LTE in FOCUS

BY REINER GOETZ, ANNE STEPHAN, AND MEIK KOTTKAMP, ROHDE & SCHWARZ -- EDN Europe, 01 May 2010

As part of the worldwide 3GPP Release 8 standard, LTE has been fully defined since March 2009. This new technology is an essential enhancement of classic mobile-communications technologies, such as GSM/ EDGE, WCDMA/HSPA(+), and CDMA2000/Ev-Do Rev A. It promises end users significant data-rate, network-capacity, and latency improvements. LTE will also enable network operators to more effectively and inexpensively deliver services, such as Web browsing, gaming, and video streaming, and it will open the door for new mobile services.

Consequently, no fewer than 59 operators in 28 countries have announced plans to deploy LTE on their networks. A sizable number of commercial networks will roll out in 2010. Therefore, vendors of LTE products and infrastructure are performing extensive testing ranging from lab tests on individual devices to small-scale field tests with just a few base stations and user devices from a variety of vendors. Major test networks, too, with numerous base stations and a significant number of initial LTE user devices, are undergoing user trials.

What requirements does this new technology have to meet, and how can you assess and effectively verify LTE perform ance in a laboratory environment by means of suitable tests? Consider the differences between LTE and other technologies before proceeding to examine the various phases in the development of LTE-enabled infrastructure and end-user equipment.

SIMILAR BUT DIFFERENT

The 3GPP participants developed LTE on the basis of available technologies. Therefore, it is hardly surprising that, despite its numerous differences, LTE has a lot in common with them. Consider HSPA(+), an established technology in packet-oriented services. In LTE, just as in HSPA(+), the allocation of resources for the transmission of data from the base station to user equipment relies on a rapid feedback mode in the user device. The device ascertains the quality of the transmission channel and informs the base station what maximum resource size to allocate. The only difference is that LTE offers a much faster feedback mode than does HSPA(+): LTE uses a 1-msec time-transmission interval, whereas HSPA(+) employs a 2-msec TTI and WCDMA uses a 10-msec TTI, allowing the data rate to adapt to current transmission conditions practically every millisecond.

The biggest differences between LTE and current 3GPP standards lie in the technologies that manufacturers use to implement the air interface. LTE employs OFDMA and MIMO. In addition, LTE works with a flat IP-based network architecture. OFDMA enables granular resource allocation because LTE uses a large number of narrowband subcarriers with a bandwidth of 15 kHz, compared with 200 kHz in GSM and 5 MHz in WCDMA (Figure 1). This enhancement combines with a maximum channel width of 20 MHz and the ability to work with as many as four transmitting and receiving antennae to create the basic conditions necessary for meeting high-data-rate and high-capacity requirements.

AT A GLANCE

LTE promises end users significant data-rate, network-capacity, and latency improvements.

The biggest differences between LTE and 3GPP standards lie in the technologies to implement the air interface.

LTE will find use in user devices alongside technologies such as WCDMA, CDMA2000, and GSM.

In RF-signaling tests, testers examine the user equipment’s transmitter and receiver in combination with all of the signaling layers.

The use of OFDM, which enables TTI-based allocation of resource blocks, leads to significant changes in testing requirements.

MIMO requires extended antenna systems at the base station that you must verify with the aid of signal analyzers.

LTE also greatly simplifies the mapping of logical channels onto physical channels. Shared channels have replaced dedicated channels, and LTE has fewer MAC-layer entities and RRC states for greater simplicity. In contrast, the number of parallel processes in the protocol has increased, and you can use MIMO to combine multiple data streams. The encryption function has also changed: In LTE, the eNodeB and MME use different keys. The data in the PDCP layer and the NAS layer are encrypted differently; in WCDMA, the NAS layer is not encrypted at all.

HIGH SPEED, COMPLEXITY

The way in which LTE integrates into networks also plays a crucial role. From the outset, developers of LTE user equipment have been under enormous time pressure. Even before LTE’s core specifications were complete, equipment makers augmented the then-unfinished specifications based on their own assumptions and implemented their own protocol “dialects.” Their aim was to be in a position to demonstrate LTE’s ability to function and its benefits as soon as possible. Now that the developers have finalized the specifications to a sufficient degree, the focus is on reducing development time so that they can deliver LTE products to the market as swiftly as possible.

LTE’s higher complexity poses significant challenges because, for the most part, LTE resides in user devices alongside technologies such as WCDMA, CDMA2000, and GSM. The presence of multiple technologies entails a variety of hand-over scenarios, all of which designers must test. In addition, LTE user devices must support other noncellular standards—Wi-Fi, GPS, and Bluetooth, for instance—that complement the wide operational coverage that cellular technologies afford.

Ideally, to enable steps in the development process for a mobile phone to take place concurrently, developers pursue an approach employing reusable modules. However, the developers must test the various components as early in the process as possible to minimize the number of potential errors during integration and to avoid creating problems in subsequent field tests. The testand- measurement equipment that they use must therefore be able to separately drive or bypass layers or functional modules. The modular approach, for example, enables developers to test the functions that they loaded in software onto a baseband chip in a virtual environment without relying on the availability of the hardware.

LTE GLOSSARY

ACLR: adjacent-channel-leakage ratio
AWGN: additive white gaussian noise
BLER: block-error rate
CDMA: code division/multiple access
CPRI: common public-radio interface
CQI: channel-quality indicator
CS: circuit-switched
EDGE: enhanced data rates for GSM evolution
eNodeB: base station
ETSI: European Telecommunications Standards Institute
Ev-Do: evolution-data optimized/evolution-data only
EVM: error-vector magnitude
GCF: Global Certifi cation Forum
GPS: global positioning system
GSM: global system for mobile communications
HSPA: high-speed packet access
I/Q: in-phase/quadrature
IP: Internet Protocol
LTE: long-term evolution
MAC: medium-access control
MIMO: multiple input/multiple output
MME: mobility-management entity
NAS: nonaccess stratum
OBSAI: Open Base Station Architecture Initiative
OFDMA: orthogonal frequency-division multiple access
PCS: personal communications service
PDCP: Packet Data Convergence Protocol
PMI: precoding matrix indicator
PRBS: pseudorandom bit sequence
PS: packet-switched
PTCRB: PCS Type Certifi cation Review Board
QAM: quadrature-amplitude modulation
QPSK: quadrature phase-shift keying
RI: rank indicator
RLC: radio-link control
RRC: radio-resource control
RRH: remote radio heads
16QAM: 16-phase QAM
64QAM: 64-phase
QAM SMS: short-message service
3GPP: Third Generation Partnership Project
TTCN2: Testing and Test Control Notation Version 2
TTCN3: Testing and Test Control Notation Version 3
TTI: time-transmission interval UL
CQI: uplink CQI
U plane: user plane
WCDMA: wideband CDMA
Other LTE features at the center of development efforts are the data rates on uplinks and downlinks. Just as in HSPA(+), they place considerable demands on the user equipment. To assess performance, manufacturers must run tests that evaluate the acknowledgments and negative acknowledgments on the RLC and MAC layers. Designers can achieve the specified data rates using the wide variety of MIMO modes. However, the test-and-measurement equipment must support these modes. Besides the ability to connect multiple antennae, the test equipment must also be able to simulate fading channels. Only then is it possible to test the functioning of receivers under realistic conditions.
Scalable bandwidths and the 17 regional frequency bands further increase the breadth of required testing. And developers must also address limitations, such as different possible power levels in tests at high bandwidths on frequency bands with low separation between receiving and transmitting frequencies.
This high complexity calls for frequent regression tests, such as testing daily software builds or the performance of endurance-testing scenarios in a realistic signal environment. Test equipment with advanced automation and remote capabilities makes it easier to efficiently conduct these tests. The range of measurements is not confined to verifying that developers have correctly implemented the specifications; it also extends to modules’ stability and robustness when facing varying interpretations of specifications (Figure 2).
At the end of the development process comes conformity testing on certified test systems, during which developers run a selected number of tests from the 3GPP conformance-test specification on the finished user device. Unlike functional tests, this type of testing formally verifies the protocol layers and the requisite RF performance.
MOBILE USER DEVICES

Testers primarily conduct pure RFperformance testing using signal generators and analyzers; combined RF-signaling tests are also important. These tests examine the user equipment’s tran smitter-and-receiver combination with all of the signaling layers. The tests are simulations that closely approximate signaling procedures and scenarios in live use under realistic conditions, with possible interfering signals, and during continuous operation of the device. The primary emphasis is not on testing the actual signaling procedures; rather, the signaling serves as a means to an end for performing realistic tests on the entire device. Testers perform separate tests for the transmitter and the receiver.

Testers apply a variety of measuring methods during the transmitter tests. First, they test LTE signals using proven methods—power and EVM measurements, for example—that they adopted from other mobile-communications technologies. Second, they verify extensive procedures, such as power control based on profiles in LTE and WCDMA. Many of the measurements may resemble well-known procedures. With LTE, they are more complex, however. A spectrum measurement is a case in point: The fact that LTE and WCDMA frequency bands may be adjacent to each other places exceptional demands on the user equipment. To help prevent interference between neighboring WCDMA and LTE systems, the transmitting power in adjacent bands must not exceed either LTE- or WCDMAspecific limits (Figure 3). An extended ACLR test can check that it doesn’t.

The use of OFDM, which enables TTI-based allocation of resource blocks, has led to significant changes in testing requirements. The measuring equipment must flexibly configure the requisite assignment tables and scheduling parameters for the uplink and downlink and send these tables and parameters to the user equipment. Meanwhile, the testers must check the correct allocation of resource blocks and the transmitting characteristics of the user equipment on the uplink (Figure 4).

Given that multiple user devices can concurrently reside within the available bandwidth, testers must measure inband emissions to determine whether the user device complies with allocation and transmitting-power requirements on the uplink. This approach ensures that the device does not interfere with other uplink signals outside its allocated resource blocks. Measurement equipment that can flexibly set limits and independently check limits greatly simplifies testing (Figure 5).

Because of the breadth of allocation options available, testing generates a large number of results. These results depend extensively on the location and size of the allocated resource blocks within the time and frequency domains, and developers must therefore interpret them in context. In addition, some RF impairments have an effect only on certain allocations.

The distribution of transmitting power across multiple subcarriers can lead to power differences between subcarriers. You can examine transmitting power at the subcarrier level by testing spectrum flatness, thereby enabling users to identify potential fluctuations with exceptional precision.

In receiver tests, the MAC layer uses localized acknowledgment/negativeacknowledgment- based BLER methods. These methods for analyzing uplink signals are familiar from HSPA. With LTE MIMO, the focus is on a scenario in which you apply various fading profiles to the downlink signal. To reduce development time and costs, you can use static channel models that simulate a static fading profile instead of dynamic fading profiles. The static models enable you to analyze the effects on receiver behavior using the BLER methods. In HSPA, too, you measure the downlink signal with fading and AWGN. In LTE, other technologies inside and outside the LTE band cause additional interference signals, calling for wider blocking tests and adjacent-channel tests.

The follow-UL-CQI test, again familiar from HSPA, is an important means of adjusting the signaling parameters and thus optimizing the receivingsignal quality that a user device reports through the CQI. Several values affect the quality in LTE, including Rank 1 or Rank 2 CQI, PMI, and RI values. Dynamically changing parameters on active connections in the measuring equipment can help to save time during these tests.

Testing user devices calls for a range of measurement methods suitable for checking the transmitter’s RF in combination with the allocated resource blocks on the uplink. Ideally, data for computing transmitter measurements should originate from a test sample and appear simultaneously in a clearly structured form (Figure 6).

PERFORMANCE TESTING

All LTE user devices to date have been data devices—in other words, USB sticks and PC cards. The data-services sector has mainly driven the motivation for introducing LTE. Nevertheless, the continuing debate over technical alternatives for the voice service and the fact that a number of special-interest groups focusing on it have formed within the industry substantiate the emphasis on equally efficient voice support. It remains to be seen which of the alternatives will gain primacy. However, data services pose greater requirements than the voice service as far as protocol test-and-measurement systems are concerned.

In LTE, only the PS domain exists, not the CS domain. In general, multiple services with different bearers operate in parallel in a manner comparable with WCDMA multicall services. Furthermore, once users power up and register their devices, the devices immediately have always-on status and can almost instantly issue requests to transmit data on the uplink or the downlink. Consequently, functional tests spanning all layers always require the measuring equipment to provide a service that delivers data through the U plane.

A mobile device’s performance is of direct relevance for end users. How fast is the data rate, and what is the latency when starting services? To what extent does performance degrade when reception is poor? Does the manufacturer guarantee that the device will interoperate with different base stations? When finding answers to these questions and optimizing a data device, it is not enough simply to check signaling procedures or individual values that you measure at the IP level. Rather, it is important to analyze bottlenecks in the protocol layers: Which level is causing unnecessary retransmissions? Why does the BLER increase under certain conditions? Protocol test-and-measurement equipment must answer these questions and verify signaling procedures. Thus, the lines between classic application tests and protocol tests are becoming increasingly blurred.

THE SWIFT AND EFFICIENT DEVELOPMENT OF LTE BASE STATIONS REPRESENTS A CORE CHALLENGE FOR INFRASTRUCTURE VENDORS.

If you test modules during development, the test equipment must provide the necessary interfaces. In the past, it may generally have been sufficient to connect to the user equipment through RF, but it is now essential to provide interfaces on the I/Q baseband because the protocol software runs on the baseband chip or on a chip emulation. It is even possible to completely test a protocol stack without hardware if you replace the physical layer with emulation software. Access to details of lower-level protocol-layer configurations is essential for meeting all these requirements.

Manufacturers will only gradually roll out LTE networks. The rollout will require thorough testing of the handover signaling, as it is important to ensure that user equipment can transition smoothly between technologies. Thus, it is essential for test-and-measurement equipment to provide a basic implementation of all technologies and support for synchronization with LTE. Given that MIMO plays a central role in enabling higher data rates, you must test complex signaling procedures and user-device feedback.

CONFORMANCE TESTING

GCF certification of LTE user equipment should begin in late 2010. Validation of the first test cases employing 3GPP specifications 36.521 and 36.523 has occurred. ETSI has chosen TTCN3 as the language for describing the tests. TTCN3 is an enhancement of TTCN2—the language used with WCDMA—and now has more in common with a traditional programming language such as C++. TTCN3 is therefore easier to learn, and adjacent areas of development, in addition to certification tests, will likely adopt it.

In addition to the essential userequipment certification that GCF and PTCRB perform, network operators must also perform certification to their own high standards. These tests place greater emphasis on the characteristics of the network infrastructure and on optimization of the available mobilecommunications services.

A CHANGE OF PERSPECTIVE

The swift and efficient development of LTE base stations represents a core challenge for infrastructure vendors. As a rule, they deploy test systems well before the commercial rollout of networks. If possible, these systems should run on the network infrastructure alongside commercial platforms during the test phase. This is why vendors conducted numerous LTE field trials in 2009. To determine which tests to perform during the base-station-development cycle, vendors drew on experience that they gathered over many years of successfully operating mobile-communications systems, such as GSM and WCDMA. Table 1 lists the measurements that they performed on transmitters and receivers.

Certain technical aspects of LTE are of special importance in the development of infrastructure products. MIMO requires extended antenna systems at the base station that you must verify with the aid of signal analyzers. To measure a precoded MIMO signal from two transmitting antennae, you must record both data streams at once. LTE technology uses complex precoding matrices on the transmitting branch. Measuring module parameters, such as EVM, for example, generally requires information from both transmitting signals. This requirement calls for a modular-test-equipment setup with two connected analyzers. The measured values from the first signal analyzer transfer to the second on a master/slave basis.

In addition, LTE employs shared frequency channels that multiple user devices access simultaneously. These devices may use different data rates and, therefore, different modulation types, including QPSK, 16QAM, and 64QAM. In addition, a base station’s transmitting signal comprises user data, reference- and channel-estimation information, and signaling data, and these elements combine into a composite signal (Figure 7).

Furthermore, MIMO requires precise time alignment for transmitting signals. The 3GPP test specifications therefore now include a test to ensure that the signals of two or more antennae are time-synchronized with an accuracy of at least 90 nsec. Once you verify this requirement, you combine the MIMO signal from each base-station antenna on the RF and apply it to a signal analyzer’s input.

In receiver testing, users need to apply standards-compliant LTE signals, along with various propagation models, to the base-station receiver path. As a rule, this approach employs PRBSs, and the base station can reconstruct these PRBSs if they are of a known length. By means of a simple comparison, you can obtain the error rates that enable the base-station receiver’s performance for verification under a variety of simulated propagation conditions.

Because the receiver tests cover a diversity of interference scenarios and propagation conditions, signal generators must be able to generate reference signals. It can be advantageous if the equipment can flexibly combine signals and if you implement reference channels and specified propagation-channel models, thereby enabling users to quickly and flexibly configure scenarios and greatly simplifying error detection.

In recent years, RRHs have become common features in the design of base stations. In designs of this kind, the RF components and the base-station amplifier reside in a remote front end directly on the antenna. The advantage of this approach is that it avoids line loss on the RF cabling connecting the antenna, thus increasing the base station’s available output power. The baseband signals proceed to the RRH over an optical connection. Two digitalinterface standards, CPRI and OBSAI, exist for WCDMA base stations. Besides increasing modularity in designs, the standardization of this interface eases error identification during the development cycle. It also allows you to combine modules from different vendors. Furthermore, this setup allows you to connect multiple RRHs to a base station’s baseband, which can help to achieve optimum coverage in buildings, for example.

Preparations are currently under way to standardize an optimized interface format for LTE. The digital interface between baseband and RF requires test-and-measurement equipment—in particular, signal generators and signal analyzers—to support this format. This requirement means that you can individually verify baseband or RF modules. As a rule, a converter module translates the measuring equipment’s digital baseband language into the standardized format.

To support rapid and flexible resource allocation, the base station assigns the user device a certain channel capacity, including both bandwidth and modulation type, based on various parameters, such as the network cell’s available capacity. The base station also acknowledges each packet that it receives correctly; in other words, it tells the user device whether it must retransmit a data packet or can proceed to transmit a new packet. To verify correct operation of the control mechanism in the base station’s receiving branch during receiver testing, signal generators must emulate the user device’s transmitting signal and provide a means of interpreting the base station’s feedback. The feedback proceeds to a separate input on the signal generator; the signal generator then decides in real time whether to request retransmission of the same packet or of a new packet.

Although LTE is simpler in some ways than its predecessor technology, WCDMA, LTE user equipment is more complex overall because it incorporates additional procedures, such as MIMO, and this complexity places tougher demands on test-and-measurement systems. LTE can adopt a number of types of measurements from the world of WCDMA, but LTE involves new meas surements and a wide range of parameterization, on both the RF side and the protocol side. LTE base-station testing uses expanded connection options, for example, to verify fast feedback procedures. These procedures are just as necessary as MIMO signal generation and signal analysis, which vendors implement using multipath approaches. When testing hand-over to earlier technologies, multitechnology platforms offer huge benefits; with a view to the future, these platforms are sound investments.

Author Information

Reiner Goetz is a product manager at Rohde & Schwarz. His responsibilities include marketing and sales of mobileradio- communication testers. Goetz holds a degree in communications engineering from the University of Applied Sciences (Esslingen, Germany) and a master’s degree in business administration in international marketing from the University of Reutlingen (Germany).

Anne Stephan is a project manager in the test-and-measurement division of Rohde & Schwarz, where she is responsible for the development of WCDMA and LTE-signaling software for mobile-radio-communication testers. She holds a degree in computer science from the University of Saarbruecken (Germany).

Meik Kottkamp is technology manager in the test-and-measurement division of Rohde & Schwarz, where he is responsible for strategic marketing and product-portfolio development covering the enhancements in 3GPP technologies. He holds a degree in high-frequency electrical engineering from the University of Hannover (Germany).

There are no related contents to this article at the moment.

LTE in FOCUS

There are no related contents to this article at the moment.