Impact of Signaling Load on the UMTS Call.pdf

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/224314505
Impact of Signaling Load on the UMTS Call Blocking/Dropping
Conference Paper · June 2008
DOI: 10.1109/VETECS.2008.552 · Source: IEEE Xplore
CITATIONS
6
READS
82
3 authors, including:
Some of the authors of this publication are also working on these related projects:
IoT Security View project
David Tipper
University of Pittsburgh
217 PUBLICATIONS 4,382 CITATIONS
SEE PROFILE
Yi Qian
Wuhan University
214 PUBLICATIONS 4,724 CITATIONS
SEE PROFILE
All content following this page was uploaded by David Tipper on 21 May 2014.
The user has requested enhancement of the downloaded file.

Impact of Signaling Load on the UMTS Call
Blocking/Dropping
Saowaphak Sasanus and David Tipper
Telecommunications Program, School of Information Sciences
University of Pittsburgh
135 N. Bellefield Avenue, Pittsburgh, PA 15260
Email: sasst128, dtipper@pitt.edu
Yi Qian
Advanced Network Technologies Division
National Institute of Standards and Technology
100 Bureau Drive, Stop 8920, Gaithersburg, MD, 20899
Email: yqian@nist.gov
Abstract— Radio resources in the third generation (3G)
wireless cellular networks (WCNs) such as the universal mobile
telecommunications system (UMTS) network is limited in term
of soft capacity. The quality of a signaling service transmission
depends on various factors (i.e., a user’s location, speed, and
data rate requirement), and has impact on quality of user data
communications where the opposite order is also true. In this
paper, we provide the first step to evaluate the impact that
various signaling service types have on call blocking and ongoing
call drop in the UMTS systems. The radio resource’s acquisition
time for various signaling services is calculated according to the
specifications in the UMTS standards. The maximum number
of sessions that a signaling service type can be transmitted
simultaneously is estimated along with the converting value
when the other signaling service type is transmitted instead.
Our analysis reduces the computational complexity in the call
admission control (CAC) and allows the preservation on classes
of services. An example of traffic scenario is given illustrating
the benefit of our study.
I. INTRODUCTION
To increase available capacity, the 3G WCNs adopt wide-
band code division multiple access technology where user-data
and signaling services are transmitted over the same frequency
range. A transmission of a user data traffic can be distinguished
from that of a signaling traffic through orthogonal codes.
However, due to the limitation of the orthogonal codes and
the code allocation algorithm [1], interference becomes the
limit to the radio capacity. In the UMTS networks that use the
frequency division duplex mode, two common types of the
interference are inter- and intra-band interference, and inter-
and intra-cell interference. By assuming that the previous is
insignificance, we consider only the later.
An increase in the heterogeneity of signaling traffic ob-
viously degrades quality of user data communications, and
vice versa. Thus, radio resource must be carefully allocated in
order to preserve quality of service (QoS) in signaling and user
data traffic. Unfortunately, the radio resources is not the only
scarce signaling resource in the cellular networks. Database
servers are also required to support seamless roaming and
secure communications. Thus, to ensure quality of signaling
services, the mechanisms include a CAC and a signaling
overload control must both be in place to maintain quality
of voice and multimedia calls. For global system for mobile
communications (GSM) networks, we proposed a signaling
overload control algorithm that considers the scarcity of both
radio and database server in [2] [3]. Here, we develop the
material that can be used to construct the similar set of
algorithms with the specific attention for the UMTS networks.
In the current literature, only a few simulation based studies
have happened on the impact of some signaling services (i.e,
location update, paging) on user data communications [4] [5].
In this work, we illustrate the impact of most fundamental
signaling services on the cellular communications (e.g., call
setup, location update, and handoff). The available radio
resources are represented by the numbers of sessions that
each of these signaling services called the saturation rate can
be simultaneously supported. The saturation rate is calculated
from the acquisition time that each signaling service needs
to utilize the orthogonal codes in up-link and down-link.
An orthogonal code holding time can be derived from the
transmission rate of the air interface with a choice of either
common or dedicated control channel (CCH or DCH) and
the signaling message length gathering from the signaling
procedures discussed in [6]. We also develop a simple equation
that allows a conversion between the saturated number of
sessions of one signaling service type to that of the other
service type based on a well known signal-to-interference ratio
(SIR) formula [4]. By realizing data traffic demand caused by
servicing various signaling, we can compare the impact that
one signaling service creates to that of the others.
From this analysis, we can efficiently plan radio resource
allocation for various classes of signaling services. In the
future work, we will apply our findings with the CAC of the
UMTS system integrated with signaling overload control at the
database servers, preserving classes of services at both air and
the database servers. The remaining of the paper is organized
as follows. In Section II, the literature on CAC is briefly
reviewed to illustrate the need of the saturated rate estimation
and the basis of the conversion number (i.e., based on the
SIR constraint). Also, the message flow of each signaling
procedure along with its length is depicted. In Section III, the
acquisition time is calculated. Followed by the approximation
of the saturation rate that each signaling service type can
be transmitted simultaneously in the air interface within the
control interval. The analytical model is given for a conversion
between saturation rate of one signaling service type to the

other. The benefit of our analysis is illustrated by an example
of traffic scenario along with the analytical results in Section
IV before we state our conclusions.
II. LITERATURE REVIEW
A. Call Admission Control
A CAC algorithm accepts or rejects the arrival service
requests based on the current system status. We address the
existing CAC algorithms in three perspectives. First is the
method to reject new calls [7] [8]. For example, complete shar-
ing allows all classes of signaling services to share the same
pool of the available radio resources, whereas the threshold-
based CAC restricts services from the lower classes by using
multiple thresholds.
Second is the parameter that represents the status of radio
resources. For example, the interference, the received signal
power, and the SIR. New calls are accepted only if the
maximum or minimum of the parameter is not violated. The
SIR-based CAC more accurately estimates the current system
status compared to the interference-based and the power-based
CACs since it can differentiate between the received signal
power and the interference.
Third is the method to find the available radio resources in
terms of a representative parameter. For example, the interfer-
ence of mobiles within the same cell may be used to estimate
the number of sessions that the available radio resources
sufficiently serve, or the interference of mobiles from the other
cells may also be included into the estimation. However, the
representative parameter are unnecessary in some CACs that
directly apply the parameter into the rejection method. For
example, a CAC that accepts a new call after a test pilot.
The SIR measured within the test pilot is compared with the
minimum SIR to decide the acceptance of a call.
We propose a SIR-based CAC that pre-calculates the max-
imum number of signaling sessions that the current available
radio resources can support within the next control interval.
B. Signaling Procedures
We study the following signaling services: new/end call
request, paging, location update (LU), handover, and SMS. We
illustrate the procedures of these services through the message
flow. The signaling message length is given for the calculation
in the code acquisition time. In the followings, the service
procedures on the originating/terminating side or from/to users
to/from the core network are denoted by the subscript org and
term, respectively.
First, we consider the signaling services that effect the
quality of the active user-data transmission on the up-link
direction (i.e., LU, call setup, handoverorg, and SMSorg). The
user equipment (UE) must perform a general packet radio
service (GPRS) attach, the security related procedures, and the
packet data protocol (PDP) context before sending the data if
any. The GPRS attach allows the system to handle the mobility
management and to obtain the detailed location information.
The PDP context characterizes sessions and assigns the PDP
address for each PDP session. These procedures are illustrated
in Figure 1.
2. RRC connection request complete (26 bytes)
1. UE sends RRC connection request
message over DCH/CCH (10 bytes)
4. Authentication and ciphering request (53 bytes)
UE RNC SGSN HLR
6. Inform HLR to update UE’s locations
with SGSN number and SGSN address
7. Insert subscriber data
8. Validate UE’ RA, MM context,
insert subscriber data ACK to HLR
VLR
5. Authentication and ciphering response (27 bytes)
11. Activate PDP context request (84 bytes)
13. Radio bearer setup (105 bytes)
10. Service accept (e.g., call setup - 11 bytes,
location update - 21 bytes, SMS)
Security
procedures
MM context
(Location update
for first attach)
68-100 bytes
3. Service request (e.g., call setup - 27 bytes, location update - 29 bytes,
SMS - 1 to 100 Kbytes
2. RRC connection setup (139 bytes)
14. Radio bearer setup complete (7 bytes)
15. Activate PDP context accept (31 bytes)
16. Deactivate PDP context request (18 bytes)
17. Deactivate PDP context accept (11 bytes)
19. Radio bearer release request (91 bytes)
20. Radio bearer release complete (7 bytes)
22. RRC connection release complete (2 bytes)
21. RRC connection release (3 bytes)
9. Cancel old MM context, send
location update ACK to SGSN
12. Radio bearer assignment
18. Radio bearer release
Send data traffic
Fig. 1. The GPRS attach and a PDP context [6]
According to [9], these signaling procedures consist of the
following steps. In step 1, the radio resource control (RRC)
connection is established over the CCH. Then, in step 2, the
radio network controller (RNC) sets up a point-to-point radio
connection as well as the signaling connection to the network
before sending acknowledgment back to the UE. After that, the
UE will start the attach process in steps 3−10 which includes
the attach request, the identity request/response for the first
time that the UE is attached to the network, the authentication
request/response if the mobility management context does not
exist for the UE anywhere else. Then, the PDP context will be
setup to characterize the radio bearer (RAB) session and RAB
request is setup in step 11 − 15. The PDP addresses that will
be used and stored at the UE and the GPRS supported nodes
(GSNs) are activated. The PDP context contains mapping and
routing information for packet transmission between the UE
and the gateway GSN (GGSN). After the UE finished data
transmission, the RAB release is initiated along with the PDP
context deactivation and the RRC release in step 16 − 22.
Second, we consider the signaling services that interfere
with the user data communications in the down-link direction
(i.e., paging, handoverterm, and SMSterm). Sometimes, a
SMSterm also needs the paging service if the terminating UE
is in the idle mode. In the UMTS network, the user locations
are tracked in terms of the location area (LA) for the circuit-
switched domain and the routing area (RA) for the packet-
switched domain. A LA consists of multiple RAs. In turn, each

RA consists of multiple UTRAN registration areas (URAs)
each of which consists of multiple cells.
In the packet-switched domain, the UE stays in the idle
mode when a UE does not establish any connection. The UE
locations are tracked with the accuracy in the level of RA. The
UE state is moved to cell-connected only when the connection
is established. If later the UE is inactive longer than timeout,
the UE state is moved to the URA connected and the tracking
accuracy is in the level of URA. If the terminating UE is not in
the RRC cell-connected state, the HLR will be queried for the
availability, the billing information, the available services, and
the last known LA or RA of the UE. Then, the core network
pages all cells within the UE’s LA or RA over the paging
channel (PCH). The larger the location area, the larger the
paging but the smaller load of the location update. After that,
the UE sends the response to the BS in the random access
control channel (RACH), which triggers the BS to assign
the traffic channel to the UE. Then, the RRC connection is
established following with the delivery of the SMS message
(for SMS service).
We note here that PCH, RACH, and another forward access
control channel (FACH) is later referred to as CCH.
III. PERFORMANCE ANALYSIS
A. The Acquisition Time
Most of signaling services can be delivered over either the
CCH or the DCH, leading to the different code acquisition
time. The CCH benefits from fast transmission since it does
not require call setup or tear-down, and the ability to share
code. Also, the interference is introduced only when the
signaling services is transmitted, not in the idle period unlike
in DCH. However, it lacks of fast power control which
anticipates higher interference than CCH. On the other hand,
the DCH allows fast power control, but the interference is
always generated even when channel is idle.
According to the study in [10], the CCH is more suitable to
lower burst size compared to the DCH. More specifically, the
CCH performs better than DCH for a signaling service session
which transmits signaling messages of size approximately up
to 250 bytes. Because the CCH access time is shorter than
the setup time of DCH. In the up-link, the maximum data
rate for the CCH and DCH are 60 kbps and 48 kbps for a
spreading factor of 32. In the down-link, the CCH and DCH
can accommodate the maximum transport channel rate of 36
kbps and 28.8 kbps for a spreading factor of 64.
Table I summarizes the acquisition time which can be
derived from the total message length according to [6], and
the channel data rate. Location update considered here is
the periodic location update where GPRS attach and security
command are not performed. We use the maximum length of
SMS message, 1Kbytes.
B. The Maximum Signaling Service Sessions
In this section, we roughly estimate the maximum amount
of the signaling service sessions that can be conveyed by mean
of a SIR analysis, based on the basic equation adopt from [11].
TABLE I
THE CHANNEL ACQUISITION TIME
Service MSG length Acquisition time
type (bytes) (ms)
DCH CCH DCH CCH
SMS 1180 1000 204.4 133.3
Location update 394 214 81.6 38.6
Call setup 652 472 148.9 88.9
End call 689 500 155.3 93.8
Paging - 9 - 2.0
Inter-RNC Handoff - 17 - 2.71
UE offline 199 45 37.7 36.6
By assuming negligible the interference and noise and the
equal received signal power from all users, signal to noise
ratio (SNR) is S
(N−1)S where N is the total number of users
in the cell and S denotes the received signal power. SIR which
is energy-per-bit to noise power spectral density is S/R
(N−1)S/W
where W is the total radio frequency bandwidth, and R is the
baseband information bit rate.
In this work, we consider arrivals within each control
interval. We assume that only the signaling service type i is
initiated at the beginning of the control interval time between
t − 1 to t. Let SP
be the received signal power of the active
signaling services initiated within the previous control interval
measured at time t which concerns the period of time before
t − 1. The requirement of the SIR for a signaling service type
i, SIRi can be calculated as shown in Eq. 1. Note here that
our analysis here is also applicable for data traffic.
SIRi =
Si/Ri
((1 − α)Iin + Sout + N0)/W
=

W
Ri

Si
Ri(1 − α)[SP + (Ni − 1)]Si + Sout + N0
(1)
α is the orthogonal factor in the down-link and the interfer-
ence reduction scheme in the up-link. There is no synchroniza-
tion among users in the up-link, so there is no orthogonality.
We assume that the transmission in one direction have no
impact to the data rate in the other direction. Only intra-cell
and inter-cell interference is included in the calculation. Iin
and Sout are defined as the interference caused by transmission
of other services within the same cell and within the other
cells, respectively. In fact, Iin is only SP
, and Sout is the
summation of Iin from the neighbor cells. Ni denotes the
maximum number of sessions that signaling service type i can
be supported simultaneously by the available radio resources
within the control interval. Ri be the baseband information bit
rate of the signaling service Si.
The BS can simply monitor the received signal power for
an analysis of the up-link transmission. For the down-link, the
received signal power is calculated according to [12] derived
from the transmitted signal power in Table II and the path
loss model adopted from [13], S = Pt − max(Pl − G, Cl).

S and Pt are the received and transmitted power in dBm.
G denotes the antenna gain in the BS (11dB), and Cl is the
maximum coupling loss (70dB). The path loss denoted by Pl
is 128.1+37.6logr in dB where r is the distance between the
UE and the BS in km.
In the interference limit system such as the UMTS, noise
is negligible compared to the interference, N0 → 0. We can
find Ni as follows.
Ni =
W
Ri(1 − α)SIRi
−
SP
Si
−
Sout
(1 − α)Si
−
N0
RiSi
+ 1
Ni =
W
Ri(1 − α)SIRi
−
SP
Si
−
Sout
(1 − α)Si
+ 1
Ni =
a
Ri
−
b
Si
−
c
Si
+ 1 (2)
where : a =
w
(1 − α)SIRi
, b = SP
, c =
Sout
1 − α
Let Ni be the maximum number of sessions that signaling
service type i alone can be supported by the available radio
resources within the control interval, and Vij indicates the
value that converts Ni to Nj.
Nj = VijNi
a
Rj
−
b
Sj
−
c
Sj
+ 1 = Vij

a
Rj
−
b
Sj
−
c
Sj
+ 1

Vij =

Ri
Rj

Si
Sj

Fj
Fi

(3)
where : Fj = aSj − (b + c)Rj + 1
Fi = aSi − (b + c)Ri + 1
Assume that only Si and Sj exists over the control interval.
From the total available number of sessions Ni, the followings
are derived for the case that X sessions are used by Si and
Ni −X sessions of Si are occupied by Sj. Denote the number
of sessions that Sj can be supported by Ni − X sessions of
Si by Ńj. The conversion value V́ij which maps the number
that signaling service type Si can be supported by the available
radio resource to the number that Sj can be supported is shown
in Eq. 4.
SIRj =
Sj
(1−α)(SP +XSi+((Ni−X)V arij −1)sj )+Sout+N0
Rj/W
(4)
Ńj = Vij(Ni − X) =
a
Rj
−
b + c + XSi
Sj
+ 1
where : a =
w
(1 − α)SIRi
, b = SP
, c =
Sout
1 − α
V́ij =
Ńj
Ni
=
a
Rj
− b+c+XSi
Sj
+ 1
a
Rj
− b+c
Sj
− X + 1
=

Rj
Ri

Sj
Si

Fi − XRiSi
Fj − XRjSi

(5)
where : Fi = aSi − (b + c)Ri + 1
Fj = aSj − (b + c)Rj + 1
By using the induction method, Eq. 5 becomes Eq. 3 when
X = 0. With the similar assumption above, Eq. 5 below is the
general form of Vij where X1, X2, ...XTy signaling service
sessions of S1, S2, ..., STy is transmitted over the control
interval for the total of Ty signaling service types.
Vij =

Rj
Ri

Sj
Si

Fi − fi(Ty)
Fj − fj(Ty)

(6)
where : Fi = aSi − (b + c)Ri + 1
Fj = aSj − (b + c)Rj + 1
fi(Ty) = RiSi(X1 + ... + Xj−1 + Xj+1 + ... − XTy )
fj(Ty) = RjSi(X1 + ... + Xj−1 + Xj+1 + ... − XTy )
From the analysis results, we can promptly plan types
of signaling services and its amount that will be accepted
based on its class at the beginning of the control interval
despite large signaling service types in the near future. At
every control interval (e.g., 1s for signaling services), the
computation complexity is reduced from O(T 2
y ) to O(Ty)
where Ty is the number of signaling service type. For O(T 2
y ),
all N1, N2, ..., NTy must be calculated first before the cal-
culation of V12, V13, ..., V1Ty . Whereas, for O(Ty), only N1
and V12, V13, ..., V1Ty are needed. Signaling service that is
most frequently occurred (e.g., location update) should be
assigned as the signaling service type 1, so the estimation
of the saturated rate or the maximum number of sessions can
be more accurate.
The actual usage of the radio resources can be very
different from the radio resource allocation plan, as user’s
characteristics (e.g., environment, mobility, and interference)
changes over times, especially in large control interval. Thus,
within the control interval, we should adjust radio resource
pool and allocation according to the current user’s status
(e.g., every 0.33s from the total of 1s control interval). The
adjustment period can be adaptively set according to change
in the user’s status. SP
becomes the received signal power of
services within the previous control interval and the signaling
services that are already admitted within the current control
interval in Eq. 1. Because of this adaptability need, using our
formulation will further reduces the computation complexity
in the admission control.
IV. ANALYTICAL RESULTS
We use the example scenario when user either connects with
low speed data 12.2 kbps or high speed data session 64 kbps
after call setup or handoff to new cell. The data rate for CCH
and DCH are set as calculation in the Table I. Other parameters
are set as shown in Table II.A. From both tables, we derive the
maximum number of sessions for various channel rate at the
beginning of the control interval in Table II.B. Low and High
indicates low and high speed data channel. Since the capacity
is limited only by load in the down-link, we perform here only
an analysis for down-link with an assumption that load in the
down-link is higher than that in the up-link. Load in up-link
is only influenced the coverage.

TABLE II
(A) POWER CONTROL PARAMETERS (B) MAX. NO. OF SESSIONS
User data parameters PS
Bit rate(kbps) 12.2 (LOW), 64 (HI)
Spreading gain 32 (UL), 64 (DL)
SIR requirement(dB) 2.5 [4]
BS transmitting power(W) 20 (DCH), 3 (CCH)
Orthogonal factor 0.5
Activity factor 1
Control interval (sec) 1
Ch. max.no.of
Type sessions
CCH 70
DCH 101
Low 305
High 183
In the analysis, only one session of data traffic is initiated for
call setup and handoff. The average message length for each
data session is set to 1Mbytes, which means that the data
session lasts longer than 1s. Table III shows the maximum
number of sessions for some fundamental signaling services
available within the control interval 1s.
TABLE III
THE MAXIMUM NUMBER OF SIGNALING SERVICE SESSIONS (OVER 1S)
Signaling Max. no of sessions
Type CCH DCH
SMS 756 346
Location update 2612 868
Call setup 219 (Low), 179 (High) 213 (Low), 179 (High)
End call 1134 476
Paging 50405 -
Inter-RNC Handoff 301 (Low), 183(High) - (Low), - (High)
UE offine 1878 2754
We illustrate the benefit of our analysis through a small
network consisting of one node B with the arrival signaling
traffic load in the Table IV. Only a low speed data session
will be initiated when a call setup or handoff service is
accepted. Here, we compare between two cases: a simple CAC
which is equipped and not equipped with the knowledge of
the estimated saturated rate in advanced, followed Table III.
The equipped CAC assigns 50%, 35%, and 15% of total
radio resource to high, medium, and low priority classes,
respectively. The unequipped CAC rejects the arrival traffic
only if there is no available radio resource. The table shows
accepted and rejected sessions within an interval time of 1s
when control is performed every 100ms. The results clearly
validate that classes of services can be improved by embedded
our analysis into the simple CAC.
V. CONCLUSIONS AND THE FUTURE WORK
In this paper, we provided a simple analysis to study the
impact of signaling load on call blocking/dropping. The calcu-
lation of the data rate requirement for basic signaling services
are given in both common and dedicated radio channels. Based
on the estimation of the maximum number of users for soft
capacity system discussed in [11], we calculate the simple
conversion that allows fast mapping between the maximum
number of signaling service sessions of one signaling service
type to the other.
Our ongoing work applies the findings of this work to the
CAC that is aware of the database servers’s available resource
for the UMTS networks.
TABLE IV
THE BENEFIT OF OUR DERIVATION
Traffic Arrival Equipped Non-equipped
load (class) session Served Rejected Served Rejected
rate Traffic Traffic Traffic Traffic
SMS (LOW) 40 17.127 22.873 33 7
LU(MED) 150 91.14 58.86 121 29
Call setup(LOW) 10 8.8605 1.1395 9 1
End call(HI) 10 23.8 0 9 1
Paging (MED) 15000 12349.225 2650.775 12001 2999
Inter-RNC HO (HI) 90 82.35 7.65 73 17
UE offline (LOW) 200 136.323 63.677 163 37
ACKNOWLEDGMENTS
This work was supported by a scholarship from TOT
Corporation Public Co. ltd. The author would like to thank
Dr. P. Krishnamurthy for his advises.
REFERENCES
[1] T. Al-Meshhadany and K. A. Agha, “A new code allocation scheme for
UMTS system,” in IEEE Vehicular Technology Conference, 2001. VTC
2001), 2001.
[2] S. Sasanus and D. Tipper, “On the level of guaranteed services for
signaling control in cellular networks,” in IEEE Global Communications
Conference (GC’07), 2007.
[3] ——, “Adaptive multi-class signaling control for cellular networks,” in
IEEE Symposium on Computers and Communications (ISCC’07), 2007.
[4] D. Lin, B. Yeo, and Y. C. Y. Kwok, “Effects of location management
signaling load on the forward link throughput of UMTS-FDD systems,”
in IEEE 16th International Symposium on Personal, Indoor and Mobile
Radio Communications (PIMRC 2005), Sep. 2005.
[5] K. Thum, B. Yeo, Y. Chew, and K. Ang, “Performance study of the
varying parameters on the paging and updating signaling loads in an
UMTS-FDD system,” in Global Telecommunications Conference, 2004.
(GLOBECOM’ 04), Dec. 2004.
[6] ETSI, TR 101 503 v8.27.0. Digital Cellular Telecommunications System
(Phase 2+); Mobile Radio Interface Layer 3 Specification; Radio
Resource Controll (RRC) Protocol.
[7] M. Andersin, Z. Rosberg, and J. Zander, “Soft and safe admission control
in cellular networks,” in IEEE/ACM Transaction on Networking, vol. 5,
no. 2, Apr. 1997, pp. 255–265.
[8] D. Kim, “Efficient interactive call admission control in power-controlled
mobile systems,” IEEE Transactions on Vehicular Technology, vol. 49,
no. 3, pp. 1017–1028, May 2000.
[9] R. Kreher and T. Ruedebusch, UMTS Signaling: UMTS Interfaces,
Protocols, Message Flows and Procedures Analyzed and Explained,
2nd ed. 111 River Street, Hoboken, NJ 07030, USA: John Wiley
Sons, Ltd., 2005.
[10] J. Sachs, S. Wager, and H. Wiemann, “Performance of shared and
dedicated resources in WCDMA,” in IEEE Wireless Communications
and Networking Conference, 2000. (WCNC’00), 2000.
[11] T. S. Rappaport, Wireless Communications: Principles and Practice,
2nd ed. Upper Saddle River, New Jersey: Prentice Hall, 2001.
[12] 3GPP, TR 25.942 v3.3.0. RF System Scenarios.
[13] ETSI, TR 101 112 v3.2.0. Selection procedures for the choice of radio
transmission technologies of the UMTS.
View publication stats
View publication stats

Impact of Signaling Load on the UMTS Call.pdf

More Related Content

Similar to Impact of Signaling Load on the UMTS Call.pdf (20)

Recently uploaded (20)

Impact of Signaling Load on the UMTS Call.pdf