An Enhanced Scheduling Scheme for QoS in Mobile WiMAX Networks

(1)

ISSN (PRINT) :2320 – 8945, Volume -1, Issue -2, 2013

49

An Enhanced Scheduling Scheme for QoS in Mobile WiMAX Networks

Jithin Raj & D. David Neels Pon Kumar

Department of ECE, Einstien College of Engineering, Tirunelveli, Tamilnadu E-mail : [email protected], [email protected]

Abstract – The Worldwide Interoperability for Microwave Access (WiMAX) standard, that is, the IEEE 802.16e standard supports point-to-multipoint (PMP). Scheduling in WiMAX become one of the most challenging issues, since it is responsible for distributing available resources among users to meet the QoS criteria such as delay, delay jitter, fairness and throughput requirements. However, no specific scheduling mechanism has been defined to provide QoS through five kinds of service class differentiation. In this paper we propose a novel Priority based Scheduling scheme to support various services by considering the QoS constraints of each class. The simulation results show the throughput, delay and jitter with standing nodes and nodes at different speeds. From these results it is concluded that the proposed scheme provides QoS support for each class efficiently for Mobile WiMAX

Keywords – Delay jitter, Fairness index, QoS, PMP, Mobile WiMAX

I. INTRODUCTION

WiMAX is an emerging broadband wireless last mile technology to provide higher speeds of upto 70 Mbps for longer distances say 30miles of radius for distributing broadband wireless data over wide geographic areas. WiMAX is used due to extend broadband capabilities, supporting very high bandwidth, providing wide area coverage. WiMAX offers a rich set of features with a great deal of flexibility in terms of deployment options and potential serviceofferings. It can provide two forms of wireless service such as Line- of-Sight (LoS) service and Non-Line-of-Sight (NLoS) service. A WiMAX system consists of two partsa WiMAX Base Station (BS) and a WiMAX receiver. Fig 1 explains the basic diagram of WiMAX technology.The IEEE 802.16e standard divides all services into five different classes namely Unsolicited grant service (UGS), Real-time polling service(rtPS), Enhanced Real-time polling service (ertPS), Non real- time polling service (nrtPS) and Best effort (BE). The

application of these service classes are shown in the Table.1. Since Mobile WiMAX system provides various real-time and non-real-time services as mentioned above, appropriate resource allocation schemes are required to support QoS (Quality of Service) of each service. A key feature of the WiMAX technology is that it is a connection oriented technology, which provides a strong support for QoS management

Fig 1: Operational principles of WiMAX technology.

Table.1 Application, band width, latency and jitter of each service classes

Class Application Bandwidth Latency Jitter 1 Interactive

Gaming 50kbps < 25 msec

N/A

2

Voice Telephone (VoIP) Video

conference

32064kbps 160 msec < 50 msec 3 Streaming

Media

5Kbps – 2

Mbps N/A < 100 msec

4 Instant 10 Kbps – N/A N/A

(2)

50 Messaging

Web Browsing

2 Mbps 5 Media Content

Download >2 Mbps N/A N/A II. LITERATURE SURVEY

In [1] the author proposed Modified Priority Algorithm to improve the throughput and fairness of the BE. The other service classes are not mentioned in this paper. In [2] various scheduling algorithms for the various traffics in mobile WiMAX is discussed.These Scheduling algorithms do not clearly consider all the required QoS parameters. In [3] surveys of the WiMAX scheduling algorithms are done and discussed the key issues and design factors in QoS scheduling. The authors of[4] give an idea about Exponential Rule Based algorithms. The advantages and drawbacks of major scheduling algorithms such as Round Robin, Weighted Round Robin(WRR), Deficit Round Robin (DRR) and Modified Deficit Round-Robin (MDRR)algorithms are summarized and compared in [5].The advantages of Round Robin polling in high traffic scenarios are explained in [6]. In [7] authors proposed a systematic framework of Mobile WiMAX QoS scheduling based on OFDMA radio resource management. The study showed the correct selection of scheduling algorithm is critical to support combinations of real-time and non- real-time traffic flows. The authors in [8] presented the implementation methodology of an ns-2 based WiMAX simulation model in which the QoS scheduling is achieved by traffic class prioritization implementation.

The simulation results in[9] show that properly chosen scheduling algorithm can provide high service standards to support the QoS required for different type of traffics and users. In [10] authors propose a packet scheduling scheme to support multiple services efficiently by considering the QoS characteristics of each class by selecting a service class first after considering characteristics of each class. Several simulation models have been proposed to support Mobile WiMAX simulation, such as QualNet [11], Opnet [12], and NS- 2[13]. The research in [14] proposed a systematic framework of Mobile WiMAX QoS scheduling based on OFDMA radio resource management. The study showed the correct selection of scheduling algorithm is critical to support combinations of real-time and non- real-time traffic flows. Similar research studies regarding WiMAX QoS scheduling can be found in [15]-[16]. On the other hand, only a few simulation studies have been published or reported in literature, largely ignoring the implementation and evaluation of various scheduling algorithms to support versatile Mobile WiMAX QoS simulations.

III. PROPOSED SCHEDULING ALGORITHM

The main objective of this work is to provide an implementation of the IEEE 802.16(e) standard using dynamic fuzzy based priority scheduler. Here scheduling is done in two stages. The requests are classified in terms of priority using fuzzy logic and the requests are scheduled using Neural Networks.

A) Classification of priority using fuzzy logic

Basically the fuzzy system consists of four blocks, namely, fuzzifier, defuzzifier, inference engine, and fuzzy knowledge base.The first step is to take the inputs and determine the degree to which they belong to each of the appropriate fuzzy sets via membership functions.

A fuzzy set A in the universe of discourse U is a set of ordered pairs {(x1, μA(x1)), (x2, μA(x2)), . . . , (xn, μA(xn))}, where μA : U → [0, 1] is the membership function of the fuzzy set A and μA(xi) indicates the membership degree of x_i in the fuzzy set A.If a fuzzy system has n inputs and a single output, its fuzzy rules Rj can be of the following general format. (Rj) If X₁ is A₁j , X₂ is A₂ j , X₃ is A₃ j , . . ., and X_m is A_mj, then Y is Bj. The variables X_i{i = 1, 2, 3, . . ., n} appearing in the antecedent part of the fuzzy rules Rj are called the input linguistic variables, the variable Y in the consequent part of the fuzzy rules R_j is called the output linguistic variable. The fuzzy sets A_ij are called the input fuzzy sets of the input linguistic variable Xi and the fuzzy sets Bj are called the output fuzzy sets of the output linguistic variable Y of the fuzzy rules Rj. The final desired output for each variable is generally a single number. By Centroid method of defuzzification, the output η is calculated using the formula

(1)

where y is the centre point of each of the output membership function in the output fuzzy set B_j and μ^outputx₁··………..xn(y) is the strength of the output membership function. In the proposed fuzzy scheduler we use two different stages namely the Primary Scheduler, FS1 and the Dynamic Scheduler, FS2. This proposed scheduler is named as Dynamic Fuzzy based Priority Scheduler (DFPS). In the proposed Primary Scheduler four inputs namely, Expiry time (E), Waiting time (W), Queue length (Q), Packet size (P) and one output, Priority index are used as shown in Fig 2.According to fuzzy rule „if packet size is low and queue length is low, then priority index is low‟.

Similarly “If packet size is high and queue length is high, then priority index is high”. The priority index, if high, indicates that the packets are associated with the highest priority and will be scheduled immediately. If the index is low, then packets are with the lowest

(3)

51 priority and will be scheduled only after high priority packets are scheduled. This is illustrated in Table.2

Fig 2: Dynamic Fuzzy scheduler

Table.2 Fuzzy Rule Base (a) Expiry Time Vs Waiting Time(b) Packet size Vs Queue length (c) (a) Vs (b)

Expiry Time

Waiting Time

L M H

L L L M

M L M M

H L M H

(a) Packet

Size

Queue Length

L M H

L L M M

M L M H

H M M H

(b)

(a) (b)

L M H

L VL L M

M L M H

H M H VH

As there are five different types of classes the priority levels are set to five different levels starting from Very High (VH), High (H), Medium (M), Low (L) and Very Low (VL). This rule is used to satisfy the QoS requirements of WiMAX

Algorithm 1:- Setting up the priority

Input: Expiry Time, Waiting Time, Packet Size, Queue Length, Type of Service, Mobility

Assumptions: The above said first four inputs are in 3 scales (H, M and L) and last input is in 5 scales (VH, H, M, L and VL).

Output: Highest Priority Request For i=1 to n do

a. Compare Expiry Time and Waiting Time which arrives at Priority index „a‟.

b. Compare Packet Size and Queue Length which arrives at priority index „b‟.

c. Compare „a‟ and „b‟ (3 scale output) which arrives at Intermediate Priority1 (5scale output).

d. Compare Intermediate Priority with Type of Service that arrives at Intermediate priority2

e. Compare Intermediate Priority2 with mobility factor which arrives at Final Priority Index.

f. Arrange the requests in descending order.

IV. SIMULATION AND PERFORMANCE ANALYSIS Here WiMAX environment is created using Network Simulator (NS2).Scheduling is carried out by using a scheduler unit. Initially the data from base station goes to scheduler. The scheduler calculate priority index according to algorithm explained above and the allows data to move subscriber station which have appropriate priority. The simulation was carried out for fixed nodes and mobile nodes up to 120 km/hr.

QoS parameters such as throughput, delay and jitter were calculated for each simulation .The simulation parameters and results are discussed below.

A) Simulation Parameters 1) Throughput (T_h)

Throughput is measure of number of packets successfully delivered in a network. It is measured in terms of packets/second. The value of throughput should be high or else it affects every service class defined in Wimax. Throughput can be calculated by the equation

(1) 2) Delay

Delay or latency could be defined as the time taken by the packets to reach from source to destination. End to end delay could be measured as the difference of

Final PriorityIndex

Type of Service Priority Index Mobility

Packet size Queue Length

Waiting time Expiry time

FS-1

^FS-2

(4)

52 Packet arrival and packet start time. Equation 2 shows the calculation of average end to end delay.

(2) 3) Jitter

Jitter could be termed as the variation in delay or packet delay variation. The value of jitter is calculated from the end to end delay. Measuring jitter is critical element to determining the performance of network and the QoS the network offers. It is the variation in the time between packets arriving. Jitter is commonly used as an indicator of consistency and stability of a network.

Equation 3 shows how to calculate jitter.

(3) B) Simulation Results

Here consider fixed and mobile nodes. For each type of service throughput delay and jitter is measured and results are given below.

1) Fixed Nodes

In this nodes are fixed and simulations are carried out. Figure 3,4 and 5 shows simulation results.

Fig 3: Throughput for fixed nodes

Fig 4: Delay for fixed nodes

Fig 5: Jitter for fixed nodes

From above graphs it is clear that throughput has a larger value for high priority and after that the throughput decreases. The delay and jitter varies inversely with priority.ie, for high priority services the delay and jitter has minimum value. As priority goes to low that parameters increases with simulation time. The average values of throughput, delay and jitter is tabulated below.

Table 3:Average values of throughput delay and jitter at non mobile condition.

Types of

service UGS ertPS rtPS nrtPS BE Throughput

(MB) 2.4 2.37 1.7 1.58 1.37 Delay

(second) 0.05 0.18 0.6 0.4 2.5 Jitter

(second) 0.03 0.08 0.15 0.12 0.17 2) Mobile nodes

Here simulations are carried out for different mobility conditions such as normal speed and high speeds

a) Normal speed

In this simulation normal speed is taken as 30 km/hr and three parameters are measured. Simulation results are given below.

(5)

53

Fig 6: Throughput at 30 km/hr

Fig 7: Delay at 30 km/hr

Fig 8:Jitter at 30 km/hr

Table 4: Average values of throughput, delay and jitter at 30 km/hr

Types of

(MB) 1.5 1.2 1.19 1.4 1.32

Delay

(second) 0.05 0.15 0.47 0.26 2.2 Jitter

(second) 0.03 0.07 0.16 0.12 0.23 b) High speed

According to WiMAX forum high speed is taken as 120km/hr and simulate at this speed. The simulation results are given below.

Fig 9:Throughput at 120km/hr

Fig 10: Delay at 120 km/hr

(6)

54

Fig 11:Jitter at 120 km/hr

Table 5:Average values of throughput delay and jitter at 120 km/hr

Types of

(MB) 1.6 1.35 1.25 0.8 0.6

Delay

(second) 0.05 0.18 0.45 0.3 0.18 Jitter

(second) 0.05 0.089 0.15 0.063 0.4 Figure 6 to 11 gives the simulation results for mobile nodes at low and high speed and the results are tabulated in tables 4 and 5.Even different speeds the proposed algorithm provides approximately equal values of throughput, delay and jitter.

V. CONCLUSION

In this paper we have proposed fuzzy based scheduling algorithm for mobile WiMAX network.

Although there are many scheduling algorithms are used but some of the algorithm didn‟t fulfill QoS requirements of both mobile and fixed subscriber stations. From our simulation results it is clear that proposed algorithm is good for both mobile and non- mobile conditions.

VI. REFERENCES

[1] D.MohdAli,K. Dimyati, K.A Noordin and Azlinaidris

“Modified Priority Algorithm For Mobile WiMAX Uplink Scheduler” Research Journal Of Applied Sciences, Engineering And Technology 4(10):1310-

[2] Jun Shen, Na Yi, An Liu, Haige Xiang "Opportunistic Scheduling for Heterogeneous Services in Downlink OFDMA System," cmc, vol. 1, pp.260-264, 2009 WRI International Conference on Communications and Mobile Computing, 2009

[3] PreethiChandur, Karthik R.M and Krishna M.

Sivalingam, “ Performance Evaluation of Scheduling Algorithmsfor Mobile WiMAX Networks”

[4] C. Ravichandiran, Dr. C. Pethuru Raj, Dr. V.

Vaidhyanathan, “Analysis, Modification, and Implementation(AMI) of Scheduling Algorithm for the IEEE 802.116e (Mobile WiMAX)”, (IJCSIS) International Journal of Computer Science and Information Security,Vol. 7, No. 2, 2010

[5] W. Park, S. Cho, and S. Bahk, "Scheduler Design for Multiple Traffic Classes in OFDMA Networks,” in Proc. IEEE Int. Conf. Communications, Istanbul, Turkey, 2006, vol. 2, pp. 790-795.

[6] C. So-In,R. Jain,A. Tamimi,“Scheduling in IEEE 802.16e mobile WiMAX networks: key issues and survey,”IEEEJ.Sel.AreasCommn, vol. 27,no. 2,pp.156- 171,Feb. 2009.

[7] C. Cicconetti, I. Akyildiz, L. Lenzini, “FEBA: A bandwidth allocation algorithm for service differentiation in IEEE 802.16 mesh networks,”

IEEE/ACM Trans. Netw., vol. 17, no. 3, pp.884-897, Jun. 2009.

[8] HarpreetKaur, Gurpal Singh, “Implementation and Evaluation of Scheduling Algorithms in Point-to- Multipoint Mode in Wimax Networks” IJCST Vol. 2, Issue 3, September 2011

[9] Eunyoung Lee, Hyung-Kun Park, “Packet Scheduling Scheme for Multiple Services in Mobile WiMAX System” Second International Conference on Computer and Networks” IJCST Vol. 2, Issue 3, September 2011.

[10] Technology DOI 10.1109/ ICCNT.2010.56 IEEE Computer Society

[11] http://www.opnet.com.

[12] W. Wang, H. Sharif, M. Hempel, T. Zhou, P.

Mahasukhon, T. Ma, “Implementation and performance evaluation of a complete, accurate, versatile and realistic simulation model for mobile WiMAX in NS-2,” submitted to IEEE ICC, 2010.

[13] http://info.iet.unipi.it/~cng/ns2mesh80216/.

[14] Q. Ni, A. Vinel, Y. Xiao, A. Turlikov, T. Jiang,

“Investigation of bandwidth request mechanisms under point-to- multipoint mode of WiMAX networks,”

(7)

55

IEEE Commun. Mag., vol. 45, no. 5, pp.132-138, May.

2007.

[15] C. Cicconetti, I. Akyildiz, L. Lenzini, “FEBA: A bandwidth allocation algorithm for service differentiation in IEEE 802.16 mesh networks,”

IEEE/ACM Trans. Netw., vol. 17, no. 3, pp.884-897, Jun. 2009.

[16] J. Huang, V. Subramanian, R. Agrawal, R. Berry,

“Joint scheduling andresource allocation in uplink OFDM systems for broadband wirelessaccess networks,” IEEE J. Sel. Areas Commun., vol. 27, no.

2,pp.226-234, Feb. 2009.

[17] D. David NeelsPon Kumar, K. Murugesan, S.

Raghavan, M. Suganthi “Neural Network based Scheduling Algorithm for WiMAX with Improved QoS Constraints”, ICETECT, Mar 23-24, 2011, available in IEEExplore since May 2011.

