Sei sulla pagina 1di 6

Performance Evaluation for Integrated High Altitude

Platforms-Terrestrial General Packet Radio Service


System.
Sami A. EL-Dolil AND Amir S. EL-Safrawey
Dept. of Electronics and Electrical Comm. Eng., Faculty of Electronic Engineering, Menouf, Egypt

Abstract- General Packet Radio service (GPRS) is a part of the evaluation is estimated considering different radio resources
evolution path towards 3G. This paper investigates the effect of allocation strategies. Different from the previous works, the
using two representative scenarios: a macro cellular system key feature of this work is that: a simulator developed in
operating from a high altitude platform line of sight system and Matlab environment is used to study the effect of using a
a terrestrial non-line-of-sight System, in each case a packet-
macro cellular layer formed from high altitude platforms on
based such as GPRS system is considered. The macro cellular
layer will serve as second server to the mobile data terminals the GPRS system performance. The simulator focuses on the
located in a microcellular layer, both layers form a Multi-layer communication over the radio interface between a MS and
cellular system. A comparison between GPRS system using only BSS, as it is one of the most crucial aspect of GPRS
microcellular structure and when using both microcellular and operation, and mainly determines the performance of any
macro cellular structures is evaluated. A simulator model was wireless network.
developed to study the behavior of data transmission between a
GPRS server and data terminals via air interface. Simulation II. GPRS NETWORK
results show the validity of using macro cells to serve as a The proposed model is a GSM/GPRS system, in which
second layer, and how this can enhance the GPRS system the GPRS subsystem is a new bearer service designed to be
performance. an extension to the GSM network. The GPRS functionalities
was achieved by introducing two GPRS support nodes
I. INTRODUCTION
Migration from the second generation such as the GSM (GSNs), the serving GPRS support node (SGSN) and the
towards the third generation such as UMTS or CDMA2000 gateway GPRS support node (GGSN). The SGSN delivers
is achieved by a change in the inherent switching technology, packets to and from mobile stations within its service area, it
from circuit switching to completely packet switching detects new GPRS MS and queries the home location
system. This sharp switching technology conversion could register to get user profile, and it keeps track of mobile
be done smoothly via a hybrid system such as GPRS, which stations locations. GGSN links different SGSNs and
considered as 2.5 generation. So it is considered as a bridge provides interface between GPRS network and other external
from the voice-centric 2G networks to the data-centric 3G Internet Protocol based networks such as (internet-other
networks and incorporates higher speeds to support data GPRS networks). GPRS greatly improves the operation of
transport. the wireless data communication networks. From the
GPRS is a standard from the European network's point of view, it employs packet switching which
Telecommunication Standards Institute (ETSI) as a new enables operators to use the limited spectrum more
bearer service based on GSM architecture [1, 2]. The rule of efficiently as the radio resources are used only when users
GPRS is to provide a packet switched service over the are actually sending or receiving data so that the system
existing GSM networks. The service providers use the scarce capacity increases. Form the customer's point of view, GPRS
radio resources in a more efficient way, by assigning it on offers much higher data rate than conventional GSM system
demand. Wireless access to external internet protocol-based by applying MSC, thus enables new and better wireless data
network such as internet, provides mixed service capability applications. So the mobile phone will be changed from the
for both real-time and non real-time applications such as traditional one to a multi media wireless access one.
WWW, ftp, email and video conference. GPRS network GPRS gets its radio resources by sharing the GSM
operator provides subscribers with instant, “always on resources. One of the most crucial question is, how to divide
connection”, high data rate with theoretical maximum speed the cell capacity between the traditional GSM and GPRS
up to 171.2 k bit/s by applying Multi Slot Capability (MSC) services. In [7, 8] it is found that a more friendly strategy is,
against 9.6 k bit/s of SMS over GSM networks. Several to reserve Q time slots from the N time slots in each
analytical and simulation models have been proposed to GSM/GPRS TDMA frame for data connection preventing
study the performance of GSM/GPRS. In [3] transmission of their use by voice calls. Consequently guarantee a minimum
data packet in the silent periods of a conversation with voice amount of band width to be shared among all active data
activity detection was proposed. The effect of using dynamic connections. The drawback of this strategy is the increase of
and fixed channel allocation techniques on the GPRS QoS is the number of blocked voice calls due to the reduction in
presented in [4]. In [5] different types of queues for mobility time slots available for voice. In order to guarantee a fixed
management related packets was considered to reduce the voice call blocking probability, it is necessary to
delay time by reducing the competition time of the Routing dimensioning the cell for a reduced offered voice load,
Area Update (RAU). A comparison between using a queue therefore a balance between the reduced number of served
and keeping a dedicated channel for GPRS users is voice calls and the required increase of data connections has
introduced in [6]. In [7-9], GPRS performance and capacity to be performed by the network operator. The remaining N-
Q time slots are shared between GSM and GPRS users with kept stationary for no less than 5 years at altitude ranging
full preemption for the voice call over data connection in this between 17 and 30Km in the stratosphere. The platform is
shared partition. When a voice call request arrived and there approximately 150 meters in length and 40 meters in
is no available channel, the voice preempt a channel used by diameter. Each platform provides instant
data connection and the data terminal (GPRS user – data telecommunications infrastructure for an entire region and
terminal are interchanged) is forced to stop its transmission does not require the deployment of additional, or a
(immediately released) and stored in a queue, waiting for constellation of, stations to provide service. Therefore, each
service. The GPRS user during its data session will send and HAPS constitutes a stand-alone regional system that can be
receive data, that must be converted into bursts. Each burst individually deployed. The deployment of 250 platforms
data bits is convoyed by a time slot each frame period. The over the 250 most populous cities will create a world wide
conversion of user's data to burst carried out through system covering more than 80% of the world's population.
multiple levels, at first the data is segmented into packets, The most promising telecommunication applications of
each packet is split into Logical Link Control (LLC) frames, HAPs are cellular telephony, broadband local multipoint
each LLC is in turn split into a number of Radio Link distribution system (LMDS) services, and access provision
Control (RLC) blocks as shown in figure (1). The number of to digital networks (Internet, ISDN). The major advantages
data bits contained in each RLC block is 456 coded bits for future services integrating HAPs based systems are the
arranged in four consecutive radio bursts. As the size of the cost-effective coverage of rural and maritime regions, the
RLC block is kept constant, the Coding Schemes (CS1-CS4) system flexibility due to the platforms’ mobility on demand,
determine the a mount of the useful payload convoyed by and the possibility of upgrading the platform’s payload in
each RLC block, the choice of CS controlled by the carrier order to reduce the risk of technology obsolescence
to interference ratio (CIR), it's a crucial parameter in any experienced with traditional satellites. In addition, HAPs can
wireless network. The RLC block is the Packet Data have a large footprint on earth offering the possibility of
Channel (PDCH) basic transmission unit, each RLC block providing services over large regions with low expected
requires four time slots in four consecutive time division traffic cost effectively.[11-13]. The platforms can be linked
multiple access frames, hence the RLC block is the unit of directly to one another by hop stations located midway
transmission. Some previous studies made their calculations between the platforms or by inter- platform links, and can
considering the RLC block level rather packet level [10]. also be linked indirectly via satellite or the PSTN. [14 ].
IV. SIMULATION MODEL AND CHANNEL ASSUMPTION
User
data This paper focuses on the performance of GPRS data
service considering the use of HAP combined with the
Packets micro cellular structure forming a Multi-layer cellular
LLC system. Both structures have a shape of a cluster of seven
frame hexagonal cells, each cell in the macro cell layer is
20 ms represented by high altitude plat form, will cover seven
RLC
clusters from the terrestrial system as shown in figure (2).
blocks
The key feature of this work is to evaluate the effect of HAP
Burst to serve as a second server for the mobile data terminal, and
how thus affect on the GPRS system performance, the
Figure (1): User data to bursts transformation. evaluation done by introducing a comparison between the
GPRS system when use terrestrial structure only and when
III. HAP OVERVIEW use both terrestrial and HAP structures. The comparison
High Altitude Platform Stations (HAPs) or High Altitude done with respect to different quality of service parameters
Very long Endurance (HAVE) vehicles. is a novel such as; the failure probability of new and handoff attempts
telecommunications network, which employs a proprietary of data terminals, the RLC block delay and throughput
stratospheric platform technology to provide broadband (channel utilization by GPRS load). Adaptive CS according
wireless communications, services around the world. HAPS to CIR was used to compensate the environmental effects,
is considered to be a revolutionary opportunity for wireless provide a suitable protection and guarantee BLER less than
industry, systems with several economic and technological 0.2 [15]. CIR is a crucial parameter in today’s network, as
advantages over both space- or ground- based counterparts the capacity of any wireless network is interference limited.
which will offer fixed, or transportable. This is sufficient to
support such multimedia services as video-telephony, low
cost video-conferencing, and high speed internet
connectivity. The services will include user authentication,
location registration, link encryption, and bandwidth-on-
demand. The build-in security measure provides wire-line
equivalent privacy (WEP) to insure high degree of privacy of
communications. The stratospheric platform is a super-
pressurized helium-filled solar/fuel- cell powered airship
with enough payload and power capacity to support all of a
metropolitan area's broadband communications requirements.
Using propulsion and thermal control the platform will be Figure (2): HAP and terrestrial systems structure
A- CHANNEL MODELS The interference between two co-channel cells separated by
The first system considered is a terrestrial non-line-of- distance d terr is.
sight system. By considering the environmental effects such
as the radio propagation loss and shadowing, both will affect I terr = p tx − terr − L ( d terr ) + λ
the Carrier to Interference Ratio (CIR) between co-channel where λ is the shadowing fading term which is a
cells. There are many radio propagation models, the path random variable following normal distribution with zero
loss Okumura and Hata model was selected to calculate the mean and eight standard deviation, this random is range
propagation path loss, because it's more practical and has limited between, [18]
intermediate results within others models [16], the path loss − 4σ ≤ λ ≥ 2σ
Okumura-Hata model given by [17] where, σ is the standard deviation.
L ( d terr ) = 69 . 55 + 26 . 16 log [ f c (MHz )] Considering the offered traffic per interfered channel is
− 13 . 82 log [h b ] − a (h m ) assumed to be constant and equal to 0.4 Erlang, and a
+ [44 . 9 − 6 . 55 log (h b )]log (d terr ) constant transmitter power Ptx-terr = 10 wat [19], standard
deviation σ = 8 [18], the frequency of operation f c = 895
where:
a (h m ) = [1 . 1 log ( f c ) − 0 . 7 ]h m MHz, the cell radius R terr = 1 Km .
− [1 . 56 log ( f c ) − 0 . 8 ] The number of co-channel interferers cells considered
here are the first two interferers tires, so there are at
fc is the carrier frequency in MHz.
maximum eighteen active interferes and nineteen
d terr is the distance in Km between the mobile terminal interference states (0 - 18). The total interference strength in
and the interferer base stations as shown in figure (3), dB, contributed by these interferers is:
n =18
the distance calculated from the following relation:
2 2
It terr dB = 10 log 10 ∑P n − terr ⋅ I n − terr wat
d terr = ri + rm − 2 ri rm cos( θ ) n=0
Where Pn-terr is the probability of n active interferers from
where ri rm are the distances from the originating base nineteen states (0 - 18) that can be obtained from the
station to the interferer base station and mobile terminal interfering traffic using the Erlang formula. Assume the
respectively as shown in figure (4). offered traffic per interfering channel equals 0.4 Erlang, so
h b is the BS antenna height (m) (effective). the total interfering traffic Aterr is 7.2 Erlang, then Pn-terr
equals:
h m is the MU antenna height (m) (above ground). A
n
/ n!
P n − terr = i = 18 terr
a (h m ) is the correction factor.
∑ A terr / i !
i

i= 0

I n−terr represents the interference power in watts for n


wat

active interferers.
The average carrier strength is calculated as;
C ( d MS − BSS ) = Ptx − terr − L ( d MS − BSS )
where; RMS-BSS is the average intra cell MS-BSS distance
that can be approximated by the radius of a circle covering
half of the hexagonal cell area, [20] given by:
AHex 1.5 ∗ 3 R 2
d MS − BSS = = = 643 m
2π 2π
Where L (dMS-BSS) is the propagation loss of the intra cell
Figure (4): Distances from mobile terminal to the
interferer and originating base stations.
MS-BSS distance dMS-BSS. The mean carrier to interference
ratio in terrestrial system is CIRterr :
CIR terr = C ( d MS − BSS ) − It terr dB

The second system is a high altitude platform line-of-


sight system, in this case only the radio propagation loss is
considered, thus will affect the total Carrier to Interference
Ratio CIR. The propagation model used in this case is an
extended version of the Hata’s model for cellular system and
free space loss [21]:
L ( d ) = 32 . 4 + 10 ∗ log 10 [ f c ( MHz ) ∗ d ( Km ) ]
The interference between two co-channel cells separated by
distance d hapi is.
I hap = p tx − hap − L ( d hapi )
Figure (5): Distances from mobile terminal to the
interferer and originating HAPs.
The distance d hapi in this case is the distance between scheme as a function of the CIR. The value of CIR defines
the mobile terminal and the interferers HAP as shown in the coding scheme that maximizes the throughput and
figure (5), the distance calculated from the following keep BLER < 0.2.
relation: C- SIMULATOR MODEL
2 2 In this paper a simulator of GPRS at the radio link
d hapi = 17 + rmhap + 4 R hap − 4 R hap rmhap cos( φ )
2
control level has been carried out and developed using
where rmhap is the distance from the center of the coverage Matlab environment. The simulator focuses on the
area covered by the originating HAP to the mobile terminal. communication over the radio interface between MS and
By considering the offered traffic per interfered channel is BSS (NLOS) or HAP (LOS) systems, because this is one of
assumed to be constant and equal to 0.4 Erlang, and a the most crucial aspects of GPRS operation. It considers a
constant transmitter power Ptx-hap = 1 wat [19], the frequency two dimension GSM/GPRS network of microcellular and
macro cellular layers, both layers are partitioned into seven
of operation f c = 1950 MHz [21], the cell radius hexagonal cells, each cell of macro cellular layer cover seven
R hap ≈ 7 R terr . clusters of microcellular layer. To allow for a finite number
The number of co-channel interferers HAPs considered of cells to be simulated while still approximately an infinite
here are the first interferers tire, so there are at maximum six system, a warp-around technique is used. This means that
active interferes and seven interference states (0-6). The total opposite sides warp-around, so that the finite size effect is
interference strength in dB, contributed by these interferers eliminated. The simulator describes the behavior of a single
is: carrier with eight time slots per cell GSM/GPRS system. In
n =6 microcellular layer, each cell has one channel (time slot) is
It hap dB = 10 log 10 ∑ Pn − hap ⋅ I n − hap wat reserved for control signaling, another channel is reserved
n =0 for GPRS data sessions, the remainder six channels are
Where Pn-hap is the probability of n active interferers from shared by both circuit switched GSM services and GPRS
seven states (0 - 6) that can be obtained as above, hence the data sessions. In macro cellular layer, each cell has one
total interfering traffic is Ahap = 2.4 Erlang, then Pn-hap equals: channel (time slot) is reserved for control signaling, the
n
A hap / n ! reminder seven channels are shared between the seven
P n − hap = i = 6 clusters in the microcellular layer.
∑ A hap / i!
i
D- TRAFFIC MODEL
i= 0
Traffic models for future broadband systems are the
I n−hap wat represents the interference power in watts for n subject of much current research. Here, a simple model is
active interferers HAP. used, whereby the arrival of voice calls and data sessions are
The average carrier strength is calculated as; modeled according to two mutually independent poisson
C ( d hap ) = Ptx − hap − L ( d hap ) processes, with negatively exponentially distributed inter-
arrival times. both of them are scheduled to share the radio
where; d hap is the distance between the mobile terminal resources. This is considered to represent a typical voice call
and the servant HAP as shown in figure (5) and data session, where the radio link is the limiting factor
2
on the speed of the downlink, and congestion control
d hap = 17 2 + rmhap techniques such as an access queue is employed. When the
where L ( d hap ) is the free space propagation loss base station is presented with a burst of packets to transmit
to a user, it selects one or more channels and coding scheme
between the servant HAP and mobile terminal. The mean to use according to the current CIR, and attempts to transmit
carrier to interference ratio in the HAP system is CIRhap: the packet, if there is no available resource in the terrestrial
CIR hap = C ( d hap ) − It hap dB layer, it will try to access HAP layer, if there is available
resources, it will be served, if there is no resources, it will be
B- CODING SCHEME CHOICE stored in the access queue waiting for a finite time. During
The CIR is obtained for each data terminal. It depends on this time if resources will be available in any layer, it will
the current location and the current operating system, both served, if not it will be cleared from the system due to buffer
will assign the propagation loss and shadow fading term. time expiring. When a new voice call request arrives, it takes
The value of CIR will control the choice of the coding a free channel, if there is no channel available and the
scheme used in GPRS. There are four coding scheme are number of voice calls in the service is below six, hence one
considered, during the user mobility the CIR is varied, thus of GPRS data sessions stops its transmission in order to
will control the choice of the appropriate coding scheme allocate a channel to the new voice calls. The interrupted
according to the following table: data session will try to access the corresponding cell in the
Table 1: Switching between CS's depending on CIR value macro cellular layer, if there is no available radio resources it
CIR ≤ CIR ≤ CIR ≤ CIR is stored in the suspend queue without limitation on the
CIR storing time. When resources are available again in any
5.9586 10.1315 16.5037 >16.5037
Coding layer, the data sessions in the suspend queue have higher
CS1 CS2 CS3 CS4 priority to be resources allocated than the new data session
Scheme
or data session in the access queue. The discipline in the
The switching between these CS's was done according to access and suspend queues follow the First In First Out
[10,22], which present the optimal choice of the coding (FIFO) principle which is the simplest scheduling and
queuing method. The average service time of circuit figure (9), this ensure that all the simulation programs
switched calls is assumed to be exponentially distributed operate under the same offered traffic and thus authenticate
with mean 120 seconds. The voice traffic load is set to 2.275 the reliability of the simulator. The data sessions served in
Erlang to satisfy the general requirements of 2% blocking the system per second and the percentage of data sessions
probability which is a typical target value for GSM completed are illustrated in figures (10,11) respectively, both
operators. figures prove the gain of using macro cellular layer. The
In this paper the Email traffic model is used, that has the normalized channel occupation time (throughput) is
following characteristics: presented in figure (12), it is clear that at low and moderate
• Calls and data sessions arrived according to poisson traffic, the macro cellular layer cause the radio resources to
processes. be free, while at high traffic, the macro cellular layer carry
• Adapted coding scheme according to the carrier to the traffic which should be blocked if there is no macro
interference ratio. cellular layer, and thus will increase the channel throughput.
• Interferences from the first two tires in the terrestrial VI. CONCLUSION
system are considered.
• Interferences from the first tire in the HAP system are In this paper, the effect of using a HAP LOS system to
considered. serve as a second servant for the mobile data terminal in a
• The macro cell's base station mounted on a high altitude micro cellular GPRS system is investigated. Results were
platform. obtained by simulating GSM/GPRS Multi-layer cellular
• Single carrier, eight time slots per each microcellular cell.
network air interface and Email traffic model. The
contribution of this paper is that, first; the proposed multi-
• Single carrier, eight time slots per each macro cellular cell
layers structure is sufficiently to enhance the GPRS system
(single HAP).
with the lowest infrastructure requirements, second; the use
• The length of the Email size is exponentially distributed
of macro cellular layer using high altitude platform will
with a mean of 320 K bit, [22]. enhance the quality of service parameters of the GPRS
• Average service time of 320000 /13400 second. system such that, blocking probability of new data sessions,
• The data terminal can operate in MSC mode, up to seven handoff failure of data sessions, RLC block delay, data
time slots were used simultaneously. sessions served in the system and the throughput.
• Mobile stations are assumed to be moving with a uniform
truncated speed between maximum value of 60 Km/h REFERENCES
(vehicles) and minimum speed of 5 Km/h (pedestrian). [1] ETSI, “Digital cellular telecommunications system (phase 2+) General Packet
Radio Service, (GPRS) service description; stage 2 (3 GPP TS 03.60 version
• Access queue of length 20 and maximum storing time of 2 7.9.0 Release 1998), ”.
minutes. [2] G. Brasche and B. Walke, “Concepts, services and protocols of the new GSM
• Suspend queue with infinite length and storing time. phase 2+, General Packet Radio Service, ” IEEE Commun. Mag, pp. 94-104,
Aug. 1997.
• Micro cell radius set to be 1Km. [3] J. H. Hung, S. L. Su and J. H. Chen, “Design and performance analysis for data
transmission in GSM/GPRS system with voice activity detection, ” IEEE Trans.
• Macro cell radius set to be 7 Km. Veh. Technology vol. 51, pp. 648-656, July 2002.
• The queuing discipline is FIFO. [4] M. D’Arienzo, A. Pescapè and G. Ventre, “A simulation environment for
analysis experimental analysis of multiple traffic scheduling algorithms over
• The average service time of GPRS data session is not GPRS networks,” CCCT, July 2003.
small, hence during the mobility of the data terminals, [5] K. Chew and R. Tafazolli, “Performance analysis for GPRS with prioritized and
handoffs are required for these terminals to complete their non-prioritized mobility management procedures,” IEE Third International
Conference on 3G Mobile Telecommunication Technologies, London, United
data sessions. Kingdom, pp 544-549, 8-10 th May 2002.
• The handoff computed only for users served in the [6] M. Ghanderi and R. Bautaba, “Data service performance analysis in GPRS
systems,” Proceedings of the 15th IEEE international Symposium on Personal,
terrestrial system, while it is neglected in the HAP system. Indoor and mobile radio communications (PRIMC'2004), Barcelona, Spain,
During the data terminal movement within the NLOS cell September 2004.
[7] R. G. Garroppo, S. Giordano and S. Lucetti, “Capacity evaluation of resource
area, it suffers both propagation loss and shadow fading due allocation strategies in GPRS system,” Academic OPNET Research and
to terrain features and building density, while in the LOS Educational Projects, University of Pisa, 2002.
coverage area only the propagation loss is considered. [8] M. Ermal, K. Müller, J. Schüler and M. Schweigel, “Analytical comparison of
different GPRS introduction strategies, ” Proc. 3rd ACM Int. Workshop on
Modeling analysis and simulation of wireless and mobile systems, pp. 3-10,
V. SIMULATION RESULTS Boston, MA, Aug, 2000.
[9] W. Chen, J. C. Wu and H. Liu, “Performance analysis of radio resource
The developed simulator was intensively utilized to allocation in GSM/GPRS networks, ”Proceedings of IEEE VTC 2002-Fall, vol.
investigate the behavior of the Email traffic model over 3, pp. 1461-1465, Sept. 2002.
[10] O. Queseth, F. Gessler and M. Frodigh, “Algorithms for link adaptation in
GPRS system. The simulation results show that the use of a GPRS ” IEEE Proc. VTC Houston., May, 1999.
practical GPRS system with added HAP layer will improve [11] S. Karapantazis, F.-N. Pavlidou, “Broadband Communications via High Altitude
the QoS parameters of GPRS system. Figure (6) shows that Platforms: A Survey”, IEEE Communications Surveysand Tutorials, first
Quarter 2005 issue, vol. 7, no. 1, pp. 2-31.
the use of macro cellular layer will reduce the blocking [12] F. Dovis, R. Fantini. M. Mondin, P. Savi, "Small-Scale Fading for High Altitude
probability of new data sessions, for example at offered platform (HAP) propagation channels", IEEE Journal on selected areain
communications, Vol. 20, No. 3, April 2002.
traffic of 14 Erlang the blocking probability reduced from [13] G. Kandus, A. Svigelj, M. Mohorcic, “Telecommunication Network over High
0.3 to about 0.06. The handoff failure probability of data Altitude Platforms”. 7th International Conference on Telecommunications in
Modern Satellite, Cable and Broadband Services, TELSIKS 2005, Nis, Serbia
sessions shown in figure (7), the use of HAP layer made an nad Montenegro, 28-30 Sept. 2005, Proceedings, Vol. 2, pp. 344-347.
enhancement by about 37% at offered traffic at 14 Erlang. [14] Yee Chun LEE, "Radio aspects of a stratospheric broadband wireless system",
Figure (8) show the RLC block delay, there is a reduction in IEEE Radio and Wireless Conference 1999, RAWCON 99, pp.: 1-2.
[15] R. Kalden, I. Meirick and M. Meyer, “Wireless internet access based on
the RLC block delay after using the macro cellular layer. GPRS, ” IEEE Commun. Mag, pp. 8-18, April, 2000.
The data sessions enter the system per second is presented in
[16] S. El-Dolil and A. El Safrawey “General Packet Radio Service Performance
Evaluation Based on CIR Calculation, Considering Different Radio Propagation
Models” proceeding of the Twenty fourth National Radio Science Conference
(24th NRSC’ 2007), Cairo, Egypt, 13-15 March 2007.
[17] W. Webb, “The complete wireless communications professional“, Artech House,
Norwood, MA, 1999.
[18] R.Steele, C.C.Lee, P.gould, “GSM cdmaOne and 3G systems“, Wiley, 2001.
[19] D.A.J. Pearce, D. Grace, "Optimising the downlink capacity of broadband fixed
wireless access systems for packet-based communications", IEEE International
Conference on Communications, ICC 03, Vol. 3, May 2003 pp.: 2149-2153.
[20] U. Fornefeld and B. Walke, “On the CIR gain from location diversity in GPRS
networks,” Proceedings of the European wireless, vol. 0, pp. 898-903, Feb.
2002.
[21] Jong-Min Park , Dae-Sub Oh, Yang-Su Kim, and Do-Seob Ahn, "Evaluation of
interference effect into cellular system from high altitude platform station to
provide IMT-2000 service" IEEE Global Telecommunications Conference 2003,
GLOBECOM 03, Vol.1, Dec. 2003, Pp.: 420-424.
[22] P. J. A. Gutiérrez, J. Wigard, P. N. Anderson, “Performance of link adaptation in
GPRS networks,” Proceedings of IEEE, VTC'2000-fall, vol. 2, pp. 492-499,
2000.
[23] R. Litjens and R. J. Boucherie, “Elastic calls in an integrated services network: Figure (9): The data sessions enter the system per
the greater cell size variability the better the QoS, Performance evaluation”, vol.
52, no. 4, pp. 193-220, 2003
second.

Figure (6): The blocking probability. Figure (10): The data sessions served in the system per
second.

Figure (7): The handoff failure probability.


Figure (11): percentage of data sessions completed.

Figure (8): The RLC block delay.


Figure (12): The channel occupation time (throughput).

Potrebbero piacerti anche