MOBILE EUROPE

INSIGHT REPORT

LTE, LTE-A AND 4G the move to the Het Net

REPORT AUTHOR Zahid Ghadialy, MD, ExplanoTech www.explanotech.com

SPONSORS

INSIGHT REPORT

LTE, LTE-A AND 4G

BY ZAHID GHADIALY, MD, EXPLANOTECH
hird Generation mobile technology, developed by 3GPP, is deployed throughout the world in many different flavours and is officially known as IMT-2000 standard by

conditions (high mobility scenario) and 1Gbps (Gigabits per second) in slow changing channel conditions (low mobility scenarios). LTE-Advanced (generally referred to as LTE-A) can achieve these speeds (detailed further) and can even better them. Hence this is one of the official 4G technologies as defined by the ITU. LTE-A is built on top of LTE and enhances some of the features to achieve the desired results. It is officially defined as part of the Release-10 standards of 3GPP. Some of the features were considered as too advanced, and not enough analysis was done due to tight schedules, and have been moved to Release-11 (September 2012) and Release-12. A key thing to remember is that LTE-A devices are backward compatible so that they can work seamlessly in an LTE cell. At the same time, the LTE-A cell should cater for an LTE device as well as LTE-A device; both should work without problems. NEED FOR LTE-A LTE promises a theoretical maximum downlink of 300Mbps which is significantly lower than what is required for the technology to be IMT-A compliant. At the same time LTE-A promises a theoretical maximum of up to 3Gbps. Even though 3Gbps may be impossible to

the ITU (International Telecommunications Union), who are responsible for coordinating efforts between different parties, including the government, private sector and industries, to achieve a common set of standards that can interwork across many different technologies and countries. Long Term Evolution (LTE) is the 3GPP standard Release-8 enhancement involving the evolution of the air interface and the core network. Even though LTE is known as 4G, it is more of a marketing term rather than being an actual 4G technology. Some networks even brand their advanced 3G networks that have HSPA+ technology as 4G. LTE is a part of the IMT-2000 family and hence would be considered by ITU as another flavour of 3G standard. Some operators brand LTE as 3.9G to signify that it is one step behind the real 4G. ITUs 4G program is officially known as IMT-Advanced. One of the main requirements for IMT-Advanced is that the technology should be able to achieve peak data rates of 100Mbps (Megabits per second) in fast changing channel

achieve, even in lab conditions, it is an indication that the data rates could be far higher than can be achieved with LTE. The higher speeds are necessary for some of the advanced applications that are promising to revolutionise our lives. Take for instance cloud based systems that could only work in a mobile environment in an acceptable way if large amounts of data could be sent and received from cloud servers. Sticking with the cloud analogy, another important requirement is that the latency should be low or it may slow down operations when small bursts of data need to be sent and received at regular small intervals. This is achieved by the small latency periods in LTE-A. Another issue faced by all mobile technologies is that cell edge coverage can be very poor leading not only to really slow data rates for users at the celledge but at the same time causing interference and affecting cell capacity. This issue is resolved in LTE-A with the help of Relays and Co-ordinated Multipoint (CoMP) technologies. Finally, with spectrum running out in the lower frequency bands, it has become even more necessary to utilise higher frequency bands. The higher the frequency of radio waves, the lower their capability to penetrate through physical

IMT-A Bandwidth Scalable, at least 40MHz

LTE (Rel-8) Scalable, 1.4MHz to 20MHz

LTE-A (Rel-10) Scalable, up to 5x20MHz (100MHz max). Rel-10 supports a max. of 2x20MHz (40MHz max)

Peak Data rates

DL = 1Gbps, UL = 1Gbps

DL = 300Mbps, UL = 75Mbps

DL = 3Gbps, UL = 1.5Gbps Rel-10 supports DL=1.2Gbps, UL = 600Mbps

Latency Peak Spectral Efficiency

User plane (UP) Control plane (CP) Downlink (DL) Uplink (UL)

10ms max. 100ms max 15 bps/Hz 6.75 bps/Hz

4.9ms 50ms 15 bps/Hz 3.75 bps/Hz

4.9ms 50ms 30 bps/Hz 15 bps/Hz

Table 1: The need for lTE-A. Increased theoretical maximum peak data rates of 3Gbps, spectral efficiency of 30bps/Hz
16 | Mobile Europe Insight Report

INSIGHT REPORT
Enhancements Carrier Aggregation Enhanced MIMO support New functionality Relays HetNets and eICIC CoMP (Co-ordinated Multipoint)
structures like walls. At the same time, data consumption is increasing indoors due to the popularity of smartphones and tablets. To cater for these requirements operators are in the process of introducing femtocells and other small cells. The biggest challenge is to make them work together without causing interference with each other, with macro cells and to the users, regardless of them being allowed to access these small cells or not. This requires the need to use advanced interference avoidance techniques, which in turn has introduced enhanced Inter Cell Interference Coordination (eICIC). LTE-A ENABLING TECHNOLOGIES LTE-A is built on top of LTE, so the underlying technology is the same. Some of the characteristics of LTE have been enhanced in LTE-A while some additional functions have been introduced. It is important to differentiate between them. A network will be considered an LTE-A network if some of these enhancements are present. The new functionality can be considered as a bonus which is good to have but not necessarily a must have. CARRIER AGGREGATION Carrier Aggregation (commonly known as CA) is one of the main features of LTEA and is the basis for converting LTE into an IMT-A capable technology. In the simplest of forms, CA can be defined as combining up to five different LTE carriers into a single carrier. In an ideal situation each of these carriers (known as CC or component carriers) would be 20MHz, thereby making the total 100MHz. At the same time is should be pointed out that for a Release 8/9 LTE device each CC should still continue to act as an independent CC as the device is not aware of CA. Even though in theory CA can aggregate any bands, there are initially some restrictions in place. Rel-10 standards define an Intra-band CA (explained later) for bands 1 (FDD) and 40 (TDD), and Inter-band CA (explained

Other non-LTE-A new features

Table 2: Difference between enhancements, and new functionality in LTE-A

Ericsson LTE equipment in Stockholm. Huawei has since announced LTE-A trials with Telenor and Tele 2.

later) between bands 1 and 5. Rel-11 adds band 41 (TDD) for Intra-band case and bands 1 and 19 for Inter-band CA. These combinations dont satisfy the likely needs for every operator and as a result the

following combinations are being worked on in Rel-11 to satisfy the needs of more operators (see table below). Even though the focus is only on two bands in Rel-10/11, there is a good

Bands Band 3, Band 7 Inter-band CA Band 4, Band 17 Inter-band CA Band 4, Band 13 Inter-band CA Band 4, Band 12 Inter-band CA Band 5, Band 12 Inter-band CA Band 7, Band 20 Inter-band CA Band 2, Band 17 Inter-band CA Band 4, Band 5 Inter-band CA Band 5, Band 17 Inter-band CA Band 3, Band 20 Inter-band CA Band 8, Band 20 Inter-band CA Band 1, Band 7 Inter-band CA Band 3, Band 5 Inter-band CA Band 4, Band 7 Inter-band CA Band 11, Band 18 Inter-band CA Band 1, Band 18 Inter-band CA Band 1, Band 19 Inter-band CA Band 1, Band 21 Inter-band CA Band 38 Intra-band CA

Operator and/or Region Europe AT&T, USA Verizon, USA USA USA Europe AT&T, USA AT&T, USA AT&T, USA Europe Europe Europe and Asia (mainly China) SK Telecom, South Korea Canada, USA, Latin America, Europe and Asia KDDI, Japan KDDI, Japan NTT Docomo, Japan NTT Docomo, Japan China and Europe

Table 3: Combinations of bands being worked on in Rel-11 for Inter-band Carrier Aggregation
Mobile Europe Insight Report | 17

INSIGHT REPORT
FDD Carrier Aggregation: The R10 UE can be allocated resources DL and UL on up to five Component Carriers (CC). The R8/R9 UEs can be allocated resources on any ONE of the CCs. The CCs can be of different bandwidths. (Image: 3GPP) data rates due to multiple antenna elements would depend on a lot of different conditions, mainly the channel condition. MIMO has existed since mid 70s. The amount of processing that is required to process the MIMO channel in real-time has been a limiting factor until recently. With advancements in technology, it is easier to process received signals much faster in much noisier environments using much more complex algorithms. Whereas once the real-time speed of these algorithms was the main limiting factor for the use of MIMO, this is no longer the case nowadays. The limiting factor now is a complex combination of the number of antennas, real-time processor requirements and power consumption. It should be pointed out that the number of antennas is limited by the form factors of the mobile devices themselves. The following enhancements have been done for the Rel-10 of the standards:

chance that CA with more than two CC may be specified in the Rel-12 timeframe. A mobile device declares its capability by using the UE category during the initial signalling at registration time. For CA, a new category (category 8) has been defined that indicates that the mobile device supports CA. The device further states its support for different frequency bands or band combinations as part of its capability. There are various types of CA within LTE, the most basic being Intra-band and Inter-band. Intra-band is when the CCs are within the same band specified in the 3GPP specifications. Since in LTE the maximum bandwidth could be 20MHz, it is possible that the band specified is much larger, say 100MHz. In this case each CC can be a maximum of 20MHz and multiple CCs can be combined from this band to achieve larger bandwidth and in turn higher data rates. This scenario would be known as Intra-band CA. Intra-band CA could be contiguous or non-contiguous. Inter-band would always be non-contiguous. For example Band 1 is 60MHz wide. It may be possible that some operators have 2x10MHz that may be non-contiguous. In this case CA would have to be used to achieve 20MHz. In Rel10, for Intra-band CA, only contiguous spectrum aggregation was supported; non-contiguous support is present in Rel11 version of standards. Even though the focus right now is mainly on the Intra-LTE CA, there are several other forms of CA in evaluation phase in Rel-12 and beyond. Some of them are: T LTE Inter-mode CA: Aggregating CCs
18 | Mobile Europe Insight Report

from FDD (Frequency Division Duplex) and TDD (Time Division Duplex) bands. T Inter-RAT CA: Aggregating carriers from different RATs (Radio Access Technologies) like LTE and HSPA+. T Licensed and Unlicensed CA: Aggregating LTE with unlicensed spectrum technologies like WiFi. T Intra-LTE Whitespace CA: Aggregating LTE with whitespace carriers that can be shared opportunistically between different operators based on the requirements. This is a very ambitious technology with many challenges, chief among them being the way to dynamically allocate the spectrum. The minimum requirement from a Rel10 deployment of LTE-A network will include some form of Intra-LTE CA; either Intra-band or Inter-band. ENHANCED MIMO SUPPORT Multiple Input Multiple Output (MIMO) is a way of referring to the multiple antennas present in the transmitter and/or receiver. A traditional network that has one transmit antenna and one receive antenna is referred to as Single Input Single Output (SISO) system. MIMO antenna configuration is specified by M transmit and N receive antenna elements. When M > N, the system benefits from transmit diversity while when M < N, the system is enhanced with receive diversity. In the simplest of explanations (though not entirely correct), a 2x2 MIMO, where there are 2 transmit antennas and 2 receive antennas, can be considered as doubling the data rate as opposed to a SISO system. In practice the increase in

REL-10 ENHANCEMENTS
Enhanced DL spatial multiplexing: The number of antennas has increased to a maximum of 8x8 in Rel-10 from 4x4 in Rel-8/9. These can be combined with beamforming. In practise though the most popular configuration still remains 4x2.

Enhanced UL spatial multiplexing: The number of antennas has increased to a maximum of 4x4 in Rel-10 from 2x2 in Rel-8/9.

Enhanced DL MU-MIMO: Even though Multiuser-MIMO (MUMIMO) was defined in Rel-8, it had limitations and did not always produce the desired results. As a result this feature was enhanced in Rel-10 to provide better results.

INSIGHT REPORT

RELAYS Repeaters have been commonly used with the mobile technologies. A repeater receives a signal that is generally weak, amplifies it and re-transmits it. The process of amplification also amplifies the noise that was received by the repeater. The repeaters are sometimes referred to as Layer-1 relay or AF (Amplifier and Forward) type relay. Relays on the other hand, receive a signal, decode it and then re-transmit it. They are also known as DF (Decode and Forward) type relays. Relays improve coverage in an area where the signals are weak but increase the time required for the signal to reach its final destination. This means that latency increases with the Relay node in between. This increase in latency is still much better than a situation where re-transmissions increase due to weak coverage, resulting in low data rates (indirectly increasing latency) and reducing the cell capacity. The main motivations for Relays are that they would make a mobile device appear closer to the transmitting antenna, thereby reducing the amount of power transmitted and at the same time resulting in improved throughput. A Relay appears as a mobile device to the eNB and as an eNB to a mobile device. This ensures that legacy Rel-8 devices are unaffected by the Relay node. The standards have put restrictions that do not allow the Relay node to move, and also multi-hop relaying is not supported. This ensures that latency does not

increase so as to affect the performance expected from LTE/LTE-A. HETNETS AND eICIC A HetNet or Heterogeneous Network is, in a broad sense, a network that consists of different types of technologies that coexist and are available to be used by the mobile device in the most convenient and cost-effective way. In case of LTE it is generally referred to as the LTE Network, with different topologies that co-exist to create a network better than the individual components of the network. Occasionally HetNets are confused with Hierarchical Cell Structures (HCS) and it is often assumed that they are different names for one and the same thing. HCS includes different types of cells in different frequency bands. HCS would generally be deployed with overlapping Macrocells with other Macrocells/Microcells where one frequency (generally the lower one) would be used to improve coverage and the other frequencies (generally the higher ones) would be used to improve the capacity of the network in a geographical location. HetNets on the other hand can use the same or different frequencies for the different types of cells. If different frequencies are used, there is no need for having advanced interference management techniques. If the same frequencies are used (even partially) then there is a need to manage the interference between all the cells that use these same frequencies. This is the biggest challenge with the deployment of HetNets.

With the advent of Small Cells (Femtocells, Metrocells, etc.) and Relays, it has become necessary that they co-exist harmoniously among themselves and with the Macro-cells. There are two main problems with the non-Macro-cell; interference to the Macro-cell (reducing capacity) and interference to other mobile users who may either be not allowed to access the services of this non-Macro-cell as in case of residential Femtocells that are configured in closed subscriber group (CSG). Or there may be dead zones in between non-Macro-cells and Macro-cells, thereby reducing coverage. In the case of the UMTS family, the general option used was to put these small cells on a different frequency so that they did not cause interference to the macro cells. There were a few other techniques that were used to avoid interference with other small cells. In case of the LTE family, with the scalable bandwidths available, it is not an efficient method to put these small cells on their own frequency band. The spectrum available for mobile technologies is expensive and rare. As a result in LTE Rel-8, ICIC (Inter-Cell Interference Co-ordination) was introduced. ICIC uses the fractional frequency reuse concept like good old GSM. Using this method, Cell Centre Users (CCUs) are allowed to use all frequencies but Cell Edge Users (CEUs) are only allowed to use certain frequencies. This ensures that the high power used by CEUs belonging to one

The Het Net (Credit: 4G Americas)

Mobile Europe Insight Report | 19

INSIGHT REPORT
complicated due to the new topologies, MDTs intention is to reduce this expense saving the operator considerable sums. MDT is often referred to as part of SON. T Enhanced Multimedia Broadcast Multicast Services (eMBMS): Certain operators think that eMBMS, which is seen as a key enabler for Mobile TV services in the operator spectrum, can provide additional services and revenue to them. Even though eMBMS was standardised in Rel-9, HetNets are seen as the best opportunity for rolling eMBMS out. T Home eNB (HeNB) mobility enhancements: HeNB is the 3GPP standard defined term for Small Cells. Even though the term Home is used, the Small-cell could be a residential, enterprise, accessible to public or closed subscriber group type. The intention is to just differentiate with the Macro-cell. To provide seamless coverage to the end user, inbound mobility was specified in Rel-9 and HeNB to HeNB mobility has been defined in Rel-10. The latter is a very important requirement for enterprise deployments. T Machine-to-Machine (M2M) Enhancements: The 3GPP standard term for M2M devices is Machine Type Communications (MTC) device. We will refer to this as M2M for ease of understanding. M2M is predicted to be the next big thing. The current number of mobile connections is estimated to be 10 billion which is expected to reach to 50 billion by 2025. The bulk of these new connections would be M2M devices that would be used in our everyday lives, mostly as sensors for various applications in smart meters, consumer products, healthcare and so forth. LTE/LTE-A is naturally suited for M2M devices because these devices use very small amounts of data that can be sent on the shared channel without affecting the capacity of the cell in any way. The enhancements include; optimisation of the networks to reduce power consumption, handling of large number of device groups so that other important services remain unaffected, new applications that can be used with the M2M devices and so forth.

Picture: Ericsson antenna, Sweden

cell do not interfere with the CEUs belonging to other cells. Small Cells could use the CEUs frequencies to keep the interference to a minimum. HetNets are expected to become a commonplace in LTE-deployments, more for higher capacity and speed rather than poor coverage issues. In HetNets, small cells would be distributed throughout the cell rather than just the cell edge. As a result the ICIC funtionality was enhanced (eICIC) to cater for small-cells distributed throughout the cell as well as the cell edge users. In eICIC, the Macro-cell sends an ABS (Almost Blank Subframe) where no user traffic is sent. During this time the small cells can send their data, reducing the interference between the Small and Macro cell. The ICIC feature is designed for a static environment whereas eICIC is designed to be dynamic. During peak times the macro can send few ABS subframes to manage the traffic in the Macro-cell whereas in the evening the number of ABS subframes can increase to cater for heavy indoor traffic. eICIC is a major tool to reduce interference between macro and pico layers, thereby increasing the capacity of the Macro-cell and at the same time improving cell-edge throughput and coverage. COMP The main aim of Co-ordinated MultiPoint (CoMP) Tx/Rx is to increase

the throughput of the mobile device at the cell edge. The basic idea of CoMP is similar to soft handover used in UMTS systems. In CoMP, interference from neighbour cells at the cell edge is turned into a useful signal that can be combined to produce an increased throughput. The main challenge is to make sure that, ascoordination is now be required between different eNBs, the processing delay is kept small and latency is not increased. Even though Co-ordinated MultiPoint (CoMP) looks like a straightforward feature, it has not provided the desired results in the field. As a result the feature has been pushed back to Rel-11. OTHER NON-LTE-A ENHANCEMENTS Other features, though not directly a part of LTE-A, are expected to play a big role in the LTE-A networks. Here is a quick summary of the important features: T Self-Organising Networks (SON): With HetNets, one of the biggest problems is optimising the transmitters to manage the power, frequencies, interference, etc. This is the main intention of the SON. SON will allow plug and play (PnP) capability for the deployment of new network nodes including Macro-cells, Relays, Small-cells, etc. T Minimisation of Drive Testing (MDT): Drive testing is one of the biggest expenses for a network operator during the deployment phase. With the deployment becoming ever more

20 | Mobile Europe Insight Report

INSIGHT REPORT
HSPA (Rel-10) Bandwidth Peak DL Data Rates Max. no of Antennas Latency Peak Spectral Efficiency DL UL UP CP DL UL Scalable, up to 4 carriers of 5MHz (20MHz total) 168Mbps 2x2 MIMO SISO 8.67ms 76ms 8.6 bps/Hz 2.3 bps/Hz LTE-A (Rel-10) Scalable, up to 2x20MHz (40MHz total) 1.2Gbps 8x8 MIMO 4x4 MIMO 4.9ms 50ms 30 bps/Hz 15 bps/Hz

Table 4: Comparison of HSPA Rel-10 and LTE-A Rel-10: We can see why LTE/LTE-A is more flexible, efficient and the way forward. HSPA+ VS. LTE-A 3G and its evolution HSPA/HSPA+ has been widely deployed throughout the world. In fact most of the networks have been now either upgraded to HSPA+ or are in the process. HSPA+ has its own evolution path that competes with and complements LTE. Operators that have deployed 3G networks would like to squeeze as much as they can out of it. In theory this sounds simple but in practice it may not always be easy. Many deployments of 3G/HSPA networks are quite dated, which means that the hardware is not as flexible as they are in the newer deployments. Many technical challenges on SDR and RF have been solved in the recent years. As a result a few of these older network deployments may not be upgradeable only with software. They may require new architecture, hardware and RF (radio frequency) components. The upgrade costs may be prohibitive or as high as a new deployment. These deployments would reach the end of their life when there is enough availability and uptake of LTE devices to justify the re-farming of the spectrum. It is still worthwhile comparing the technologies so that we know the differences and can see why LTE/LTE-A is more flexible, efficient and the way forward. There are further enhancements defined in Rel-11 that that will take the HSPA+ data rates even higher. It is unlikely though that all these enhancements would be deployed widely. It is accepted that MIMO works better with OFDMA (Orthogonal frequency division multiple access) which is the basis for LTE/LTE-A as compared to WCDMA (Wideband code division multiple access) which is the basis for 3G/HSPA. The consensus in the industry is that when the Voice/SMS issues are completely resolved in LTE/LTE-A and there are enough mobile devices STATE OF MARKET available at competitive prices then the operators would start phasing out their HSPA/HSPA+ networks and refarm the spectrum for LTE-A. With Carrier Aggregation it would be comparatively easy to create a wider

Sprint, USA Clearwire, USA

Rel-10 network with CA support in H1, 2013 Plans to deploy an LTE Advanced-ready network by June 2013, will be able to deliver theoretical peak speeds of up to 168 Mbps by 2014

SK Telecom, Korea

Successfully demonstrated CoMP at Mobile World Congress (MWC) 2011 and CA at MWC 2012. Plans to launch a Rel-10 network with CA support in H2, 2013 to achieve data transmission speeds up to 150Mbps . A recent press release showed that SK Telecom have successfully demonstrated the use of eICIC with Qualcomm and NSN.

T-Mobile, USA AT&T, USA

Rel-10 network with CA support in AWS bands by 2013 Plans to deploy LTE-A network in 2013 to eventually use CA to glue together the 700 MHz spectrum with its existing AWS, 1900 MHz or 850 MHz spectrum holdings. Though it has not been mentioned, the requirements will only be met by Rel-11 capable network.

NTT Docomo, Japan

Conducting field experiments of LTE-A in real radio environments in the cities of Yokosuka and Sagamihara. NTT DoCoMo has confirmed the performance of LTE-Advanced technologies using simulators in its R&D centre, achieving transmission data rates of approximately 1 Gbps on the downlink and 200 Mbps on the uplink. Plans commercialisation around 2015.

Mobitel, Slovenia

Plans a full deployment of LTE and LTE Advanced using 800 MHz, 1800 MHz and 2.6 GHz bands.

3 Italia, Italy

Deploying LTE1800 and from 2013 intends to offer CA allowing LTE1800 and LTE2600 to be used together.

Tele2 and Telenor, Sweden

Announced joint trials of an LTE-Advanced (LTE-A) network.

Table 5: Operators with announced plans for LTE-A features

Mobile Europe Insight Report | 21

INSIGHT REPORT
Name Huawei Features and Timelines Unveiled the LTE-A inter-band carrier aggregation (CA) solution at the Mobile World Congress 2012 in Barcelona and also showcased interband carrier aggregation at 800MHz and 2.6GHz with peak throughput of over 225Mb/s. ZTE Demonstrated the base station's interband carrier aggregation capabilities on two bands - 20MHz channels in 2.6GHz and 1.8GHz and MIMO to achieve peak rates of up to 270Mbps in the downlink. Ericsson Demonstrated CA of 3 x 20 MHz channels over the air in a mobile environment using their multi-mode, multi-standard radio base station, RBS 6000. It showed speeds more than 10 times faster than the existing LTE network (June 2011) in Sweden. 8x8 MIMO was used on the downlink. Expects first stages of LTE-A to be in commercial operation in 2013. Nokia-Siemens Networks The Flexi Multiradio 10 Base Station has been used for a demonstration of super-fast LTE-Advanced data speeds delivering mobile broadband at 1.4 Gbps on 100 MHz spectrum. Commercially available early 2013. Also demonstrated speeds of 1.3 Gbps throughput over TD-LTE by aggregating 60 MHz of spectrum and Multi User MIMO (MU-MIMO) using the commercial Single RAN Flexi base station hardware.
Table 6: Network Equipment Manufacturers: leading innovations to date. Picture: Nokia Siemens Networks, Flexi Base Station bandwidth and as a result achieve higher data rates. CHALLENGES AND OUTSTANDING ISSUES We have seen above that LTE-A will help achieve the higher data rates that are needed for demanding applications like video-conferencing, pictures/files transfer, video/audio streaming, etc. It will also help lay the foundation for new and innovative applications like cloud-based services and multi-part video conferencing applications that require high speeds and low latency. The benefits are not without challenges though. The biggest challenge for LTE/LTE-A networks is band fragmentation. With the possibility of having LTE in many different bands and on top of that different band combinations for CA, it can be a real challenge to have devices that will work for different operators, regardless of the geography. This could cause two issues; economies of scale would be difficult to achieve and roaming interoperability would be compromised. The various cells used in CA are mostly non-overlapping. This means that a mobile device that is moving will change different carrier combinations quite often and this in turn would require signalling overhead in addition to the radio frequency (RF) complexities associated with the device. The new antennas that are introduced as part of MIMO require additional

Name Qualcomm

Features and Timelines The third generation of LTE modem chipsets MDM9225 and MDM9625 support CA, will begin sampling in Q4 2012

Freescale

The QorIQ Qonverge 4860 base station-on-chip(SoC) is already available and supports March 2011 Release 10 LTE Advanced standard.

Table 7: Chipset Manufacturers: support for LTE-A releases.

Name Aeroflex

Features and Timelines The TM500 Test Mobile, the de facto industry standard for testing LTE base stations or eNodeBs, supports all of the carrier aggregation scenarios specified in 3GPP Release 10, and is also ready for all those currently proposed for Release 11. Recently, China Mobile Research Institute (CMRI) and Aeroflex have signed a memorandum of understanding for a cooperation covering both LTE (Long Term Evolution) and LTE-A (LTE-Advanced). Under the terms of the agreement, the companies will collaborate on the testing and verification of LTE functionality and performance, as well as key LTE-A technologies, such as carrier aggregation, eICIC (enhanced inter-cell interference coordination), and UL-MIMO (uplink MIMO), using the Aeroflex TM500 Test Mobile and EAST500 capacity test system.

Agilent

Signal Studio N7624B signal generation software and 89600B VSA LTEAdvanced software option has specific LTE protocol to set up the measurement correctly for that technology. These software packages work

22 | Mobile Europe Insight Report

INSIGHT REPORT
on multiple hardware platforms including MXG Signal Generators and PXA Signal Analyzers. Key features of these two products for FDD and TDD, Rel-10 in UL and DL signal configurations include CA for both contiguous and noncontiguous component carrier configurations for up to 100 MHz I/Q bandwidth using Agilent's MXG vector signal generators and up to 140 MHz analysis bandwidth with the Agilent PXA signal analyser; independent setup parameters for each component carrier, including any LTE-Advanced specified bandwidth or modulation type; simultaneous analysis of up to five component carriers, a feature unique to the 89600B, and troubleshooting of each component carrier using a rich selection of measurements, including EVM, CCDF and more; and enhanced uplink-clustered SC-FDMA and simultaneous control and data channel (PUCCH and PUSCH) support. Also for designing chipsets, the following SystemVue software W1918 LTE-A library and W1715 MIMO Channel is available. The library mostly functions at Layer 1, but in terms of protocol, allows some active HARQ feedback in closed-loop Throughput simulations. Anite Currently developing device testing solutions for development, conformance and interoperability testing related to LTE-A in collaboration with key industry players. Road map is aligned to market requirements and notable features include carrier aggregation, MIMO enhancements (including higher order antenna configurations, multi-user MIMO and uplink MIMO) and data-throughput support for higher UE categories. Anritsu Demonstrated full-stack Carrier Aggregation capability at CTIA Wireless 2012 with a single MD8430A, based on the network simulators unique availability of two active LTE base stations and four downlink RF transmitters in a single unit. When each active base station is configured with two RFs in a 2x2 MIMO configuration, 150 MB/s downlink throughput is available from each base station, and 300 MB/s downlink throughput is available after aggregation. Both intra- and inter-band aggregation are available with the MD8430A, using either contiguous or non-contiguous carriers. Recently MD8430A-085 Carrier Aggregation option was rolled out. Rohde & Schwarz The R&SSMU-K85 option allows testing of LTE-Advanced physical layer features in line with release 10 of the 3GPP LTE standard. It covers downlink and uplink signal generation and adds features such as carrier aggregation of up to five carriers including cross-carrier scheduling as well as enhanced SC-FDMA uplink. Mandatory prerequisite for the R&SSMU-K85 option is the R&SSMU-K55 LTE option.
Table 8: Test & Measurement Equipment Manufacturers: LTE-A support. reference signals for these new antennas. These reference signals now increase the overhead for Rel-10 mobile devices as they need them to recognise the new antennas. For the Rel-8/9 devices these reference signals are interference. Different technologies in LTE-A have been discussed but different operators may choose different deployment patterns. This may cause instability in the network for initial periods where the results may be worse than deploying without the enhanced features. For example, Relays and small-cells are independent of each other yet they would need to be dealt with in a similar fashion using SON algorithms. With small-cells power on/off is not a guaranteed event, so these algorithms may have to keep working and make significant changes dynamically which may result in stability issues.

REFERENCES/FURTHER READING
Books 1. 4G: LTE/LTE-Advanced for Mobile Broadband by Erik Dahlman, Stefan Parkvall, Johan Skold 2. LTE Advanced: 3GPP Solution for IMT-Advanced by Harri Holma, Antti Toskala 3. An Introduction to LTE: LTE, LTE-Advanced, SAE and 4G Mobile Communications by Christopher Cox Whitepapers 1. Nomor research: Whitepapers on LTE-A http://www.nomor.de/home/technology/whitepapers 2. 3GPP: Industry whitepapers http://www.3gpp.org/Industry-White-Papers 3. NTT Docomo Technical Journal vol.12 no.2: Special articles on LTE-Advanced technology http://www.nttdocomo.co.jp/english/corporate/tech nology/rd/technical_journal/bn/vol12_2/index.html 4. LTE-Advanced: An Operator Perspective, Prakash Bhat et al., IEEE Communications Magazine, February 2012 5. Overview of enabling technologies for 3GPP LTEadvanced, Tran et al., EURASIP Journal on Wireless Communications and Networking 2012, 2012:54. Websites 1. 3GPP: LTE-Advanced - http://www.3gpp.org/lteadvanced 2. 4G Americas: LTE-Advanced http://www.4gamericas.org/index.cfm?fuseaction=pa ge&sectionid=352 3. 3G4G: LTE-Advanced (IMT-Advanced) http://www.3g4g.co.uk/LteA/
Mobile Europe Insight Report | 23

????????

INSIGHT REPORT

INDUSTRY LTE-A VIEWPOINTS

HAT DO YOU THINK ARE THE CHIEF TECHNICAL CONSIDERATIONS OF THE 3GPP RELEASES THAT MAKE UP LTE-A?

account are Carrier Aggregation, HetNet technology and User Data Convergence. (Avinash Joshi - Director,New Business Unit (LTE),CTO Office, Tech Mahindra) LTE-A is a paradox. LTE was called evolution but was a dramatic step. People are positioning LTE-A as though it were another big step, but this time it is an evolution. Or rather, it is a portfolio of different evolutions: more MIMO, Carrier aggregation & COMP. Of these CA is the most immediately useful and valuable. COMP is very clever and potentially most important but it will take longer to get solved and deployed.(Rupert Baines, VP Corporate Strategy & Marketing, Mindspeed) WILL LTE-A CHANGE ANYTHING, OPERATIONALLY OR COMMERCIALLY, FOR MOBILE OPERATORS? IF SO, WHAT? The primary commercial driver for LTEA is the continued need to deliver increased capacity while lowering cost per bit. Beyond that, LTE-A will not bring about any fundamental operational or commercial changes for operators. This is an evolution, not a revolution. (Manish Singh, CTO, Radisys) LTE-A capabilities will be used to further add capacity to manage the ever increasing data usage in all regions and operators of the world. The capacity increase and flexibility in operating models for managing capacity is the main area that operators are looking for. (Per Kangru, LTE technical expert, JDSU) LTE-A requires that operators swap-out or upgrade previous infrastructure. In order to control cost, this will initially be done in limited areas and will result in islands of LTE-A coverage. Maximisation of cost-effectiveness means that these LTE-A sites need to be surgically placed where the benefit will be greatest. As such, this constitutes an important aspect of heterogeneous network (hetnet) activity involving location-aware, subscriber-

LTE-A provides features like multicarrier aggregation over different frequency bands, coordinated multi point transmission, multi carrier aggregation UMTS+LTE, improved MIMO Mode. The capability to enable carrier aggregation for more than 20 MHz is key for higher data rates.Refarming of existing spectrum enables us to use additional spectrum for LTE-A. We may consider further LTE Frequency planning (fractional frequency reuse if necessary) to reduce interference. LTE-A will providehigherdata ratesover the entire cell and increasedata rateat the cell edge. Multicarrier aggregation over different frequency bands will enable more flexible network design. In order to use full LTE-A capabilities one would have to upgrade existing Antenna installations from MIMO 4x4 up to 8x8 (Static, authorities, costs, time). (Armin Sumesgutner, Director Network Planning Telekom Austria Group) There are many technical considerations within LTE-A. Hetnets and small cells are of primary importance, but carrier aggregation, CoMP and SU-MIMO are also important, particularly when you consider the user equipment updates that these will require. (Jens Voigt, Principal Research Engineer, Actix) Achieving LTE-As promised peak data rates of 3 Gbps over a mobile network poses one of the biggest challenges the industry has ever had to face. Higher data rates are achieved through higher bandwidths, by using carrier aggregation and evolved antenna configurations. In order to support highly intensive data sessions, LTE-A can utilise MIMO transmit and receive antennas in both base stations and devices. (Paul Beaver, Products Director, Anite) The chief considerations to take into
24 | Mobile Europe Insight Report

centric analysis & decision-making. Operators will also need to manage inter-Radio Access Technology (IRAT) handovers where subscribers transition from the LTE-A coverage areas to UMTS & GSM as seamlessly and effectively as possible. (Dr. Michael Flanagan, CTO, Arieso) LTE-A is likely to evidence a generational gap in applications delivery. Mobile operators face another significant competitive inflection point. Disruptors, LTE-A applications and services start-ups, are gifted with a marvellous commercial opportunity. (Robert Marcus, CEO, QuantumWave Capital) Bringing more out of HetNets is one of the major component of LTE Advanced. So, mobile operators, in addition to deploying macro networks, also have to focus on deploying small cells, and increase their density, in accordance with the increase in data traffic. (Qualcomm) With LTE-Advanced, operators will see significantly reduced operating costs that they will be able to use to address broader customer bases and market needs. (Andrew Green, VP Marketing, Mobile Computing, Sierra Wireless) The most obvious advantage is the higher capacity and data rates which can be provided towards our customers.This gives us the potential for upselling and affirm our Best Network proposition. (Armin Sumesgutner, Telekom Austria Group) On the commercial front, operators will gain the ability to further extend products and service offerings into the business sector, which to date has been dominated by fixed network centric applications. From an operational perspective, migration to LTE-A will increase network management complexity, and the deployment of business applications and media services will further add to the explosion in data traffic. (Paul Beaver, Anite)

INSIGHT REPORT

WHAT KEY CHALLENGE, OR OPPORTUNITY, THAT WILL OCCUR AS OPERATORS INTRODUCE LTE-A CAN YOU HELP WITH? At Tech Mahindra we provide an array of services for Carrier Aggregation including device testing, network optimisation and managed services.(Tech Mahindra) Backhaul elasticity is critical to the viability of HetNets. Operators have to focus on radically enhancing backhaul supply to meet increasing and unpredictable data demand. This means looking beyond existing technologies that just meet this demand to those that fully maximise the significant 4G opportunity, overcome spectrum limitations, and are cost efficient. (Shayan Sanyal, CCO, Bluwan) Through increased automation of network analysis, management and optimization, the industry can come closer to operating in line with the SelfOrganizing Networks (SON) concept and Actix can provide the unified, vendorindependent and multi-radio technology solutions to enable this. (Actix) Customers will demand a high Quality of Experience for this evolution in wireless technology. Anite is uniquely placed to help as we continue to lead the way in LTE signalling conformance. Anite is responsible for the verification of over 80% of the LTE Release-8 protocol tests and offers the greatest number of unique Global Certification Forum validated test cases. (Anite) Where 3G worldwide roaming could be achieved with 4 or 5 bands, the standards now describe 40+ specific bands for LTE use. Carrier Aggregation adds to this the need to support combinations of bands. Finding effective RF solutions for the spectral requirements of each customer is a huge challenge that were working hard to solve. (Sierra Wireless) The first key challenge will be liberating LTE-A-capable devices from UMTS & GPRS networks.The second key challenge is managing the hetnet interactions of LTE-A

with the more ubiquitous LTE, UMTS & GPS network. (Arieso) LTE-A upgrades of existing macro sites will in most cases require some sort of modifications to the RBS during deployment including antenna adjustments etc. JDSUs tools can ensure that this is delivered in a smooth and effective way and ensure that the parts have been fully qualified in the lab to ensure no negative operational impact due to issues with regression on legacy (LTE) functionality. (JDSU) We have and continue to develop the advanced interference management techniques which are part of the HetNet component of LTE Advanced. We are also the first to announce chipsets supporting carrier aggregation. (Qualcomm) Ixia will evolve our testing to address LTE-A features as they are introduced. Ixia is currently focused on Carrier Aggregation support. (Joe Zeto, Ixia) The rollout of LTE-A overlaps with the critical need for more capacity and for increased CapEx efficiency. SoCs will (as in other markets) increase competition and drive down cost, transforming the wireless infrastructure market. (Mindspeed) WHEN WILL WE SEE COMMERCIALLY AVAILABLE EQUIPMENT (NETWORK AND DEVICE) AND COMMERCIAL LTE-A NETWORKS/ SERVICES? This will vary by region as a factor of local competition and local spectrum access. US will see LTE-A networks launch in early 2013 if not in late 2012. (JDSU) ZTE announced LTE-A base stations in February 2012, and Huawei announced trials with Tele2 and Telenor in Sweden in May 2012.Less has been publicly stated about device readiness for LTE-A, although test vendor Rhode & Schwarz announced Rel.10-related gear in August 2012.So, Id say the forecast for 2013 is cloudy with a chance of LTE-A.Clearwire LTE-A rollout announcements suggest brisker action in the 2014 timeframe. (Arieso) We expect there to be multiple

product launches at MWC 2013 and that these products will mature in the latter part of 2013. We also expect much of the focus to be in carrier aggregation, relay and the SON enhancements in the 1st phase of product launches. (Radisys) The industry is gated behind the R&D evolution of hardware vendors and chipset manufacturers, the operators won't move until their development roadmaps becomes clear. I don't believe LTE-A will be mainstream until the middle of this decade." (QuantumWave Capital) Many of the LTE networks deployed are already LTE-A ready.Both SKT and ATT have said they will have LTE-A next year.This is may not be fully commercial, but the first instances of LTE-A eg with CA to bond small channels will be deployed quite quickly. (Mindspeed) It is possible some specific operators may skip LTE and go straight to LTE-A, or some specific broadband to home deployments, this is possible from late 2014. (Ixia) Were likely to hear lots of news about LTE-Advanced in 2013 and Id expect to see significant launches by the middle of 2014. (Sierra Wireless) We expect a commercially launched LTE-A networks/ services for 2014/15 andcommercially available CPEs for 2015/16. (Telekom Austria) Some LTE-A features will likely see deployment in 2012, but industry estimates suggest that upgrade migration will start during 2013 with the availability of the first production grade handset, and continue to roll out through 2014/15 and beyond. (Anite) HetNets are already available today, with the active deployment of small cells in high traffic areas. I estimate that this will be augmented in the next 12 months with the arrival of carrier aggregation and, finally, the implementation of CoMP or SU-MIMO technologies three years from now. (Actix)
Mobile Europe Insight Report | 25