFERENCES [1] S. Sen, J. Jawanda, K. Basu, and S. Das, Quality-of Service Degradation Strategies in Multimedia Wireless Network, Proc.IEEE Vehicular Technology Conf., vol. 3, pp , May [2] S. Singh, Quality of Service Guarantees in Mobile Computing, Computer Comm., no. 19, pp , [3] M.R. Sherif, I.W. Habib, M.N. Nagshineh, and P.K. Kermani, Adaptive Allocation of Resources and Call Admission Control for Wireless ATM Using Generic Algorithm, IEEE J. Selected Areas in Comm., vol. 18, no. 2, pp , Feb [4] T. Kwon, Y. Choi, C. Bisdikian, and M. Naghshineh, Call Admission Control for Adaptive Multimedia in Wireless/Mobile Network, Proc. First ACM Int l Workshop Wireless Mobile Multimedia, pp , Oct [5] S. Choi and K.G. Shin, Location/Mobility-Dependent Bandwidth Adaptation in QoS-Sensitive Cellular Networks, Proc. IEEE Vehicular Technology Conf., vol. 3, pp , [6] Y.B Lin, S. Mohan, and A. Noerpel, Queuing Priority Channel assignment Strategy for PCS Handoff and Initial Access, IEEE Trans. Vehicular Technology, vol. 43, no. 3, pp , Aug [7] R. Ramjee, R. Nagarajan, and D. Towsley, On Optimal Call Admission Control in Cellular Networks, Proc. IEEE INFOCOM 96, vol. 1, pp , [8] K. Mitchell and K. Sohraby, An Analysis of the Effects of mobility on Bandwidth Allocation Strategies in Multi-Class Cellular Wireless Networks, Proc. IEEE INFOCOM 01, vol. 2,pp , [9] A. Sutoving and J.M. Peha, Novel Heuristic for Call Admission Control in Cellular Systems, Proc. IEEE Int l Conf. Universal Personal Comm., vol. 1, pp , 1997.