International Journal of Scientific & Engineering Research, Volume 5, Issue 3, March-2014 1146

ISSN 2229-5518

Call Admission Control in 3GPP LTE Systems at

High Vehicular Communications

Rajeshkumar RAMRAJ, Daryoush HABIBI, Iftekhar AHMAD

Abstract — In high-speed wireless communication, call dropping and blocking probability plays an important factor in providing connection-level Quality of Service (QoS). The main goal of this research is to develop an admission control algorithm to reduce call block and call drop. These parameters are mostly affected due to insufficient resources in the target cell. Radio resources management (RRM) plays a major role in providing better QoS in high speed vehicular communication. In this paper, we presents a novel CAC algorithm for high vehicular speed communication systems where the UE (mobiles) contain real time constant bit rate (rt-CBR) and non-real time variable bit rate (nrt-VBR) and the performance for these two traffic conditions are compared at different vehicular speeds.

Index Terms— Call Admission Control (CAC), 3GPP LTE, Call Dropping, Call Blocking, Channel Utilization, Bit Error Rate, Vehicular

Communication.

1. INTRODUCTION

——————————  ——————————
ong Term Evolution (LTE) of Universal Mobile Telecom- munication System (UMTS) has become one of the ad- vanced series in mobile telecommunication industries. In
1947, Bell labs, USA started the development of cell concept in mobile communications to increase the capacity of mobile networks by dividing the coverage area into smaller areas with its own base station, operating in different frequency. LTE or the Evolved Universal Terrestrial Access Network (E- UTRAN) which is a part of Evolved Packet Access (EPS), spec- ified by Third Generation Partnership Project (3GPP) has con- tinued dominating the cellular communication industries and progressed its development towards Fourth Generation (4G) networks. Considering the use of multimedia applications like, online mobile games, video calls, video chatting, online video streaming, mobile TV etc. in mobile communications, the telecommunication industry needs to satisfy the require- ments (data rates) of mobile users. These requirements suc- ceeded by the next generation wireless technology such as LTE, which is marketed as 4G LTE, as per 3GPP standard and LTE can operate at high-speed mobile communications.

————————————————

• Rajeshkumar RAMRAJ is currently with Centre for Communications and Electronics Research, pursuing PhD degree program in Electronics and Communication engineering in Edith Cowan University, Perth, Australia. E-mail: rramraj@our.ecu.edu.au

• Daryoush HABIBI, Professor and Leader, Centre for Communications and

Electronics Research, Edith Cowan University, Perth, Australia. E- mail: d.habibi@ecu.edu.au

• Iftekhar AHMAD, Sr Lecturer, Centre for Communications and Electron- ics Research, Edith Cowan University, Perth, Australia. E- mail: i.ahmad@ecu.edu.au

LTE system is based on Global System for Mobile Communi- cations / Enhanced Data Rate for GSM Evolution (GSM/EDGE), Universal Mobile Telecommunication System / High Speed Packet Access (UMTS/HSPA) technologies and some of the important requirements of this technology are high spectral efficiency, high peak data rates, and short round trip time and frequency flexibility [1]. LTE designed to be an IP-centric network that can provide QoS. LTE consists of en- hanced nodeBs (eNBs) in Radio Access Network (RAN), Mo- bility Management Entities (MMEs) and Serving Gateways (S- GW) in the core. X2 interface is used to interconnect eNBs in RAN and S1 interface to connect MMEs and S-GWs at the core (figure 1).

Fig 1. A Simple LTE Architecture

International Journal of Scientific & Engineering Research, Volume 5, Issue 3, March-2014 1147

ISSN 2229-5518

LTE supports peak download rates up to 300 Mbps and up- link rates up to 75 Mbps depending on the user equipment (UE) category by adopting Orthogonal Frequency Division Multiple Access (OFDMA) and Multiple-Input and Multiple-Output (MIMO) systems. LTE supports both Time Division Duplex (TDD) and Frequency Division Duplex (FDD) with same RAN and supports different spectrum flexibility with at least 200 ac- tive UE in every 5 MHz cell. Furthermore, LTE targets for smooth evolution and co-existence with legacy standards like GSM/EDGE, UMTS and CDMA2000, table 1 gives you the key specifications of 3GPP LTE. In the recent development of mobile communication industry, LTE technology plays a vital role and leads the industry towards the next generation networks. Even though the previous standards had some good development, researchers throughout the world are working on the current technology where the upcoming mobile wireless infrastructures are considered to support high-speed services with different QoS. Hence, some of the key aspects like multimedia services with real-time QoS guarantees are expected to improve further. To achieve this goal in multimedia services, this paper focuses on improving wireless transmission [2]. This can be achieved in many ways and here we aim to develop a Call Admission Con- trol (CAC) algorithm at vehicular speeds in LTE communication systems. In our simulation, we are considering the scenario (figure 3) where we are comparing real time-constant bit rate (rt-CBR) traffic against non-real time variable bit rate (nrt-VBR) traffic. BER and high data rate will have significant impact on UEs QoS in non-real time traffic (ftp data, web browsing) condi- tion. Radio Resource Management (RRM) handles the BER guarantee, by adaptive power allocation. In real time traffic condition (VoIP, live online video streaming), in addition to the QoS for non-real time traffic, delay is another important factor that need to be taken into account. When it comes to multime- dia service, delay should be given high importance in terms of QoS [3, 4] and bandwidth management in admission control framework [5].
The rest of the paper is organised as follows, section 2 will

talk about the pre-study on call admission control, section 3 will
talk about call admission control in high vehicular speeds, and section 3 will talk about our proposed call admission control algorithm and strategy, followed by results and discussion and conclusion in section 5 and 6 respectively.

2. CALL ADMISSION CONTROL – A PRE-STUDY

The wireless network for satisfying multimedia services must include a resourceful CAC, which is a process of regulating traffic in Voice over Internet Protocol (VoIP). This CAC will ensure the quality and performance level of audio and voice. CAC uses the strategy to limit the number of call connections to the network, so that the network congestion and call drop- ping is reduced. CAC is the process of regulating traffic vol- ume in voice communications, mainly in wireless mobile communication networks and in VoIP protocol. The objective of this CAC is to maintain a certain level of audio and voice quality in the communication network and to provide a de- sired QoS [6-9]. Ahmad discusses some of the main reason for CAC to have a guarantee QoS like improving signal quality, reducing call dropping and increasing the packet level per- formance. The author also discussessthe CAC classification with their pros and cons [10]. There are number of factors that the signal quality of a UE relies on, such as base station dis- tance, path loss, fading and noise. Motorola has proved in [11] that, AMC uses link adaption method to raise the overall sys- tem capacity [12]. CAC algorithm uses a key technology called Adaptive Modulation Coding (AMC) to optimize spectrum utilization [13]. In order to satisfy the QoS and utilize the available resources efficiently in vehicular communication, the call-level and packet-level service requirements should guar- antee an effective communication. A simple adaptive modula- tion technique used in LTE network in shown in figure 2. Link adaption technology is widely used in most wireless technolo- gy for an OFDMA downlink systems. Zuh et al, proposes a new CAC algorithm for OFDMA systems, where the link adaption technique is implemented to provide better QoS in a cell considering the probability distribution of the incoming traffic to determine the resource utilization [14]
CAC algorithm is of two types, static and dynamic CAC. Static CAC deals with resource allocation for hand-off calls [15,
16] and dynamic CAC deals with the radio channel status and available resources to perform the admission control [17, 18]. There are several research methods identified for new and
hand-off request, Zarai et al, addressed on a scheme where the load balance is taken into account to reduce the hand-off drop- ping. The author also proposes a CAC algorithm where the RB allocation is considered and it prioritises the hand-off calls over the new calls [19]. Bae et al, propose a resource-estimated CAC algorithm for a guarantee QoS and packet delay [9]. In order to provide better QoS, RRM is very important in CAC, Based on the resources available, the RRM will decide whether to accept or to drop the call. In [20] a delay aware CAC is proposed which guarantee a certain level of packet delay and prevents conges- tion in the base station. The traditional guard channel scheme has been the foundation to reserve a fixed number of RBs and this was proved by Ramjee et al. in [21], where he also defines

International Journal of Scientific & Engineering Research, Volume 5, Issue 3, March-2014 1148

ISSN 2229-5518

the linear objective functions for dropping and blocking proba- bilities. Angelos et al. propose a dynamic CAC mechanism con- sidering the issues of busy hour and the bandwidth allocation and request mechanism like Grant per Connection (GP) and Grant per Subscriber Station (PSST) were discussed [22].

Fig 2. Adaptive Modulation in LTE

S(f) =

𝐴

⎪ 2

𝑓

�

𝑑

⎪

⎩ 0,

|𝑓| < 𝑓𝑑

|𝑓| > 𝑓𝑑

(1)

3. CALL ADMISSION CONTROL IN HIGH VEHICULAR

SPEED

LTE offers a high data rate of 300 and 75 Mbps downlink and uplink capacity up to release 10. Release 11 has been consid- ered as LTE-A with 1 Gbps and 300 Mbps for downlink and uplink [13, 23]. In reality due to line of sight (LOS) issues, the recommended target is difficult to achieve. However research- ers attempt to prove their work with the help of simulators. LTE communication standard incorporates OFDMA for down- link and SC-FDMA for uplink to achieve better QoS and spec- tral efficiency. OFDMA extends system support by adopting PHY architecture and implementing the scalability factors from 1.5 MHz to 20 MHz. LTE anticipates the bandwidth spec- trum range from 1.4 MHz to 20 MHz and this choice of SC- FDMA in uplink places the complex, power hungry operation of frequency domain [24]. During high vehicular speeds due to the rapid variation of time and frequency, the relative mo- tion between transmitters and receivers reacts with very high Doppler shift. In such conditions, it causes signals to overlap and signal interference. This is widely addressed by various researchers as a fading problem and many solutions are pro- vided at various simulation conditions. Here in this paper we are going to address such kind of problem with a different simulating scenario. A simulation scenario used in this re- search with real time LTE communication system shown in figure 3.
The classical Doppler spectrum derived mathematically for a sin wave is given below.
Fig 3. A Simple Simulation Scenario

Fading problem causes bit error rate (BER) issues at the re- ceiver side, therefore BER probability increases at high vehicu- lar speed resulting unused bandwidth. OFDMA reduces inter- symbol interference (ISI) by utilizing guard channel. Due to the complexity at the receiver during transmission, SC-FDMA, chosen as a medium access for uplink in LTE [25]. The ISI is widely measured by “rms delay”, τrms seconds, a magnitude of components and the time difference. The maximum delay, τmax seconds, is the difference between shortest and the longest path. If the greatest path difference is Dmax km, then τmax = Dmax /0.3µs. Dmax depends on PHY characteristics and cell cov- erage area [24]. When it comes to BER estimation, Rayleigh fading is proved as a first-rate model which can rival error rate in the radio signal. Abdelhadi et al. proposed a unified ap- proach to derive BER for mobile WiMAX in uplink direction. The analytical expression was derived and the bit rate was obtained for BPSK, QPSK and 16QAM in adaptive modulation [26]. Gangane et al. in his paper, studied about the influence of LOS on SC-FDMA uplink BER behaviour for Rayleigh fading [27].

CAC in high vehicular communication systems is an im-
portant factor in providing better QoS. In our scenario, con- sider we have a eNBs controlled by MME/S-GW, and UE which is at “time t” (current time) moving at a initial speed of vt and a maximum speed of vmax . Consider we have a total us- able (available) bandwidth of Btotal in Mbps, when the eNB receives a request from the UE with a demand of Breq in Mbps, the CAC scheme should satisfy the below condition.
𝑈𝑐𝑎𝑝 ≥ 𝐵𝑟𝑒𝑞 + 𝐵𝑡𝑜𝑡𝑎𝑙 (2)

International Journal of Scientific & Engineering Research, Volume 5, Issue 3, March-2014 1149

ISSN 2229-5518

Ucap is the usable capacity of the UE and it varies according to the spped. When the UEs increases its speed the Ucap drops and the dropping rate increases significantly resulting in QoS degradation.

4. PROPOSED SCHEME

4.1. Bit Error Rate Estimation at High Vehicular

Speed

In the proposed CAC scheme, our initial step is to estimate
BER using Rayleigh fading model. Based on this the CAC de-
From equation (5) and (6) we can derive the bit error for 1 bits/symbol Pb

𝑃𝑏 = 𝑄(�𝛾𝑠 ) (7)
When QPSk transmits 2 bits/symbol, symbol energy is
𝐸ℎ = 2𝐸𝑏 Eb is bit energy and 𝛾𝑠 = 2𝛾𝑏 which will derive equa-
tion (4).
When we consider bit verses symbol error probability, the

probability decision error P[e] otherwise known as PM is de- rived in equation (5). Though we are interested in error prob- ability Pb , and the data symbols corresponds to log2 M data bits, we have,

cision is made. For high speed vesicular communication, error control mechanism like Forward Error Control (FEC) is used as they use extra parity bits to recover the corrupted bits and

𝑃𝑀
𝑙𝑙𝑙2 𝑀
≤ 𝑃𝑏
≤ 𝑃𝑀
(8)
with this we can avoid re transmission of packets in wireless channels. At high vehicular speeds the rate of transmission goes down rapidly and this creates very high impact on effec- tive throughput. To overcome this issue we use a FEC mecha- nism in OFDMA DL to avoid relatively high throughput. In LTE communication channels, we have parameters illustrated as follows:
Hence, Pb and PM relate each other, we have:

𝑃 ≈ 𝑃𝑀 (9)

𝑏 𝑙𝑙𝑙2 𝑀

When we calculate the error probability for Rayleigh fad-
ing, the received bit and symbol energy-to-noise ratios 𝛾𝑏 and
𝛾𝑠 have exponential PDFs, therefore we have,
We have N as the number of OFDMA sub-carriers, fD as maximum Doppler frequency, VUE as the speed of the UE, C

1
𝑃𝛾𝑏 (𝑥) = 𝛾�

−𝑥

𝑒 𝛾�𝑏
, 𝑥 ≥ 0 (10)
as the speed of the wavelength (light), and fc as carrier fre-
quency [28].
Now we have the Doppler frequency as,
𝑃𝛾𝑠 (𝑥) =

1
𝛾�𝑠

−𝑥

𝑒 𝛾�𝑠 , 𝑥 ≥ 0 (11)

𝑉𝑈𝐸

Since there are log2 M bits/modulated symbol, we have:

𝑓𝐷 = 𝑓𝐶 � 𝐶 � (3)

𝛾 = 𝛾 𝑙𝑙𝑙 𝑀 (12)

𝑠 𝑏 2

We have error probability of various forms of PSK signals, BPSK transmits 1 bits /symbol, so energy En = Eb , where Eb is the bit energy. Therefore, error probability is given by,

𝛾�𝑠 = 𝛾�𝑏 𝑙𝑙𝑙2𝑀 (13)
For such channels the error probability is given as,

∞

𝑃𝑏 (𝛾𝑏 ) = 𝑄�√2𝛾𝑏 � (4)
Where 𝛾𝑏 is defined as received bit energy-to-noise ratio and
𝛾𝑏 is defined as average received bit energy-to-noise ratio.
The below mathematical calculation will derive equation (4) to
𝑃𝑏 = � 𝑄�√2𝑥�𝑃𝛾𝑏 (𝑥)𝑑𝑥

∞

𝑃𝑏 = � 𝑃𝑏 (𝑥)𝑃𝛾𝑏 (𝑥)𝑑𝑥

(14)
(15)
calculate the error probability.
In a cell area with x and y axis, the QPSK signal can be ro-
tated at 45’ without changing the error probability due to rota-
tion invariance property. In such cases the decision bounda-

1
= 2 �1 − �
𝛾�𝑏
1 + 𝛾�𝑏
� (16)
ries corresponds to real and imaginary axis of complex signals. In OFDMA DL air interface, inter-carrier interference causes main issue for BER; inorder to overcome this, Additive white Gaussian noise (AWGN) is used. Considering AWGN, the
noise vectors in a cell area is 𝑛� = 𝑛� 𝐼 + 𝑗𝑛� 𝑄 where 𝑛� 𝐼 and 𝑛� 𝑄 are
The above BPSK and QPSK bit error probability when plot-
ted for an AWGN channel and a Rayleigh fading channel, it
will result in a huge loss of performance [28]. Therefore for M-
PSK the average symbol error probability is given as:

∞

independent Gaussian variables with variance No . Consider- ing this the probability symbol error PM is given as,
𝑃𝑠 = � 𝑃𝑀 (𝑥)𝑃𝛾𝑠 (𝑥)𝑑𝑥

(17)

𝑃𝑀 = 1 − �1 − 𝑄��𝛾𝑠 ��
(5)
Now with the above equation, the average received symbol
energy-to-noise ratio is given as:
Where 𝛾𝑠 is defined as the received symbol energy-to-noise ratio and 𝛾�𝑠 is defined as the average symbol energy-to-noise
ratio. Let Pb be the probability of bit error, then we have,
𝛾�𝑠 =

1
1 − 1 [𝑁 + 2 ∑𝑁−1(𝑁 − 𝑖)𝐽 (2𝜋𝑓 𝑇 𝑔 𝑖)] + 𝑁𝑇𝑠
(18)
𝑃𝑀 = 1 − (1 − 𝑃𝑏 )2 (6)

𝑁2

𝑖=1

0 𝐷 𝑠

( 𝐸𝑠 )
𝑁0

International Journal of Scientific & Engineering Research, Volume 5, Issue 3, March-2014 1150

ISSN 2229-5518

The average received bit energy-to-noise ratio 𝛾�𝑏 can be de-
rived from equation (13), therefore we have the following:

𝛾� = 𝛾�𝑠 (19)

𝑏 𝑙𝑙𝑙2 𝑀

Now 𝛾�𝑏 is expressed as,
𝛾�𝑏 =

𝑙𝑙𝑙2𝑀
( 1 )
(20)

1 − 1 [𝑁 + 2 ∑𝑁−1(𝑁 − 𝑖)𝐽 (2𝜋𝑓 𝑇 𝑔 𝑖)] + 𝑁𝑇𝑠 𝐸𝑏
𝑁2

𝑖=1

0 𝐷 𝑠

𝑙𝑙𝑙2 𝑀
𝑁0
The above average received bit energy-to-noise ratio 𝛾�𝑏 and average received symbol energy-to-noise ratio 𝛾�𝑠 is verified in-

order to examine ICI and the Bessel function J0 (.) order zero is

used to estimate the symbol rare for M-ary QAM. And finally for the value M=2 and M=4, the error probability Pb can be derived from the equation (15). With higher order of M and with a speed factor v, the error probability can be calculated using the expression below,

𝑃 (𝑣) = 𝑃𝑀 (𝑣) (21)

𝑏 𝑙𝑙𝑙2 𝑀

In OFDMA with a low 𝛾�𝑏 AWGN dominates the perfor-
mance and with the extra noise due to ICI has little effect. At
the same time high 𝛾�𝑏 , the ICI dominates the performance and
cause an error floor [28].

4.2 Call admission Control Strategy and

Throughput Estimation

The call admission control strategy and its decision is based on the throughput estimation, consider the simulation scenario from figure 3, say we have a UE moving at a speed of 120 km/hr, speed can be represented as vmax , with M-ary the sym- bol and bit error probability Ps , Pb is represented as,

𝑃𝑠 (𝑣𝑚𝑎𝑥 )𝑀 = 1 − (𝑃𝑏 (𝑣𝑚𝑎𝑥 ))𝑀 (22)

As we are going to compare the real time and non-real
time traffic conditions in our scenario, let us assume that the
traffic generation rate from the traffic source is g bits per se- cond and the traffic arrival rate in the cell area is i, therefore
the mean traffic arrival 𝜑𝑖 is expressed as:
𝜑𝑖 = 𝑛� 𝑖 𝑙 (23)
Where 𝑛� 𝑖 is the number of UEs in the cell area, now let us
consider the UE in the cell is sending a request for N number
of RBs which contains Y symbols, the corrupted or ruined
symbols Z can be calculated as follows:
Table 2: Simulation Parameters
Where, K is the parity bit recovered through forward error correction (FEC).
The CAC algorithm used in this paper is given below:
• When the UE is active in a cell and its speed and di- rection is calculated by the eNB. The eNB provides the bandwidth information to all the active UE in the cell area. The average bit error rate probability
𝑃𝑏 (𝑣𝑚𝑎𝑥 ) and average symbol error probability
𝑃𝑠 (𝑣𝑚𝑎𝑥 ) at the maximum speed is estimated at this
stage.
• UE will now send a request to the eNB asking N
number of RBs to answer the call.
• Now the eNB will calculate the requested RBs with the available RBs, based on this the decision will be made and the required RBs will be allocated to the UE.
• However, if there are insufficient RBs, then the re- maining available RBs will be reserved.
• eNB will now receive a new request from a UE and both the reserved RBs and allocated RBs will be sub- tracted from the total cell capacity to allocated the new required RBs
• If the requested RBs is less than or equal to the total available RBs, then the bandwidth is allocated. This also should satisfy the condition specified in equation (2)

• Otherwise the call is rejected.

Algorithm 1: Proposed CAC Algorithm initialization;

gather UE information;
compute total available RBs;
verify call type;
𝑍𝑚𝑎𝑥 = 𝑌 × 𝑃𝑠 (𝑣𝑚𝑎𝑥 ) (24)

if U

cap

≥ Breq
+ Btotal

then

When the UE receives the requested RBs from the header h, it is very important to check for an overhead rate (Hover ) per bit received and this can be expressed as:

𝐾(𝑣max)

| accept call;

else

| drop call and go begin the process again;

end

𝐻𝑜𝑣𝑒𝑟 = ℎ + �
� (25)
𝑌

International Journal of Scientific & Engineering Research, Volume 5, Issue 3, March-2014 1151

ISSN 2229-5518

5. SIMULATION RESULTS AND DISCUSSINON

The simulator used for this research is NS2, and the simple simulation scenario is given in ﬁgure 3. The simulation pa- rameters was given in Table 2. The maximum bandwidth for the LTE channel was 20 MHz with data rate capacity of 20
Mbps. The queue length is 10mb and trafﬁc rate was distribut-
ed in the range 128 kbps to 512kbps. In the simulation, we are
considering two different types of trafﬁc, ﬁrst as real time CBR
and second as non-real time VBR. The simulation was con-
ducted at a speed on 120km/hr, ﬁnally at the end of the simu- lation call blocking rate, dropping rate and channel utilization during the call duration was calculated and analyzed for above mentioned trafﬁc conditions.
The simulation results are evaluated as per the following
performance evaluation metrics:
• Estimating the bit error rate probability Pb from equa- tion (4).
• Estimating the amount of available RBs before band-
width allocation.
• Evaluate the type of call based on the trafﬁc condi- tion.
• Estimate the amount of available RBs after resource allocation.
• Physical RBs utilization, (i.e.), number of allocated RBs for the UE during the whole simulation time. Considering the scenario, when the UE starts to move at a
speed of 120km/hr, the corresponding eNB will get all the
relevant information such as speed, UE direction, position etc.
and the eNB will compute the available BW (Ucap) every mi- nute. Now the UE will send a request (Breq ) to BS and the eNB
will check for the condition given in the equation (2). If the above conditions satisfy, the eNB will allocate the requested RBs. At the same time, the eNB will be checking the call type and trafﬁc, because the eNB will consider only request which follows the CAC algorithm. Here in this simulation, the speed factor plays an important role in estimating the error probabil- ity (Pb ) and throughput estimation given in section 4.2. When the UE moves at a maximum speed vmax the channel usable capacity is calculated based on the estimated throughput at the maximum speed vmax and the CAC decision is made as follows:
𝑈𝑐𝑎𝑝 (𝑣𝑚𝑎𝑥 ) ≥ 𝐵𝑟𝑒𝑞 (𝑣𝑚𝑎𝑥 ) + 𝐵𝑡𝑜𝑡𝑎𝑙 (𝑣𝑚𝑎𝑥 ) (26)
The simulation will be scheduled to perform in the cell ar-
ea where the UE is active and the call blocking rate, call drop- ping rate channel utilization is calculated. When the UE moves from one cell to another cell (adjacent cell), the packet genera- tion is continued and the correspondence is started in the hand-off in time and when the UE connects to the next eNB and that is the hand-off out time, and here the delay is calcu- lated. In RRM of CAC algorithm, Breq is decided based on the service type and trafﬁc.
Fig 4. Call Blocking Rate when the UEs at 120km/hr Speed
Figure 4 represents the call blocking rate, x-axis shows the number of call arrival per minute and y-axis shows the call blocking rate in %. The simulation is analyzed for the above mentioned two trafﬁc types and the results shows the call blocking rate for rt-CBR is less compared to nrt-VBR. The sim- ulation results shows the blocking percentage for rt-CBR is around 6.8% for 10 calls per minute and 8% for nrt-VBR for 10 calls per minute. As explained in [3, 4] CBR trafﬁc has a regu- lar and constant trafﬁc ﬂow compared to VBR, where the trafﬁc conditions have high impact on user condition.

Fig 5. Call Dropping Rate when the UEs at Speed 120km/hr
Figure 5 represents the call drop ration for the proposed CAC algorithm, where the percentage of call drop for real time trafﬁc is less than the non-real time trafﬁc. Call drop percent- age for rt-CBR is around 6.8% and 7.5% for nrt-VBR, this was calculated at 10 calls per minutes with maximum of 24 UEs. In rt-CBR traffic, after the requested the RBs allocated, the RRC connection is established with MME and UE. The connection session is smooth because of the constant flow in the traffic, where the connection between the MME and UE drops in nrt- VBR, where the traffic flow is not stable.

International Journal of Scientific & Engineering Research, Volume 5, Issue 3, March-2014 1152

ISSN 2229-5518

Fig 6. Channel Utilization

Figure 6 shows the channel utilization in both trafﬁc con- ditions, which conﬁrms the better bandwidth utilization of rt- CBR, where the ﬂow of data rate is at a constant speed when compared to nrt-VBR, where the data rate differs on users us- ability. Channel utilization for rt-CBR is calculated as 62% and the utilization for nrt-VBR varies due to variable bit rate. We should notice the utilization ratio reduces gradually for real time trafﬁc and for non-real time trafﬁc, the utilization curve varies according to the variable bit rate. RRM is the primary reason for the CAC RBs reservation by adjusting the attribu- tion of resources intelligently.
Fig 7. UEs Speed Vs Simulation Time
Figure 7 shows the UEs position and speed at various simulation time. When the UE increases its speed, and moves away from the eNB, the usable capacity of the UE drops signiﬁcantly which will affect the drop and block probability and other QoS factors.

6. CONCLUSION

This research developed a new CAC algorithm which re- duces the call blocking, call dropping rates and improved channel utilization. The proposed algorithm based on RBs reservation for current and ongoing calls, and new calls. The eNodeB will calculate the available and total RBs from the available information and the eNB manages the RB allocation. We evaluate the call drop probability and call block probabil- ity via the proposed CAC scheme and simulation. The future
work will target on cell management and resource allocation for high-speed wireless vehicular communications.

7. REFERENCES

[1] S. Sesia, I. Toufik, and M. Baker, LTE: the UMTS long term evolution: Wiley Online Library, 2011.

[2] "IEEE Draft Standard for Local and Metropolitan Area Networks - Specific Requirements - Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications - Amendment 3: Enhancements for Very High Throughput in the 60

GHz Band," in IEEE P802.11ad/D8.0, May 2012 (Draft Amendment based on IEEE 802.11-2012), ed, 2012, pp. 1-667.

[3] H. Haci, Z. Huiling, and W. Jiangzhou, "Novel scheduling for a mixture of real-time and non-real-time traffic," in Global Communications Conference (GLOBECOM), 2012 IEEE, 2012, pp.

4647-4652.

[4] E. Yaacoub and Z. Dawy, "A Survey on Uplink Resource Allocation in OFDMA Wireless Networks," Communications Surveys & Tutorials, IEEE, vol. 14, pp. 322-337, 2012.

[5] R. Ramraj, "Bandwidth allocation and routing information for wireless mobile Ad-Hoc networks," in Fuzzy Systems and Knowledge Discovery (FSKD), 2010 Seventh International Conference on, 2010, pp. 2968-2972.

[6] V. K. N. Lau and S. V. Maric, "Mobility of queued call requests of a new call-queueing technique for cellular systems," Vehicular Technology, IEEE Transactions on, vol. 47, pp. 480-488, 1998.

[7] F. Yuguang and Z. Yi, "Call admission control schemes and performance analysis in wireless mobile networks," Vehicular Technology, IEEE Transactions on, vol. 51, pp. 371-382, 2002.

[8] C. Chang, C. Chung-Ju, and L. Kuen-Rong, "Analysis of a hierarchical cellular system with reneging and dropping for waiting new and handoff calls," Vehicular Technology, IEEE Transactions on, vol. 48, pp. 1080-1091, 1999.

[9] S. Bae, J. Lee, B.-G. Choi, S. Kwon, and M. Chung, "A Resource- Estimated Call Admission Control Algorithm in 3GPP LTE System," in Computational Science and Its Applications – ICCSA 2009. vol. 5593, O. Gervasi, D. Taniar, B. Murgante, A. Laganà, Y. Mun, and M. Gavrilova, Eds., ed: Springer Berlin Heidelberg, 2009, pp.

250-260.

[10] M. H. Ahmed, "Call admission control in wireless networks: a comprehensive survey," Communications Surveys & Tutorials, IEEE, vol. 7, pp. 49-68, 2005.

[11] "Adaptive Modulation and Coding (AMC)," in 3GPP vol. 17, ed.

Sweden Stockholm: Motorola, 2000, p. 7.

[12] 3GPP, "LS on Technical Report on HSDPA " 3GPP TSG RAN, vol.

16, 2000.

[13] R. Ramraj, I. Ahmad, and D. Habibi, "Multi-Access Environments in Next Generation Networks," in Wireless Multi-Access

International Journal of Scientific & Engineering Research, Volume 5, Issue 3, March-2014 1153

ISSN 2229-5518

Environments and Quality of Service Provisioning: Solutions and

Application, ed: IGI Global, 2012, pp. 1-17.

[14] Z. Xinning, L. Xin, and Z. Qian, "A link adaptation based call admission control algorithm for OFDMA systems," in Advanced Communication Technology (ICACT), 2012 14th International Conference on, 2012, pp. 158-163.

[15] V. Phan-Van, "Soft decision call admission control versus static fixed channel assignment and interference measurement based policies for WCDMA cellular PCN's," in Personal, Indoor and Mobile Radio Communications, 2001 12th IEEE International Symposium on,

2001, pp. E-27-E-31 vol.2.

[16] D. McMillan, "Delay analysis of a cellular mobile priority queueing system," Networking, IEEE/ACM Transactions on, vol. 3, pp. 310-319, 1995.

[17] L. Gan, Z. Guangxi, and W. Weimin, "An adaptive call admission policy for broadband wireless multimedia networks using stochastic control," in Wireless Communications and Networking Conference, 2004. WCNC. 2004 IEEE, 2004, pp. 1324-1329 Vol.3.

[18] S. Wu, K. Y. M. Wong, and B. Li, "A dynamic call admission policy with precision QoS guarantee using stochastic control for mobile wireless networks," IEEE/ACM Trans. Netw., vol. 10, pp. 257-271,

2002.

[19] F. Zarai, K. B. Ali, M. S. Obaidat, and L. Kamoun, "Adaptive call admission control in 3GPP LTE networks," International Journal of Communication Systems, pp. n/a-n/a, 2012.

[20] S. J. Bae, B.-G. Choi, M. Y. Chung, J. J. Lee, and S. Kwon, "Delay- aware call admission control algorithm in 3GPP LTE system," in TENCON 2009-2009 IEEE Region 10 Conference, 2009, pp. 1-6.

[21] R. Ramjee, D. Towsley, and R. Nagarajan, "On optimal call admission control in cellular networks," Wireless Networks, vol. 3, pp. 29-41, 1997.

[22] A. Antonopoulos and C. Verikoukis, "Dynamic Call Admission Control for Enhanced GoS of UGS Connections during “Busy Hour” in WiMAX," in Mobile Lightweight Wireless Systems, ed: Springer, 2010, pp. 346-355.

[23] G. Araniti, C. Campolo, M. Condoluci, A. Iera, and A. Molinaro, "LTE for vehicular networking: A survey," Communications Magazine, IEEE, vol. 51, pp. 148-157, 2013.

[24] H. G. Myung and D. Goodman, Single carrier FDMA: a new air interface for long term evolution vol. 8: John Wiley & Sons, 2008.

[25] J. J. Sánchez-Sánchez, U. Fernández-Plazaola, and M. C. Aguayo- Torres, "BER analysis for SC-FDMA over Rayleigh fading channels," in Broadband and Biomedical Communications (IB2Com),

2011 6th International Conference on, 2011, pp. 43-47.

[26] R. Abdelhadi, H. Elbadawy, and S. El-Ramly, "A unified approach for performance analysis in mobile WiMAX networks with Adaptive Modulation and Coding schemes over Rayleigh fading

channels," in Personal, Indoor and Mobile Radio Communications,

2009 IEEE 20th International Symposium on, 2009, pp. 3183-3187.

[27] J. Gangane, M. Aguayo-Torres, J. Sanchez-Sanchez, and U.

Fernández-Plazaola, "SC-FDMA performance over Ricean channel," in Broadband and Biomedical Communications (IB2Com),

2011 6th International Conference on, 2011, pp. 91-96.

[28] G. L. Stüber, Principles of mobile communication: Springer, 2011.