Performance Evaluation of Scheduling Algorithms with Different ... - ijcst

IJCST Vol. 6, Issue 3, July - Sept 2015

ISSN : 0976-8491 (Online) | ISSN : 2229-4333 (Print)

Performance Evaluation of Scheduling Algorithms with Different MIMO Techniques in LTE Systems 1

Krishna Teja Yadav CH. T, 2C.Y. Gopinath, 3Mohankumar N. M., 4Devaraju J.T Dept. of E&C, Bangalore Institute of Technology, Bangalore, India Dept. of Electronic Science, Bangalore University, Bangalore, India

1,2 3,4

Abstract MIMO techniques are used in Wireless Broadband Access (BWA) networks to maximize spectrum efficiency and minimize the bit error rate. LTE is one such BWA network which has adopted MIMO techniques in both the uplink and downlink along with Radio Resource Management (RRM) aspects like scheduling to improve the data rate. Scheduling is mainly concerned with allocating the available radio resources among the users depending upon the metrics such as Quality of Service (QoS) requirements of users, channel conditions etc. Hence in this paper, an attempt is made to study and compare the performance of scheduling algorithms (RR, PF, MT and BET) with MIMO techniques such as SISO, SIMO, SFBC and OLSM for Constant Bit Rate (CBR) traffic scenario. The performance metrics used are average throughput and average delay.

called eNodeBs (eNB), which is responsible for handling the radio related communications between the User Equipment (UE) and the EPC. the architecture of E-UTRAN is shown in the Fig2.

Keywords SISO, SIMO, SFBC, OLSM, Round Robin, Proportional Fair, Maximum Throughput, Blind Equal Throughput. I. Introduction Ever increasing demand for multimedia applications such as video streaming, video conferencing, IP TV and many more has driven network vendors to adopt LTE. LTE technology provides enhanced peak data rates in mobile platform by adopting MIMO techniques and advanced Radio Resource Management (RRM) mechanisms such as scheduling. MIMO techniques provide a way of utilizing the multiple signal paths that exist between a transmitter and receiver to significantly improve the data throughput which can be achievable on a given channel bandwidth. Whereas, RRM mechanisms are critical for dynamic management of radio spectrum which is the most valuable resource in mobile technology. One of the key RRM mechanisms is packet scheduling which allocates suitable radio resources to each to each User Equipment (UE) for transmission through the air interface. Hence in this paper an attempt has been made to study the effect of MIMO and Packet scheduling on LTE system by considering average throughput and average delay as performance metrics. The rest of the paper is organized as follows. Section II de-scribes the architecture of LTE network. MIMO techniques are discussed in section III. Section IV describes the scheduling algorithms used in LTE. Simulation studies and results are given in section V and paper is concluded in Section VI. II. LTE Network Architecture The systems designed using LTE technologies are packet-based network, which consists of Evolve Packet Core (EPC) and Evolved Universal Terrestrial Radio Access Network (E-UTRAN). The architecture of EPC is shown in the Fig1. EPC functionalities include, charging and rate policing, setup of end-to-end connections, mobility management and authentication[2]. EPC is a single framework for packet-based real-time and non-realtime services. The Evolved Universal Terrestrial Radio Access Network (E-UTRAN) contains network of LTE base stations

214

International Journal of Computer Science And Technology

Fig. 1: Core-network (EPC) Architecture

Fig. 2: Radio-Access-Network Interfaces III. Multiple Input Multiple Output (MIMO) Antenna Techniques MIMO systems are used mainly for high speed wireless communications, since the use of multiple antennas provides better performance gains over single antenna system. There are different types of MIMO techniques[3] such as Single Input Single Output (SISO), Single Input Multiple Output (SIMO), Space Frequency Block Coding (SFBC) and Open Loop Spatial Multiplexing (OLSM). w w w. i j c s t. c o m



Single Input Single Output (SISO)

Fig. 3: SISO-Single Input Single Output. Single Input Single Output (SISO) technique uses one an-tenna at the transmitter and one antenna at the receiver for radio transmissions (Fig. 3). Single Input Multiple Output (SIMO)

Fig. 4. SIMO-Single Input Multiple Output. A system which uses a single antenna at the transmitter and multiple antennas at the receiver is termed as Single Input Multiple Output (SIMO) (Fig. 4). In SIMO, The received signals are added up for maximizing the signal to noise ratio (SNR) after linear combination thereby achieving higher datarates. Space Frequency Block Coding (SFBC)

OLSM is one of the downlink transmission modes that can support the higher data rate in current releases of LTE. OLSM consist of two transmit antennas at the eNB and two receive antennas at the UE (2x2 antenna configuration), sending either one or two simultaneous data streams (code words in LTE parlance) from the eNB to the UE (Fig. 6). In a 2x2 antenna configuration, sending one data stream is known as Rank1 MIMO and sending two data streams is known as Rank2 MIMO. The number of independent data streams that can be sent to the UE is restricted to either one or two steam, even if the number of transmit antennas at the eNB is increased to four. IV. Scheduling Algorithms In LTE Scheduling Algorithms are mainly concerned with allocat-ing the available radio resources among various users de-pending upon the metrics like conditions of the channel, Quality of Service (QoS) requirements of users etc. Scheduling Algorithms are required to increase the throughput of the individual users. There are no standardized scheduling algorithms implemented at the MAC layer of eNB in the LTE architecture. Hence the choice of different scheduling algorithms leads to different level of system performance[4-5]. In this section, Round Robin (RR), Proportional Fair (PF), Maximum Throughput (MT) and Blind Equal Throughput (BET) scheduling algorithms are discussed. A. Round Robin (RR) In Round Robin (RR) Scheduling Algorithm, the available resources are shared among various users one after the other without any priority. The scheduling is only based on the available Resource Blocks (RBs) during scheduling process [6]. It carries out fair sharing of time resources among all the users B. Blind Equal Throughput (BET) Throughput Fairness can be achieved with Blind Equal Throughput (BET) which considers the past average throughput achieved by each user and uses it as metric [3]. BET metric for the ith user is calculated as in (1)

m

BET i ,k

=

1 R (t ) i

Where is Ri (t ) given in (2). Fig. 5: SFBC-Space Frequency Block Coding SFBC is a transmit diversity technique which is used to reduce the effect of multipath effect and interference. SFBC technique employs two transmit antennas and one receive antenna to improve the signal quality (Fig. 5). Open Loop Spatial Multiplexing (OLSM)

Fig. 6: OLSM)-Open Loop Spatial Multiplexing

(1)

(2)

Where and r (t ) is the data rate achieved by the ith user at time t. The factor Ri (t − 1) represents the past average throughput experienced by the ith user at time t which is calculated as a moving average and it is updated every Transmission Time Interval (TTI). BET scheduling algorithm allocates resources to users with lower past average throughput at each TTI. Hence the user with lowest throughput will be served till same throughput as that of other users in the cell is achieved. In particular, users with bad channel conditions are allocated more often leading to consequent fairness improvement. i

C. Maximum Throughput (MT) MT scheduling algorithm aims at maximizing the overall throughput by assigning radio resources to the user that can achieve the maximum throughput in the current TTI. MT metric for ith user is calculated as in (3)

w w w. i j c s t. c o m

(3)

International Journal of Computer Science And Technology 215



PHY-Tx Power

Where d ki (t ) is expected data rate for ith user at time t on the kth RB is given in (4) i d (t) = log ( 1 + SINRk (t)) (4) MT scheduling algorithm maximizes cell throughput, on the other hand it performs unfair resource sharing since users with poor channel conditions (e.g., cell-edge users) will only get a low percentage of the available resources. i k

D. Proportional Fair (PF) The PF scheduling algorithm provides a good tradeoff between system throughput and fairness by selecting the user with highest instantaneous data rate achievable relative to its past average throughput. PF metric is given in (5)

(5)

Where d ki (t ) is expected data rate for ith user at time t on the kth RB is given in (4) and Rki (t ) given in (2). The factor Rki (t ) represents the past average throughput experienced by the ith user until time t, is calculated as a moving average and it is updated every TTI for each user. Since the past average throughput act as a weighting factor of the expected data rate, users in bad conditions will be served within as certain amount of time. The parameter β is related to the time window T f , over which fairness wants to be imposed, according to the relation (6)

PHY-Rx Antennas UE

Antenna Height MAC-Scheduler-Type

12 dBm 1(SISO), 2(SIMO), 1(SFBC), 2(OLSM) 1.5m Simple Scheduler

The snapshot of the scenario designed is shown in the Fig 7. The constant shadowing model and two ray path loss models are considered for simulation studies and the remaining simulation parameters are listed in the Table 1. Initially simulation study is carried out for RR scheduling algorithm and SISO antenna technique in Rayleigh fading environment by enabling vehicular mobility to all UEs along pre-defined paths from centre to edge of the cell. Simulation studies are repeated for PF, MT and BET scheduling algorithms one at a time. Similar simulations are carried out for SIMO, SFBC and OLSM multiantenna techniques

(6)

The choice between these schedulers is necessary to satisfy some degree of fairness by exploiting fast variations in channel conditions.

Table 1: Simulation Parameters Property Simulation-Time Downlink-Channel Frequency uplink-Channel Frequency Shadowing mean Channel-Fading-Model Channel-Bandwidth Antenna-Model PHY- Tx-Power PHY-Num-Tx-Antennas Antenna-Height eNB

216

MAC-Tx-Mode

Value 100 sec 2.4GHz 2.5GHz 4dB Rayleigh 10MHz Omni directional 23dBm 1 12m 1(SISO), 1(SIMO), 2(SFBC), 2(OLSM)

International Journal of Computer Science And Technology

Fig. 7: Snapshot of Scenario Designed With Pre-Defined Movement of UEs from centre to edge of the cell 1000000

Average Throughput (bps)

V. Simulation Studies and Results The performance of RR, PF, MT and BET scheduling algorithms for different multi-antenna techniques like SISO, SI-MO, SFBC and OLSM are evaluated using Qualnet 7.1 network simulator. The performance metrics considered for simulation studies are average throughput and average delay. The scenario designed for simulation studies consists of single cell with an eNB and 10 UEs which are placed at the centre of the cell in a terrain area of 5kmX5km. A downlink CBR connection with a data rate of 32Mbps is established between eNB and each UE.

900000 800000 700000 600000 500000 RR PF MT BET

400000 300000 SISO

SIMO

SFBC

OLSM

Antenna Techniques

Fig. 8: Average Throughput performance of various scheduling Algorithms in Rayleigh Fading Region Fig. 8 and Fig. 9 shows the average throughput and average delay for different multi-antenna techniques in Rayleigh fading channel w w w. i j c s t. c o m



respectively. From fig 8, it is observed that MT and PF scheduling algorithm achieves higher system throughput compared to RR and BET scheduling algorithms, since PF and MT algorithms consider channel conditions (highest instantaneous data rate achievable for each UE for each RB) in radio resources allocation[8-10].

Average Throughput (bps)

1000000

SISO SIMO SFBC OLSM

900000 800000 700000 600000 500000 400000 300000 RR

PF

MT

BET

Scheduling Algorithms

Fig. 9. Average Throughput performance of various Antenna Techniques in Rayleigh Fading Region

Average Delay (ms)

250.0 237.5 225.0 212.5 200.0

RR PF MT BET

187.5 175.0 SISO

SIMO

SFBC

OLSM

Antenna Techniques

Fig. 10: Average Delay Performance in Rayleigh Fading Region

[2] D. Astely, E. Dahlman, A. Furuskar, Y. Jading, M. Lindstrom, S.Parkvall,“LTE- The Evolution of Mobile Broadband”, IEEE Commun. Mag., Vol. 47, No. 4, April 2009, pp. 4451. [3] Suk-Bok Lee, Ioannis Pefkianakis, Sayantan Choudhury, Shugong Xu, Songwu Lu,“Exploiting Spatial, Frequency, and Multiuser Diversity in 3GPP LTE Cellular Networks”, IEEE Transactions on mobile computing, Vol. 11(11), 2012. [4] Salman A. AlQahtani, Mohammed AlHassany –“Performance Modeling and Evaluation of a Novel Scheduling Algorithm for LTE Networks”, IEEE 12th International Symposium on Network Computing and Applications, pp. 101 - 105, 2013. [5] Tsung-Yu Tsai, Yao-Liang Chung, Zsehong Tsai, “Introduction to Packet Scheduling Algorithms for Communication Networks”, Communications and Networking, Chapter 13, 2010. [6] Swetha, Mohankumar N M , Devaraju J T,“Performance Evalua-tion of Round Robin and Proportional Fair Scheduling Algorithms for Constant Bit Rate Traffic in LTE” International Journal of Computer Networks and Wireless Communications (IJCNWC), Vol. 3 (1), 2013. [7]. Mohankumar N M, Swetha, Mohana H K, Devaraju J T,"Effect of Adaptive Modulation and Coding Schemes on Scheduling Algorithms for LTE Downlink", International Journal of software & Hardware Research in Engineering (IJSHREE), Vol. 1 (2), 2013. [8] Mohsen Mollanoori, Majid Ghaderi,“Fair and Efficient Scheduling in Wireless Networks with Successive Interference Cancellation”, IEEE Wireless Communications and Networking Conference (WCNC), pp. 221 - 226, 2011. [9] Raymond Kwan, Cyril Leung, Jie Zhang,“Proportional Fair Multiuser Scheduling in LTE”, IEEE Signal Processing Letters, Vol. 16, No. 6, 2009. [10] Capozzi, G. Piro, L. Grieco, G. Boggia, P. Camarda, “Downlink Packet Scheduling in LTE Cellular Networks: Design Issues and a Survey”, IEEE Communications Surveys & Tutorials, pp. 1-23, 2012.

VI. Conclusion The simulation is performed for predefined movement of the UEs from the center to the edge of the cell to compare the performance of different scheduling algorithms such as RR, PF, MT and BET in the Rayleigh fading region. The PF and MT scheduling algorithm provides comparatively better results. Further, SIMO multi antenna technique provide better average throughput and average delay performance. VII. Acknowledgment The authors wish to thank UGC for providing Junior Research Fellowship under ‘At Any One Given Time Basis Scheme’ to carry out the research work. References [1] Daniel Biiltmann, Torsten Andre, Rainer Schoenen, “Analysis of 3GPP LTE-Advanced Cell Spectral Efficiency”, 21st Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, pp. 1876 - 1881, 2010. w w w. i j c s t. c o m

International Journal of Computer Science And Technology 217