In todays information age, telecommunication technology has the ability to access everything, everywhere, all the time. The cellular networks fulfills this dream by aggregating of all kind of traffic such as voice, multimedia data transfer etc. in the same transport technology.
Due to the unprecedented demand of multiple services such as voice, data and other multimedia applications over cellular networks, there is a need to provide Quality of Service (QoS) in cellular networks. If QoS is not administered at the cellular network, a cellular or mobile data communication system cannot protect itself when too many undifferentiated users try and access the cellular network for data services. For example, if scores of teenagers are downloading heavy bandwidth applications (MP3 files) simultaneously, system may crash disrupting service for millions of enterprise users and power consumers.
Compared to their wired counterpart, wire-less/mobile networks must deal with several problems which make it very difficult to provide Quality of Service (QoS): (1) resources (e.g., bandwidth) in wireless networks are scarcer than in wired networks; (2) wireless channels are prone to location-dependent, bursty and time-varying errors; (3) users tend to move around during a communication session causing hand-over between cells.
QoS, is a collective measure of the level of service delivered to the customers by service providers, characterized by basic performance criteria such as throughput, delay or jitter, loss and reliability. ISO standard defines QoS as a concept for specifying how good the offered networking services are. In another words, QoS defines nonfunctional characteristics of a system, affecting the perceived quality of result. QoS parameters related to technology are throughput, delay, jitter, loss and reliability; QoS parameter related to static management are specification, negotiation, admission control and resource reservation.
DiffServ, QoS model endorsed by IETF, is architecture for implementing scalable service differentiation in the Internet. Service defines some significant characteristics (throughput, delay, jitter, loss or relative priority of accessing to network resource) of packet transmission in one direction across a set of one or more paths within a network. Service differentiation is desired to accommodate heterogeneous application requirements and user expectations and to permit differentiated pricing of internet service.
CELLULAR SYSTEM ARCHITECTURE
Increases in demand and poor quality of existing service led mobile service providers to research ways to improve the quality of service and to support more users in their systems. Because the amount of frequency spectrum available for cellular use was limited, efficient use of the required frequencies was needed for cellular coverage. Each cellular coverage region is planned according to an engineering plan that includes cells, clusters, frequency reuse and handover. The cellular networks (Fig. 1) consist of the following four major components that work together to provide cell service to the subscriber:
|| Cellular system component
||Public Switched Telephone Network (PSTN)
||Mobile Telephone Switching Office
||Cell Site with antenna System
||Mobile Subscriber Unit or Mobile Terminal (MSU/MT)
In cellular networks, a service coverage area is divided into smaller hexagonal areas referred to as cells. Each cell is served by a base station. The base station is fixed. It is able to communicate with mobile stations such as cellular telephones using its radio transceiver. The base station is connected to the Mobile Switching Center (MSC), which is in turn connected to the Public Switched Telephone Network (PSTN). The MSC controls calls, tracks billing information and located cellular/Mobile subscribers.
There are two types of link in the cellular system: Forward link that are used to carry traffic from the base station to mobile station, Forward link are used to carry traffic from mobile to the base station.
The frequency spectrum allocated to wireless communications is very limited, so the cellular concept was introduced. The concept of frequency reuse is based on assigning to each cell a group of radio channels used within a small geographic area. Cells are assigned a group of channels that is completely different from neighboring cells to avoid radio interference. The same channels can reuse by two cells that are far apart such that their frequencies do not interfere. By reducing the size of cells the cellular network is able to increase its capacity and therefore to serve more subscriber.
One of the key issues of cellular system is Handover, which occurs when mobile subscriber travels from one cell to another during a call. As adjacent areas do not use the same radio channels, a call must either be dropped or transferred from one radio channel to another when a user crosses the line between two adjacent cells. Because dropping the call is unacceptable, the process of handover is created. Handover occurs when the cellular telephone network automatically transfers call from one channel to another channel as mobile crosses adjacent cells.
AN OVERVIEW OF DiffServ IN CELLULAR NETWORKS
Differentiated Service Model (DiffServ) is currently a popular research topic as a low-cost method to bring QoS to todays Internet and cellular Networks. A very little research regarding QoS and DiffServ have been done on cellular networks but intensive research efforts have been done on wired networks. However, at the end of 20th century, some research efforts have done on wireless. For example, Sadeghi and Knightly discussed about the architecture and algorithms for scalable mobile QoS implemented a DiffServ router on linux platform, Borg et al. talked about multifield packet classification, Yeom and Reddy studied packet marking issue in DiffServ,
Mahadevan and Sivalingnam described the design and implementation of an enhanced Differentiated Service (DiffServ) architectural framework for providing QoS in wireless networks. QoS for wireless or cellular networks is also described by Mahadevan and Sivalingam, Pekrins[9-10].
Differentiating service in a network is achieved using different Quality of
Service. (QoS). There are two kinds of Differentiating service in a network,
integrated Service (Intserv) and Differentiated Service (DiffServ).
IntServ is used to reserve resource between a sender and a receiver for each
session using signalling. DiffServ is used to classify traffic into different
classes where each class will get some QoS. Two important (QoS) measures for
current cellular networks are the fraction of new and handover calls that are
blocked due to unavailability of channels (radio and/or computing resources).
The usage of IP as transport technology solves some of the internetworking problems
between different technologies but at the same time, IP quality of service mechanism
are not fully adapted to the reality of QoS aware mobility between different
technologies. This causes strong problems in terms of network performance, QoS
provisioning, AAAC (Authentication, Authorization, Accounting and Charging)
|| Differentiated Service field (DS field)
DiffServ is architecture for implementing scalable service differentiation in the Internet. It adopts a purely datagram architecture which defines simple forwarding mechanisms for interior network routers, pushing most of the complexity onto the network edges. The DiffServ architecture proposes a new format and use of the IP header for packet classification, called the DS (Differentiated Services) field. In IPv4, it defines the layout of ToS (Type of Service) octet; In IPv6, the Traffic Class octet. The IP header is divided into the 6-bit Differentiated Services Code Point (DSCP) and a 2-bit unused field (Table 1). A base set of packet forwarding treatments (per-hop behaviuor, or PHB) is also defined. Traffic flows are aggregated into a few classes handled by routers according to a set of PHBs. The DiffServ architecture is composed of a number of functional elements, namely packet classifiers, traffic conditioners and Per-Hop forwarding Behaviors (PHB). By marking the DS field of packets differently and handling packets based on their DS field several differentiated service classes can be provided, i.e., premium service, assured service and best-effort service.
The DiffServ architecture has been designed to implement the service differentiation
in the Internet, but only a few proposals[15-17] address the issue
of providing service differentiation in cellular networks. All these contributions
consider only two fundamental types of traffic (i.e., voice and data) and the
higher priority is simply given to voice calls over data. In 3G mobile communication
systems, e.g., UMTS, high-rate data services will be supported. Data transmission
will no longer be the supplementary service. Therefore, it is necessary to set
up appropriate service classes for supporting heterogeneous services beyond
the fundamental voice and data classes and implement service differentiation
in cellular networks to support multiple traffic types. Corresponding to the
DiffServ architecture, service classes in cellular networks can be mapped as
Premium service: for real-time highly delay-sensitive connections such as voice and low-rate video.
Assured service: for non-real-time delay-sensitive connection-oriented connections with limited delay bounds such as FTP and WWW browsing.
Best-effort service: for delay-tolerant message-oriented connections such as paging and electronic mail.
QOS IN CELLULAR ENVIRONMENTS
Now a days cellular networks are becoming to target IP (eg. VoIP) applications in their service offerings. The 3G cellular networks intend to extend this strategy to a higher degree, transforming the voice service in an IP service as well. So IP is using to transport all type of data between nodes in the core networks. The IP protocol was not originally designed to provide different grades of QoS. Nevertheless an IP approach to e2e QoS is required to be able to build an all IP world without relying on any layer 2 specific QoS mechanism.
During proceeding VoIP call, Mobile terminal periodically move between technologies (vertical handover) and between cells of the same technology (horizontal handover). This causes serious problems: these VoIP calls require the reservation of certain network resources, the assurance of certain QoS parameters and often the confirmation of administrative information. So handovers have to be done in a short period, without disturbance in the established communications. Issues as QoS assurances in mobile environments can be structured in five layers of complexity: I) physical; ii) cell; iii) network, iv) management and v) administration.
||The physical layer problems are associated with the fundamental
connection capabilities. If any of the transmission links (wired, wireless)
places any limits to QoS (e.g. delay assurances), then this will impact
the overall end to end (e2e) service.
||The cell problems are motivated by the flow handover across
neighbouring cells. QoS parameters have to be satisfied at the time of handover
between similar cells.
||The network problems are due to changes in the IP network
of attachment. This can be described as the classical mobile IP problem,
increased with QoS constraints and strong timing limitations for the handover
||Management problems are concerned with network management,
core network configuration in order to assure the required QoS levels.
||Administration is here used in the context of accounting,
logging and charging users and ISPs for resource usage.
The issue of handover illustrates the network changes in heterogeneous mobile environments. An analysis on the issues of QoS provision in a mobility environment has to consider these multiple Context Switches (CS), which are as follows:
||CS I: Mobile terminal changing wireless cells: A MT crosses
the border between two wireless cells (either UMTS or WLAN). There is no
||CS II: Mobile terminal changing IP Autonomous Domain (AD):
A MT becomes connected to a different AD. This will require AAAC traffic.
||CS III: Mobile terminal changing wireless provider: A MT switches
wireless provider. This will require AAAC traffic.
||CS IV: Mobile terminal changing transport technology (either
wireless or wired): A MT switches from wired to wireless connections or
between different wireless technologies. This may, or may not, require AAAC
traffic. When QoS issues are considered in these CS, the handover problem
becomes much more complex.
Problems and possible solution: Marques et al. discussed some network topologies to help solving or minimizing the QoS problems that may occur at the cell and network layers. The mobility support of IPv6 is better than the IPv4 support. The layer 3 handovers are currently a problem in terms of the time they take to be done. The followings are some problems concerning handover and some possible solutions discussed by Marques et al..
Wireless-wireless (same technology) handover: There exist layer 2 handover in first and second generation networks but in 3rd Generation Networks there exist layer 3 handover as well, which causes the overall delay of layer 2 plus layer 3 handovers delay.
To minimize the overall delay, both layer 2 and layer 3 handover should be decoupled. To fulfill these objectives, the following two conditions must be satisfied:
||Every layer 3 router must serve several radio cells
||Every layer 2-border radio cell must be served by more than
one layer 3 router
Wireless-wireless (different technologies) handover: In this case one cannot guarantee that there will be an overlap of both technologies coverage to allow the handover. The decision on the handover must be taken in a cost basis, that is, every time the cost is lower, the handover should be made and the opposite should only occur if:
||The user explicitly asks for it;
||The user has a premium service and the mobile is configured
to change to the most powerful technology.
Wired-wireless handover: If a MT is connected to a fixed technology (such as Ethernet) and the cable is suddenly unplugged, all the connections will be dropped, unless some measures are taken. Two different situations may occur:
|| Targeted network services
||The wireless network resources are free of charge for instance,
a wireless LAN belonging to the same company;
||The wireless network resources are to be paid for instance,
changing to an UMTS network.
As it is known the main purpose of DiffServ model is to provision end to end
QoS guarantees by using service differentiation in cellular network. Maarquesta
represent end to end QoS architecture for DiffServ model. The offered network
service parameters i.e. DiffServ network parameters of the said architecture
is given in Table 2, where the network services will be one
of three basic DiffServ types: EF-expedited forwarding, AF-assured
forwarding and BE-best effort.
As far it is known, there has been very few works addressing the problem of providing QoS and Differentiated Service in cellular networks. Recently some researches have been going on in Europe (Mobidick) and North America (Bell Lab) on QoS service in Cellular and wireless networks. I hope the discussion of the said topic in this study will give some idea. The key problems surrounding the architecture mentioned are concerned with efficient handover solutions across different cells. This article briefly described some of the handover problems and possible proposed solutions. There is a need for more study and research on this topic.