ACM / Crossroads / Xrds9-4 / Inside Wireless Evolution

Inside Wireless Evolution

Introduction: A New Age of Networking

Over the past decade, there has been a tremendous growth in the popularity of wireless networking technologies. Schools and universities are providing wireless access to students and faculty members. Malls are providing customers with wireless connectivity to allow them to search products, virtually navigate shops, and interact with services. Tourist sites are providing wireless devices to aid tourists in navigating, exploring and learning about attractions. Moreover, the emergence of new computing paradigms such as pervasiveness, ubiquity, and mobility has necessitated the rapid deployment of wireless networks as an infrastructure underneath such technologies.

Typically, wireless connectivity is provided using one of two networking technologies: Wireless Local Area Network (WLAN) technology or Wireless Wide Area Network (WWAN) technology. IEEE 802.11 is de facto standard in WLAN technologies; it was designed to provide cheap, robust and effective means for wireless connectivity in limited geographic areas. The standard is continuously developing to accommodate the increasing bandwidth intensive nature of applications. Global System for Mobile Communications (GSM) is the widely adopted standard for wide area wireless connectivity. The GSM platform has been deployed in over 179 countries worldwide. The standard is continuously evolving to accommodate the needs of subscribers and to support a wide segment of end-user services. Today’s GSM infrastructure includes General Packet Radio Service (GPRS), Enhanced Data for GSM Evolution (EDGE), 3GSM technology, Multimedia Messaging Service (MMS), and Short Message Service (SMS) [2].

This article explores the latest developments in wireless networking technologies, in particular, it investigates IEEE 802.11 and GSM technologies. It analyzes the basic features, advantages and disadvantages provided by both wireless infrastructures. In addition, it advocates that the future of wireless technologies will be a combination of both WLAN and WWAN technologies creating so-called overlay networks. Finally, it presents an insight about the future of wireless networking in the coming decades.

IEEE 802.11

Overview of IEEE 802.11

The IEEE standard provides wireless connectivity in limited geographic areas. The standard provides two modes of network configuration: ad-hoc and infrastructure. The former provides a means to create networks on the spot from a group of mobile nodes without any infrastructure support, while the latter is used to provide continuous coverage using base stations known as access points. 802.11 supports 14 partially overlapping channels of which three channels do not overlap. This allows network designers to aggregate the bandwidth within highly loaded geographic areas by overlapping coverage areas using non-interfering channels. 802.11 uses the ISM (Industrial, Scientific, and Medical) band, thus utilizing unlicensed frequencies that could be exploited by any number of technologies. The standard specifies two distinct Medium Access Control (MAC) protocols, namely Distributed Coordination Function (DCF) and Point Coordination Function (PCF), these offer contention-based medium access and contention-free medium access respectively[5].

There are two factors that determine who seizes the shared medium in 802.11: a static element represented by a fixed Inter-Frame Space (IFS) and a random backoff interval. Figure 1, shows the relationship of these intervals as mobile stations send their frames.

Figure 1: Medium Access in 802.11.

IFSs provide fixed priority levels for accessing the wireless media. The standard defines four IFSs in ascending order of time: Shortest IFS (SIFS), PCF IFS (PIFS), DCF IFS (DIFS), and Extended IFS (EIFS). Control packets such as acknowledgments use SIFS, while data packets and management packets (such as beacons) use DIFS. Therefore, in an event of a contention, acknowledgements will have higher priority than data and management packets[3].

Association in IEEE 802.11

Typically, 802.11 systems are configured with multiple access points, of which, some overlap to provide higher aggregate bandwidth within heavily loaded geographic areas. 802.11 supports the association/re-association/dissociation of mobile stations with access points to enable roaming. Each mobile station is equipped with a wireless adapter that implements the roaming algorithm. Creating the initial association of a mobile station with an access point starts with a process called scanning. The IEEE 802.11 standard defines two methods of scanning: passive scanning, in which the station switches to a channel and listens to beacons from access points that use that channel and active scanning, in which the station switches to a channel and issues a so-called Probe Request, to which a Probe Response is expected within a given time frame from the access point. Most 802.11 product vendors implement active scanning as it provides a faster and a more efficient mechanism for detecting the access points in the vicinity of the mobile stations. Performing a series of scans on different frequencies is called sweeping. There are two types of sweeps: full sweeping, which goes through all the channel-list, and short sweeping which skips the channels that do not have sufficient frequency distance from known active channels. Figure 2, illustrates a scenario with a single mobile node trying to associate with one of two overlapping access points. The mobile node is in the overlapping coverage area of AP1 and AP2, it associates with AP1 as it has a stronger Signal-to-Noise Ratio (SNR) [1].

Figure 2: Association in 802.11 Networks.

Roaming in IEEE 802.11

When a mobile unit moves away from the access point, the SNR of the link drops, and will eventually drop below a threshold value known as the cell search threshold. When this event occurs it triggers the roaming algorithm to start looking for other access points to associate with. In this process, the mobile station initiates a sweeping to find a suitable access point to bind to. When the SNR drops below a second threshold known as cell switching threshold and defined as 'cell search threshold - Delta SNR’, the roaming algorithm triggers a re-association by selecting another access point with a better signal. Figure 3 shows the relationship of the SNR of two access points as a mobile node roams from AP1 to AP2 [4].

Figure 3: Roaming and SNR in 802.11.

Recent Developments in 802.11

The IEEE 802.11 workgroup has extended the original standard by introducing the 802.11b specifications for high data rate support including two new speeds 5.5 Mbps and 11 Mbps. In addition, several task groups have been formed to continuously develop the standard, this includes task group 802.11f which is responsible for enhancing IEEE 802.11 to provide seamless interoperability across access points developed by different vendors. Furthermore, task group 802.11e has been formed to enhance the current 802.11 MAC protocols and to expand support for applications with Quality of Service (QoS) requirements. The QoS baseline proposal contains two different access methods and three QoS levels to accommodate the needs of Integrated Services and Differentiated Services. The former provides end-to-end connection-oriented QoS with support for streaming and centralized scheduling, while the latter provides a simple mechanism for prioritizing traffic within each cell.

IEEE 802.11 has proven successful as a WLAN technology, however, it fails to become a technology that provides coverage across wide areas. This necessitated the deployment of a different infrastructure that could operate effectively for wide area coverage. In the next section of this article, we investigate GSM; the widely used standard in wireless wide area connectivity [3].

GSM (Global System for Mobile Communications)

Overview of GSM

GSM is an open, non-proprietary platform that provides wide area wireless access in more than 170 countries across the globe. Typically, coverage is provided using terrestrial means in particular using special base stations mounted on towers to create coverage cells. GSM also provides extended service access in areas where terrestrial coverage is not present using a network of GSM satellites. GSM differs from the first generation wireless systems in that it utilizes digital technologies and Time Division Multiple Access (TDMA) methods in transmitting encoded voice. This permits effective and efficient human speech emulation. In addition, second generation GSM has been enhanced to provide better support for media centric data services that require high bandwidth [2].

GSM is a standard that is continuously evolving. Providing seamless integration with the Internet and high-speed support have been at the top of the agenda of global groups developing the specifications for 3GSM (next generation of mobile communications services). It is expected that the new infrastructure will be capable of supporting high-end service capabilities such as video on demand and high-speed multimedia such as streaming applications. This will include substantially enhanced capacity, quality and data rates than the ones available in second or 2.5 generations.

In the coming sections, we explore the basic technologies that constitute second generation and third generation GSM.

GPRS

GPRS (provided as part of the 2.5 generation GSM) is an add-on technology that enhances today’s GSM mobile telephone networks with support for packet-based applications such as Internet applications. In detail, GPRS involves overlaying the existing circuit-switched GSM networks with a packet-based data service interface.

One important advantage in GPRS is the minimal overhead of deploying it in a pre-existing GSM network, in particular, enabling GPRS on a GSM network requires the addition of two core modules, the Gateway Service Node (GGSN) and the Serving GPRS Service Node (SGSN). The GGSN acts as a gateway between the GPRS network and the public data networks such as IP-based networks. The SGSN provides packet routing to and from the SGSN service area for all users in that service area. However, upgrading a GSM network to become GPRS-enabled might also require adding some additional features such as mobility management stations to locate mobile stations and security features like ciphering.

GPRS-enabled networks provide many advantages over their predecessors. Packet-switched networks enable better utilization of radio resources when users send and receive data. More clearly, circuit-switched wireless technologies require the dedication of a radio channel to a mobile data user for a fixed period of time resulting in the loss of valuable bandwidth, however, GPRS allows the efficient use of scarce radio resources among a larger number of users by splitting the information into separate (but related) packets that get transmitted and reassembled at the receiving end. Packets from different users get transmitted concurrently and instead of having a dedicated connection it is a virtual one. Another important advantage in GPRS is its full inter-working with the existing Internet, therefore any service available on the fixed Internet today can be supported by GPRS [8].

It is clear that GPRS has been a significant leap in the world of mobile data services, however it is important to note that there are some limitations with GPRS. Radio resources that are used are simultaneously shared with voice calls, thus reducing the cell capacity if particular channels are reserved for GPRS use only, however, GPRS is capable of dynamically managing channel allocation therefore reducing the impact of this limitation. Moreover, the shared spectrum is divided into eight consecutive timeslots. To reach the maximal transmission data rate a single user must utilize all eight slots, this turns out to be unrealistic in a practical GPRS installation and the actual data rate experienced at the end points of a network is much lower therefore reducing the overall data rate experienced by GPRS users. In addition, GPRS suffers from severe transit delays as packets are transmitted in different directions to reach the same destination. These limitations have forced network engineers to search for more efficient solutions in third generation GSM to overcome them. EDGE technology is seen as a major improvement in 3G platforms[2].

EDGE

Like GPRS, EDGE is a migratory technology to enhance GSM and GPRS to deliver greater network capacity and higher end-user data rates. EDGE is a technology that gives GSM networks the capability to handle services for the third generation of mobile telephony. EDGE has been designed to allow the transmission of larger amounts of data at a high speed, 384 kilobits per second. EDGE uses TDMA frame structure, logic channels and 200 kHz carrier exactly similar as today’s GSM networks, which allows existing cell plans to remain intact. One important advantage in EDGE technology is the utilization of the existing, already-deployed infrastructure, therefore minimizing the cost of deployment. Unlike GPRS, EDGE will have an important effect on voice communication; in particular, the greater data capacity demonstrated by EDGE will allow freeing-up channels in the existing spectrum for voice traffic [2].

Upgrading to EDGE is at a relative low cost, it requires just one EDGE transceiver unit to be added to each cell, with the base stations receiving remote software upgrades. GPRS uses a modulation technique called Gaussian minimum-shift keying (GSMK). This modulation technique does not allow as high a bit rate across the air interfaces as 8 PSK modulation introduced into EDGE systems. In detail, 8 PSK modulation automatically adapts to local radio conditions, offering the fastest transfer rates near the base stations in good conditions. Figure 4 shows the allowable data rates using each different technology per channel [2]:

Figure 4: End User Data Rates Per Channel.

3GSM

Third generation GSM refers to the next generation of mobile communication systems that provide enhanced services to those available today such as voice, text, or data. 3GSM is based on today’s GSM standard, but evolved, extended, and enhanced to include an additional radio air interface, better suited for higher bandwidths and multimedia services. 3GSM will provide on demand video and seamless integration with the Internet. Unlike existing GSM systems, 3GSM requires a different core platform, in detail, 3GSM uses Code Division Multiple Access (CDMA) unlike TDMA technology that is currently used in 2G networks. 3G platforms will be able to provide a wide spectrum of attractive applications including video applications, mobile multimedia, and games. This will require new terminals (phone sets) that are capable of supporting video-centric content. The data rate supported by 3G networks will depend on the environment the call is being made in, Figure 5 shows the expected data rates in different environments [2]:

Figure 5: 3GSM User Data Rates.

Deploying 3GSM will require substantial financial investments. 3G networks require new radio and core network elements including new base station systems, new radio network controllers, and probably new cell planning methods to suit the new frequency allocations. However, it is necessary for future applications [2].

SMS and MMS

Finally, we explore two popular add-on technologies in GSM networks. SMS is a technology that allows sending and receiving alphanumeric text messages to and from mobile phones. Each short message is up to 160 characters of length when Latin letters are used and 70 when non-Latin characters are used. The popularity of SMS has grown rapidly in the past decade. MMS can be seen as an evolution to SMS, it is a store and forward messaging service that allows mobile subscribers to exchange multimedia messages. With this new service, clients can create multimedia messages using built-in accessory cameras or can use images and sounds stored on their phones. Similar to SMS, if the mobile phone is off, messages get stored on the network and get delivered at a later time [2, 8].

Towards 3G Migration

Third generation (3G) GSM addresses the limitations of second generation (2G) GSM with its limited data capabilities. Different migration paths are being adopted by GSM operators in the evolution from 2G to 3G systems. For example, prior to full scale 3G rollout, 3G services will be delivered as enhancements to the existing 2G networks using technologies such as GPRS and EDGE which extend the capabilities of GSM networks.

One important aspect in the development of 3G systems is the standardization of the technology and ensuring its global interoperability. The existence of several diverging standards and technologies and the enormous financial investments in them have led to the formation of different bodies such as Third Generation Partnership Project (3GPP) and Third Generation Partnership Project 2 (3GPP2) whose task is to harmonize the various proposals for global 3G technologies. Nevertheless, the absence of a particular 2G standard in USA complicated the process of standardization, in detail, there are three main 2G standards in USA: a Time Division Multiple Access (TDMA)-based standard (IS-136), a Code Division Multiple Access (CDMA)-based standard (IS-95), and a GSM derivative known as Personal Communication Services (PCS) 1900. In Europe and Japan, the process was less complicated and the European Telecommunications Standards Institute (ETSI) and Japanese proposals merged and were completed and refined by 3GPP, which had the role of harmonizing the proposals based on wideband CDMA as the multiple access technology. This resulted in standardizing Universal Mobile Telecommunications System (UMTS) to deliver 3G services. In Japan, NTT DoCoMo has been the forefront for 3G research and collaborated with other operators across Asia in setting up a number of UMTS test beds. Similarly, in Europe, several operators are running pre-commercial UMTS trials. UMTS licenses have been awarded across Europe and Asia and commercial operation of UMTS is indeed in progress. For interoperability with North American networks, a group known as 3GPP2 was formed to develop global specifications for 3G networks with focus on North American and Asian interest. Their proposals evolved from the IS-95 and IS-136 standards making them suitable for deployment in the existing cellular and PCS bands in North America.

In spite of these efforts to reach a universal 3G standard, the 3GPP and 3GPP2 proposals diverged. This prompted a group of international operators known as the Operator Harmonization Group (OHG) to suggest the harmonization of the 3GPP and the 3GPP2 concepts; hence, developing a Global Third Generation (G3G) to allow the interoperability of networks world wide. This process of harmonization produced a standard following three modes of operation including:

CDMA-DS (CDMA - Direct Sequence) based on UMTS Frequency Division Duplex (FDD)
CDMA-MC (CDMA - Multi Carrier) based on cdma2000
CDMA-TDD (CDMA - Time Division Duplex) based on UMTS TDD

The standardization of G3G is becoming an exciting prospect that promises a wide range of spectacular media centric services with global interoperability [6].

A Glimpse into the Future of Wireless Networking

There has been a great controversy regarding the future of wireless computing in the new century. Is it going to be based on wide area technologies like GSM or on local area technologies like IEEE 802.11?. One crucial advantage in local area wireless networks is their capability to support very high bandwidths, unlike wide area networks, WLANs can support up to 10 Mbps in existing commercial products and up to 100 Mbps in research labs. It seems that WWAN technologies are incapable, at least in the near future, to compete in that area with WLAN technologies. In addition, the increasing media centric content of the Internet requires the availability of higher bandwidths to satisfy the requirements of emerging applications. However, IEEE 802.11 utilizes an unlicensed frequency spectrum namely ISM bands, which makes them unattractive as a commercial product for selling by industry giants. For example, it is possible for an individual to setup an access point (base station) at his home using that spectrum, and therefore interfering with a local area network of a particular vendor. Nevertheless, numerous industry leaders envision a future of both networks overlaying each other: both providing coverage at a particular physical area creating so-called overlay network. For example, an individual while driving could use the 3GSM platform as it provides a city wide coverage, but when the same person enters a particular building it would be possible for him to roam to a local area overlay network and to continue his applications seamlessly at a higher bandwidth [9, 10].

Furthermore, the newly emerging wireless systems pose numerous research challenges that require innovative engineering solutions. Satyanarayanan in [7] provides a comprehensive list for the major areas of research requiring exploration including:

Mobile networking (eg. TCP mobility adaptation, mobile IP, handovers)
Mobile information access (eg. disconnected operation, weak consistency)
Adaptive applications (eg. transcoding, agility)
Energy-aware systems (eg. disk spin-down)
Location sensitivity (eg. GPS, triangulation, context-awareness)

In conclusion, 802.11 and GSM are becoming increasingly popular as the platforms for local and wide area networking, respectively. They complement each other and are continuously evolving to accommodate the increasing demands of existing applications. With no doubt, both standards will determine how wireless networking evolves in the coming decades.

References

1: Al-Bin-Ali, Fahd. Design and Implementation of an Inter-cell Management System: The Sabino System. Computer Science Department Technical Reports TR02-02, University of Arizona, 2002 <http://www.fahd.albinali.name/papers/TR02-02.pdf>.
2: GSM Association. <http://www.gsmworld.com/index.shtml> (10/28/2002).
3: IEEE 802.11 Work Group. <http://grouper.ieee.org/groups/802/11/> (10/20/2002).
4: Lucent Technologies Inc. Roaming With WaveLAN/IEEE 802.11. WaveLAN Technical Bulletin 021/A, December 1998.
5: Lucent Technologies Inc. IEEE 802.11 Channel Selection Guidelines. WaveLAN Technical Bulletin 003/A, November 1998.
6: Roke Manor Research. UMTS - Universal Mobile Telecommunications System. <http://www.roke.co.uk/communications/cellular/systems/umts.asp> (12/8/2002).
7: Satyanarayanan, M. Pervasive Computing: Vision and Challenges. IEEE Personal Communications, August, 2001.
8: Scourias, John. Overview of the Global System for Mobile Communications. University of Waterloo, Canada, May 1995.
9: Stemm, M. and Katz, R. Vertical handoffs in wireless overlay networks. ACM MONET Special Issue on Mobile Networking in the Internet. 1997.
10: WMCSA 2002. Personal Notes. New York, 2002.

Biography

Fahd Al-Bin-Ali (fahd@albinali.name) received his BSc and MSc in Computer Science from the American University in Cairo in 1999 and the University of Arizona, Tucson, in 2002, respectively. He has been working as a research assistant in Arizona with Toshiba Information Systems, Japan, in the area of distributed computing. After that, he joined Hewlett Packard Laboratories, CA, as a research intern where he was responsible for building a pervasive infrastructure for navigating the web. Fahd will be joining the distributed multimedia research group at Lancaster University, UK as a visiting researcher starting from January 2003. His research interests lie in the areas of pervasive, mobile, and distributed computing.

Crossroads

The ACM Student Magazine