Mapping the Wireless Technology Migration Path:
The Evolution to 4G Systems

 

By J.-P. Rissen

 

Wireless service providers around the world are at a business and technological tipping point. Having made investments in legacy technologies that were designed primarily to support voice traffic they now need to cope with new standards, protocols and business imperatives. In so doing, existing business models and technology platforms will be rendered moot. Consider this:

 

• While the International Telecommunication Union (ITU)-R M.2072 expects voice traffic to account for the lion’s share of volume through 2015, a shift is already underway for revenue to be driven by rich multimedia entertainment services like video messaging or all forms of mobile commerce.
• This shift will place a premium on technologies that maximize bandwidth and throughput while maximizing spectral efficiency.
• Fourth Generation (4G) wireless architectures are rapidly maturing to address these requirements while leveraging the capabilities of Third Generation (3G) technologies, taking advantage of the features associated with an all-IP network system.

As a result of these trends, service providers are scrambling to assess, purchase and deploy new wireless service delivery technologies that will address emerging demand, while maximizing the revenue generated from traditional sources. Those organizations that most effectively manage the transition from Second Generation (2G), 3G and ultimately 4G infrastructures will be best positioned to grow through the rest of the decade and into the next. To that end, this article puts these overlapping wireless standards (2G, 3G and 4G) into a context that will explain the evolution of the wireless technology infrastructure and provide a basis for optimizing a company’s investment in next generation systems and business models.

 

Public wireless communications have evolved considerably since the emergence of 2G digital wireless cellular technologies in the early ’90s. In their earliest incarnation, Groupe Spécial Mobile (GSM) and its North American counterpart Code division multiple access cdmaOne™, primarily were designed for voice services (although they did support some limited data capabilities such as Short Message Service [SMS] and low-speed circuit-switched data).

 

By the late ’90s, 3G cellular systems emerged, driven by the need for a universal and interoperable technology with greater user performances. While 3G technologies offered an improvement on both fronts, incompatibility between different systems remained. From 2003 to 2006, early 3G technologies like Universal Mobile Telecommunications System (UMTS) and cdma2000® offered two key improvements over 2G:

 

• Simultaneous use of circuit services (like voice and video calls) and packet services (like web browsing or instant messaging)
• Higher speeds (up to 384 kb/s for data in UMTS Release 99 and a peak value of 2.4 Mb/s for the initial release of cdma2000® Evolution Data Optimized (EV-DO)

Later, these early 3G systems were enhanced with the introduction of faster uplink and downlink packet access channels such as High-Speed Packet Access (HSPA) for UMTS and EV-DO Revision A for cdma2000®. These technologies, which began to be deployed in the 2005-06 timeframe, improved both the user experience and spectral efficiency for both uplink and downlink transmission. Fast and adaptive radio interface packet transmission schemes and advanced re-transmission techniques such as Hybrid Automatic Repeat Request (HARQ) made these benefits possible.

 

The All-IP Migration

 

Examining fixed network evolution over time always has been a good way to predict the major changes in the wireless system. This is particularly true when considering the ever-increasing transmission speeds (from modem-based dial-up connections to high-speed Asymmetric Digital Subscriber Line [ADSL] and service offerings from circuit-switched to Internet Protocol [IP] packet-based architectures).

 

Clearly, wireless technologies had to be prepared for the migration toward all-IP service platforms. That is why work on developing an IP Multimedia Subsystems (IMS) standard was initiated in 2002, during the same timeframe as High Speed Packet Access (HSPA) radio access evolution. IMS provides an interoperable IP-based framework for supporting multimedia services on a single network topology.

 

Similarly, Evolved Packet System (EPS), also known as a combination of Long-Term Evolution (LTE) and System Architecture Evolution (SAE) is being defined as a major UMTS evolution, proposing an all-IP integrated architecture for both the Access Network and Core Network. A similar evolution is now under development for CDMA markets with the introduction of Ultra Mobile Broadband (UMB), also known as Evolution Data Optimized (EV-DO) Revision C.

 

 

During this same time period, a third major mobile access technology, Worldwide Interoperability for Microwave Access (WiMAX), was introduced, initially using the Institute of Electrical and Electronics Engineers (IEEE) 802.16e standard. WiMAX, LTE and UMB are all based on Orthogonal Frequency Division Multiplex (OFDM) radio access technologies and all have adopted an IP-based network architecture.

 

Toward 4G Systems

 

Now the industry is looking ahead to 4G systems that aim to expand the capabilities of 3G systems to meet increased demand in terms of user bit rate and capacity (which translates into spectral efficiency).

 

Like the 3G/International Mobile Telecommunications (IMT)-2000 evolution (Figure 1), the framework for 4G systems will be ruled by the Radiocommunication group of the ITU-R under the name of IMT-Advanced.

 

IMT-Advanced roadmap aims to finalize the requirements for candidate 4G radio access technologies by the end of January 2008 and make available the radio interface specification for the new systems by mid-2010. Completion of these steps would allow full commercial availability in 2015.

 

Analyzing 4G from the Technical Perspective

 

As of Nov. 1, 2007, the technical requirements of 4G had not yet been finalized, but industry insiders anticipate that 4G Radio Access Technologies will be able to deliver 1 GB/s over a 100 MHz channel, which corresponds to a peak spectral efficiency of 10 bit/s/Hz. The average 4G spectral efficiency will be around 5 bit/s/Hz, which represents approximately three times the performance of EPS networks.

 

From an architectural perspective, 4G is in line with the latest 3G EPS and UMB evolutions, as seen as an all-IP system. So it’s likely that the latest 3G evolutions (like EPS and UMB) and 4G will have more in common than were seen between early 3G systems and EPS.

 

4G systems will support built-in “always-on” support, as the traditional "dial-in" model doesn’t comply with the new set of services like "Presence,” "Instant Messaging" and all other real-time information or interactive services. This is a key enabler of "anytime/anywhere" user experience currently supported by technologies like Wi-Fi. In addition, 4G also promises to improve resource efficiency and enhance the user experience.

 

Systems based on 4G will support the full range of features that public 3G cellular systems already provide including:

 

• Optimized link adaptation and power control to allow increased user rates and efficient radio resource usage
• Configurable time or frequency duplex modes to better accommodate asymmetric services
• Seamless service provision in case of intra- and inter-system mobility
• Security protection for user data, control signalling and subscriber information.

In the future, special care has to be given to the radio frame structure definition, which is needed to enable different systems to coexist and to minimize handover time interruption when moving between disparate systems.

 

What Will be the 4G Standard?

 

As in IMT-2000, 4G/IMT-Advanced will not be subject to a single Radio Access Technology, but rather a family of technologies. Late in 2006, IEEE launched 802.16m, which is a candidate for IMT-Advanced services. Specifically, 802.16m will be the 4G version of the Wireless Ethernet standard and will be backward-compatible with existing 802.16e and WiMax, the IEEE standard for broadband mobile operation in licensed frequency bands.

 

In early 1999, the Third Generation Partnership Project (3GPP) and 3GPP2 groups were created to specify a candidate standard for 3G/IMT-2000. The mandates of these two consortia will be extended to encompass future 4G standard specifications.

 

In mid-2007, the 3GPP2 consortium began work on the definition of a candidate for the IMT-Advanced family, which eventually should replace the current UMB standard. The 3GPP consortium also started an IMT-Advanced activity at the end of 2007 and most likely will submit an evolved version of EPS/LTE as its 4G candidate.

 

How these Developments will Affect Migration to all-IP Networks

 

This is a major challenge for operators engaged in a comprehensive migration to an all-IP network. This change deeply impacts the Core Network and service layers, as well as the Radio Access Network (which moves toward non-hierarchical flat-IP models, starting with LTE evolution).

 

Beyond the complexity of network migration, it’s also critical to ensure a smooth transition of the subscriber base from the existing circuit-based architecture toward the new all-IP networks. There are three key points to consider:

 

• The need to maintain the Quality of Experience, not only in terms of voice or video quality but also in terms of call set-up time and network response time. Existing circuit-based systems have been optimized since the early ’90s and subscribers will not tolerate a noticeable degradation of such performance indicators.
• The need to maintain service continuity when moving between systems. cdma2000® EV-DO already provides a solution to ensure voice call seamless continuity between circuit-based and packet-based systems. A similar solution recently has been standardized to ensure seamless service continuity in UMTS systems between Circuit Switched and IMS domains.
• The need to maintain network capacity – or make sure that full wireless Voice over Internet Protocol (VoIP) does not translate into radio resource inefficiency. Circuit voice over the radio interface was initially designed in a much optimized way, based on efficient channel protection (known as UEP for Unequal Error Protection) and later extended using Adaptive MultiRate (AMR) speech codecs. In the IP domain, thanks to Robust Header Compression (RoHC) scheme, adaptive channel coding and the gains in terms of statistical multiplexing provided by the use of high-speed shared radio channels, VoIP is able to reach the same, if not better, level of performance over the radio interface.

In some operational networks, the first step in all-IP migration is performed at the transport level, long before the migration at the service level. This is made possible in GSM and UMTS networks thanks to the 3GPP-Release 4, which opens the possibility to dissociate the communication call control (which still relies on classical circuit-switched protocols) from the transport level. Similarly, the UMTS Terrestrial Radio Access Network -Internet Protocol (UTRAN-IP) evolution allows supporting any of the circuit or packet-based services over an IP-based access network.

 

What about 4G Technology Disruptions?

 

There are a number of technology enablers that add value to 4G network corresponding to Alcatel-Lucent research areas and domains of expertise:

 

Collaborative MIMO

 

Inter-cell interference is very often a limiting factor to cellular wireless system capacity. The Collaborative Multi Input Multi Output (Co-MIMO) technique relies on non-coherent signal combining, which dramatically reduces inter-cell interference. As opposed to conventional MIMO (where a terminal is served by one unique Base Station), a terminal in Co-MIMO conditions is served by multiple base stations.

 

Network MIMO

 

Network-MIMO is a coherent interference coordination which allows suppressing inter-cell interference by coordinating transmission and reception of users’ signals at many base stations. This method relies on coherent transmit- and receive-beamforming across different base stations. It mandates high-bandwidth, low-latency backhaul network and highly synchronized bases to be able to share channel knowledge information among coordinated base stations.

 

Network MIMO performance is higher than Conventional MIMO and Collaborative MIMO, at the expense of higher performance backhaul and increased complexity. Figure 2 describes the gain of Network MIMO against conventional MIMO techniques.

 

 

Multi-User MIMO

 

This technique aims to improve the overall system throughput by simultaneously serving multiple users during a frame on multiple spatial channels. These channels are formed using knowledge of the users’ channels at the base station transmitter and by applying a different set of coherent weights across the antennas for each user’s data stream. This technique is known as beamforming or precoding. It is performed in such a way to minimize interference between beams.

 

Software Defined Radio

 

Software Defined Radio (SDR) is a technology which enables multi-standard/multiband base stations. In an SDR product, a significant amount of the processing is performed by software libraries over a hardware platform which can serve as a common basis to many kinds of wireless standards. The obvious benefit of SDR is in the reduction of the number of dedicated developments for different standards and frequency bands, as such evolutions can be supported by software upgrade and at much lower cost.

 

Self-Organized Radio Networks

 

Mobile network deployment and operation are still cost-intensive, especially at installation, network optimization and day-to-day network operation and failure handling. All these operations require manual interaction and a high level of technical expertise.

 

Future systems have to provide innovative means for reduction of operational cost and ease overall system operation. Basically, this can be achieved through the three following axis:

 

• Self-configuration of the newly added site
• Self-optimization for radio parameters and neighboring list configuration
• Self-adaptation to network load variation (e.g., a sports event or unpredictable condition)

Wireless Relay and Mesh Networking

 

The emerging wireless relay and mesh networking aims to provide improved capacity and network reliability. Relaying helps extend coverage in areas suffering from excessive path loss while mesh networking architectures based on interconnected wireless base stations help provide data path redundancy. Those two techniques can facilitate the deployment of a 4G network in a cost-effective way.

 

Femto and Small Cell Coverage

 

Femto and small cell solutions – like the Alcatel-Lucent Base Station Router (BSR) – offer connectivity and access to legacy Core Network domains over a 3G radio and fixed Internet line (ADSL/Ethernet) to subscribers equipped with legacy 3G terminals. This architecture is a key element to providing increased end-user experience; knowing that 70 percent of calls are made indoors, but only two percent of buildings have purpose-built indoor coverage.

 

Besides, such a solution will extend cellular network capacity and provide wireless service providers with new revenue opportunities.

 

The Evolution of Global Wireless Traffic

 

The evolution of the volume and characteristics of global wireless traffic is driven by four factors:

 

• Mass evolution – the growing use of electronic devices such as cameras, personal digital assistants, etc. is increasing the number of people who download, exchange or share data;
• Virtual evolution – more and more user-related content is digitized;
• Socialization – people are becoming more comfortable with one-to many or many-to-many (peer-to-peer) forms of communications;
• Personalization – the development of a user-centric vision induces various new types of multimedia services with new end-user behaviors.

It’s hard to predict to what extent these drivers will influence wireless traffic. Still, by using world population forecasts, as well as service penetration and usage models, it is possible to make estimates (Figure 3). The ITU-R M.2072 report projects that traffic growth will be linear from 2007 to 2020, with voice traffic still dominant through 2015. But, the development of rich multimedia entertainment services like video messaging or all forms of mobile commerce will shift that trend toward more multimedia communications.

 

 

The next figure (Figure 4) shows the total wireless cellular spectrum demand, resulting from the traffic forecast described above. This estimation takes into account the evolution in spectral efficiency brought by 3G, evolved 3G and future 4G/IMT-Advanced technologies, as well as the fact that different generations of technology will co-exist on operated networks.

 

 

Conclusion

 

Improved spectral efficiency is a key aspect of future wireless technologies. Even more than the constant technological race toward increased bit rate, spectral efficiency will be essential to the wireless ecosystem if it is to cope with the future new forms of communication, behaviors and business models that are emerging for the next decade.

 

From that perspective, the evolution from 3G toward 4G/IMT-Advanced should be pursued in a sensible and phased manner, while providing enough backward compatibility to secure and maximize investments in existing operator assets.

 

Jean-Paul Rissen is Director of Wireless Technologies, Wireless CTO, Vélizy, France.
E-mail: jean-paul.rissen@alcatel-lucent.fr

 

 

Associated Items

 

Brochure: Transforming Communications to Enrich People's Lives

 

Press Release: Alcatel-Lucent has been awarded the “SAM Silver Class” Sustainability Distinction

 

Press Release: Alcatel-Lucent signs the United Nations Global Compact “Caring for Climate” initiative

 

Video: Making Universal WiMAX a Reality

 

Web Page: Alcatel-Lucent Corporate Social Responsibility

 

Web Page: Broadband for All