**6. Conclusions and future QoS testbed extensions**

Coming back to the tests in Sections 2, 3, PPP tunnels between the server and users can be temporarily closed, when packet transimission is slowed down by interference or machinery stops.

The first problem is that, in case the PPP LCP surveys trouble situations, the channel is closed and the client disconnected. An authomatic procedure is in charge of reconnection, but a time waste in PPP tunnel setup as well as abrupt disconnections are bound to take place.

A second problem derives from traffic limitation and control being handled by a single QoS server: in this case, data are properly limited only after they have crossed one or two links. In other words, in case an authenticated user sends an UDP data flow larger than his or her maximum upload bandwidth, such flow will be diminished only after reaching the QoS server. Meanwhile, the available bandwidth will be unproperly occupied by such flow.

On the basis of such considerations, the testbed will be extended according to two different scenarios of distributed QoS architectures [26-28]. The first one is depicted in Fig. 14 and aims at avoiding tunnel closure in case of interference and packet loss.

In case many relays occur between the client and PPPoE concentrator, packet loss can increase; the idea, thus, is to shorten the tunnel, so as to integrate the PPPoE concentrator and the transmitter. In this way, all PPP features could be maintained and its limitations diminished. The tunnel, in fact, would be established between the client's CPE and the nearest transmitter and communication between the transmitter and the main server could be based on TCP/IP.

Furthermore, if interferences between transmitter and main server would take place, packets could be relayed without PPP tunnel drops.

A disadvantage could concern uncoded communication between the main server and pylons. Possible solutions could be the activation of encrypted systems or a PPP tunnel to the main server. In this case, the user would not even perceive any link failure.

A Testbed About Priority-Based Dynamic

configuration changes.

from an exceeding traffic in case of network expansion.

needed that prevents connections from being denied.

the joint use of the actually developed parts and simulators.

Fig. 16. A schema for the simulation of QoS server impact.

Connection Profiles in QoS Wireless Multimedia Networks 95

As for the PPPoE concentrator, the QoS manager itself could be integrated in the transmitter. On the one hand, this kind of control logic decentralization would solve the problem of link saturation in case of heavy UDP uploads. On the other hand, the server would be spared

Two further difficulties arise: firstly, connection plans are not anymore managed by a single server in a centralized and transparent way. In consequence, a new communication protocol is required for the authomatic configuration of devices on the pylon when the main server

In addition, logic decentralization can cause more frequent failures of important components. If a failure occurs of QoS or PPPoE components, thus, an infrastructure is

A switch, for instance, could be used to disconnect out of order devices and the main server

At present, the idea is to avoid a complete implementation of the above scenarios, using simulation tools instead. In particular (Fig. 16), the QoS impact could be evaluated through

would be in charge of guaranteeing connectivity until the problem is solved.

Fig. 14. first extended testbed scenario.

The second scenario (Fig. 15) aims at solving the second problem arisen: the idea is to apply the first control on users' bandwidth at the pylon.

Fig. 15. second extended testbed scenario.

The second scenario (Fig. 15) aims at solving the second problem arisen: the idea is to apply

Fig. 14. first extended testbed scenario.

Fig. 15. second extended testbed scenario.

the first control on users' bandwidth at the pylon.

As for the PPPoE concentrator, the QoS manager itself could be integrated in the transmitter. On the one hand, this kind of control logic decentralization would solve the problem of link saturation in case of heavy UDP uploads. On the other hand, the server would be spared from an exceeding traffic in case of network expansion.

Two further difficulties arise: firstly, connection plans are not anymore managed by a single server in a centralized and transparent way. In consequence, a new communication protocol is required for the authomatic configuration of devices on the pylon when the main server configuration changes.

In addition, logic decentralization can cause more frequent failures of important components. If a failure occurs of QoS or PPPoE components, thus, an infrastructure is needed that prevents connections from being denied.

A switch, for instance, could be used to disconnect out of order devices and the main server would be in charge of guaranteeing connectivity until the problem is solved.

At present, the idea is to avoid a complete implementation of the above scenarios, using simulation tools instead. In particular (Fig. 16), the QoS impact could be evaluated through the joint use of the actually developed parts and simulators.

Fig. 16. A schema for the simulation of QoS server impact.

A Testbed About Priority-Based Dynamic

10.1109/ISCC.2003.1214112

10.1109/ICME.2008.4607463

264790

and Networking, 2010, Article ID 168357

51-60, doi: 10.1049/iet-com.2009.0694

doi: 10.1109/ICNICONSMCL.2006.62

http://esa.un.org/unup/ [24] EU, Eu mobility and transport, 2010,

Connection Profiles in QoS Wireless Multimedia Networks 97

[12] Yaya Wei; Chuang Lin; Fengyuan Ren et al; Dynamic priority handoff scheme in

[13] Xiaorong Li; Chuah, E.; Jo Yew Tham; Kwong Huang Goh; An optimal smooth QoS

[14] Huang, J.-H.; Li-Chun Wang; Chung-Ju Chang; Capacity and QoS for a Scalable Ring-

Communications, 24, 11, 2070-2080, 2006, doi: 10.1109/JSAC.2006.881622 [15] Bai, B; Chen, W.; Cao, Z. et al; Uplink Cross-Layer Scheduling with Differential QoS

[16] Montazeri, S.; Fathy, M.; Berangi, R.; An Adaptive Fair-Distributed Scheduling

[17] Almeida, M.; Sarrô, R.; Barraca, J.P. et al; Experimental Evaluation of the Usage of Ad

[18] Ganesh Babu, T.V.J.; Le-Ngoc, T.; Hayes, J.F.; Performance of a priority-based dynamic

[21] Zhang, F.; Verma, P.K.; Cheng, S.; Pricing, resource allocation and quality of service in

[22] Kamosny, D.; Novotyny, V.; Balik, M.; Bandwidth Redistribution Algorithm for Single

[23] UN, World urbanization prospects: The 2007 revision population database, 2008,

Areas in Communications, 19, 2, 355-369, 2001, doi: 10.1109/49.914513 [19] Naser, H.; Mouftah, H.T.; A class-of-service oriented packet scheduling (COPS)

Communications and Networking, 2007, Article ID 62967

and Systems, 1645-1652, 2006, doi: 10.1109/CISS.2006.286399

http://ec.europa.eu/transport/publications/statistics/

[25] ecoage, Independent ecology portal, 2003, www.ecoage.net

differentiated QoS wireless multimedia networks, in Proc. of Eighth IEEE International Symposium on Computers and Communication, 131-136, 2003, doi:

adaptation strategy for QoS differentiated scalable media streaming, in Proc. of IEEE International Conference on Multimedia and Expo, 429-432, 2008, doi:

Based Wireless Mesh Network, IEEE Journal on Selected Areas in

Requirements in OFDMA Systems, EURASIP Journal on Wireless Communications

Algorithm to Guarantee QoS for Both VBR and CBR Video Traffics on IEEE 802.11e WLANs, EURASIP Journal on Advances in Signal Processing, 2008, Article ID

Hoc Networks as Stubs for Multiservice Networks, EURASIP Journal on Wireless

capacity allocation scheme for wireless ATM systems, IEEE Journal on Selected

algorithm for EPON-based access networks, in Proc. of 7th Int. Conference on Transparent Optical Networks, 232-236, 2005, doi: 10.1109/ICTON.2005.1505793 [20] Song, S.; Manikopoulos, C.N.; A Priority-based Feedback Flow Control System for

Bandwidth Control, in Proc. of 40th Annual Conference on Information Sciences

multi-class networks with competitive market model, Communications, IET, 5, 1,

Source Multicast Networking, in Proc. of Int. Conference on Systems and Int. Conference on Mobile Communications and Learning Technologies, 147-156, 2006,

### **7. Acknowledgment**

More than an acknowledgment, a dedication: To the little Gabriele Toppan, the son of Paolo and the nephew of Andrea, go our very best wishes to grow up strong, responsible and enthusiastic about life and its numerous miracles.

#### **8. References**


More than an acknowledgment, a dedication: To the little Gabriele Toppan, the son of Paolo and the nephew of Andrea, go our very best wishes to grow up strong, responsible and

[1] Gringeri, S.; Shuaib, K.; Egorov, R. et al.; Traffic shaping, bandwidth allocation, and

[2] Frank Yong Li; Stol, N.; QoS provisioning using traffic shaping and policing in 3rd-

[4] Gozalvez, D.; Monserrat, J.F.; Calabuig, D., et al; Policy-based channel access mechanism

[5] Flegkas, P.; Trimintzios, P.; Pavlou, G.; A policy-based quality of service management

[6] Fangming Zhao; Lingge Jiang; Chen He; Policy-based radio resource allocation for

[7] Conchon, E.; Pérennou, T.; Garcia, J. Et al.; W-NINE: A Two-Stage Emulation Platform

[8] Heithecker, S.; do Carmo Lucas, A.; Ernst, R.; A High-End Real time Digital Film

[9] Chang Wook Ahn; Ramakrishna, R.S.; QoS provisioning dynamic connection-admission

[10] Xiang Chen; Bin Li; Yuguang Fang; A dynamic multiple-threshold bandwidth

[11] Huan Chen; Kumar, S.; Kuo, C.J.; Dynamic call admission control scheme for QoS

Network, IEEE, 12 , 6, pp. 94 – 107, 1998, doi: 10.1109/65.752648

quality assessment for MPEG video distribution over broadband networks,

generation wireless networks, in Proc. of IEEE Wireless Communications and Networking Conference, 2002, 1, 139 – 143, 2002, doi: 10.1109/WCNC.2002.993478 [3] Yongdong Wang; Jurczyk, M.; Impact of traffic shaping in ATM networks on video

quality, in Proc. of International Workshops on Parallel Processing, 1, 485 – 492,

selection for QoS provision in IEEE 802.11e, Vehicular Technology Magazine, IEEE,

system for IP DiffServ networks, Network, IEEE, 16, 2, 50-56, 2002, doi:

wireless mobile networks, in Proc. of IEEE International Conference on Neural Networks and Signal Processing, 476-481, 2008, doi: 10.1109/ICNNSP.2008.4590396

for Mobile and Wireless Systems, EURASIP Journal on Wireless Communications

Processing Reconfigurable Platform, EURASIP Journal on Embedded Systems,

control for multimedia wireless networks using a Hopfield neural network, Vehicular Technology, IEEE Transactions on, 53, 1, 106-117, 2004,

reservation (DMTBR) scheme for QoS provisioning in multimedia wireless networks, Wireless Communications, IEEE Transactions on, 4, 2, 583-592, 2005, doi:

priority handoff in multimedia cellular systems, in Proc. of IEEE Wireless Communications and Networking Conference, 114-118, 2002, doi:

**7. Acknowledgment** 

**8. References** 

enthusiastic about life and its numerous miracles.

2000, doi: 10.1109/ICPPW.2000.869154

and Networking, 2010, Article ID 149075

10.1109/65.993223

2007, Article ID 85318

doi:10.1109/TVT.2003.822000

10.1109/TWC.2004.843053

10.1109/WCNC.2002.993474

2, 3, 29-34, 2007, doi: 10.1109/MVT.2008.915326


**0**

**5**

Wei Zhuang

*P.R.China*

*China Telecom Co. Ltd. (Shanghai)*

**End to End Quality of Service in UMTS Systems**

About ten years ago, WCDMA <sup>1</sup> based the third generation mobile systems started to be deployed worldwide. Besides to support basic mobile data services such as file transfer and internet surfer, etc., UMTS has one of the most significant archievements which can support a richer variety of services with QoS guarantee, such as video, VOIP, etc. Quality of Service (QoS) is defined as "the collective effect of service" performance, which determines the degree of satisfaction of a user of the service in the ITU-T recommendation E.800. At a technical level, QoS can be characterized by service availability, delay, jitter, throughput, packet loss

3GPP has put many efforts to define and standardize a QoS framework for data services, specially IP-based services. The standardization of a UMTS QoS model started in 1999. the development was based on the following key principles: operation and QoS provisioning needed to be possible in the wireless environment, usage of the Internet QoS mechanisms, applications and interoperability. This chapter is aimed to provide an overview of the UMTS end-to-end QoS architecture, describe how the QoS requirements to be realized from top layer

QoS standardization in UMTS PS domain enables UMTS to provide data service with end-to-end QoS guarantees. 3GPP proposed a layered architecture for supporting end-to-end

• Mapping of end-to-end services provided by the UE, UTRAN, Core Network (CN), and

<sup>1</sup> Wideband Code Division Multiple Access W-CDMA - the radio technology of UMTS - is a part of the

QoS. It includes the following key elements (Sudhir Dixit et al., 2001):

**1. Introduction**

to wireless links.

rate (Nortel White Paper, 2002).

**2. WCDMA QoS architecture**

external IP networks;

• Location of QoS functions;

• QoS negotiation;

• Traffic classes and associated QoS parameters;

• Multiplexing of flows onto network resources;

• An end-to-end data delivery model.

ITU IMT-2000 family of 3G Standards.

