LTE: Testing of the Physical Layer

By: By Moritz Harteneck, Rohde & Schwarz
( 1 Sep 2009 )

Long term evolution (LTE) is the next step in mobile communications and aims at making wireless access technology ubiquitous. Manufacturers of mobile phones and mobile infrastructure are working on a number of improvements of the current UMTS technology to enhance the users’ experience, but also to drive down cost by reducing capital as well as operating expenses. They achieve this by measures that include high data rates (in excess of 100Mbps), shorter user plane latency (less than 5msec), shorter control plane latency (less than 50 and 100msec to change from a camped or dormant state to the active state) and higher spectral efficiency (three to four times of HSPA). With the first set of specifications available for the new 3GPP air interface, the challenges of testing LTE have become clearer. Since commercial roll-out is planned for 2010, the first test systems are already available and are currently enhanced. However, there are a number of new challenges that engineers are facing when testing the physical layer, but respective testing features are already available.

The air interface of LTE is based in the downlink on orthogonal frequency domain multiple access (OFDMA) and in the uplink on single carrier frequency domain multiple access (SC-FDMA) technology. Both use the concept of densely stacked low data rate subcarriers to convey the information. Since every subcarrier has a very low bandwidth (15kHz), the experienced channel can be considered flat fading. In addition to this, the use of a cyclic prefix combats intersymbol interference. This technology, however, makes the physical layer inherently frequency selective and consequently, all procedures defined in the specifications are also frequency selective. Engineers have to keep this in mind when performing tests on wireless devices.

To boost the data rate of the physical layer and to make it more robust, LTE uses multiple input multiple output (MIMO) schemes in conjunction with the OFDMA scheme. This adds another level of complexity to the testing challenge. As 3GPP defines various modes of operation from the MIMO bouquet – such as single-user MIMO, multi-user MIMO, open and closed loop transmit diversity – the number of different setups increases. Thus, automated testing becomes a must-have to be able to verify the correct operation of the complete system.

When testing the physical layer, engineers typically increase the coverage in accordance with the integration status of the system. It starts with low level block testing verifying the receiver and transmitter of the system. Further processing is added step by step until all required functional blocks are integrated.

During the testing campaign, it is important that engineers are able to observe the internal signals such as payload transmitted and received or measurements of the timing offset between the uplink and downlink frames. Otherwise, the debugging process could be hindered because relevant data is unavailable. Furthermore, the system should be easy to use and not require substantial configurations outside of the layer under test. For example, users should not have to configure the packet data convergence protocol (PDCP) layer when testing the physical layer. Respectively, when testing the radio link control (RLC) layer, configuring the physical layer should be as easy as possible.

Since most of the testing is done concurrently with the development and debugging process, it is extremely important that engineers are able to automate the testing and embed it into a regression environment with as little manual intervention as possible. This should include all components that are used in the testbed such as fading channel simulators and external noise and interference sources. Engineers have to go through several steps in the testing campaign which are described in the following in more detail, along with features of the test system for efficient physical layer testing.

DATA PATH TESTING
The first step within the testing campaign is to make sure that the individual channels operate in an open loop fashion. This is to validate the correct implementation of 3GPP specifications 36.211 and 36.212 which define the downlink and uplink transmission with the forward error correction. In order to verify the correct reception and transmission to and from the user equipment (UE), the tester provides a downlink signal to the UE and/or receives an uplink signal. During this phase, intermediate points in the encoding and decoding chains of the test equipment must be visible to assist in the debugging process. Also, special test features such as the ability to corrupt downlink transmissions are required. Frequently, off-the-shelf signal generation and analysis equipment, such as the SMU, SMJ and FSQ from Rohde & Schwarz, are used during this phase.

FUNCTIONAL TESTING
Once the engineers have established the operation of the individual data paths in a noise free environment, the functional testing phase begins: The engineers test the feedback procedures within the UE in a controlled and static environment. This makes the behavior of these procedures easy to predict. The procedures to be tested include channel quality indication (CQI), hybrid ARQ (HARQ) and timing control, which are all defined in the 3GPP specification 36.213. During this phase, the tester has to respond in real time to the control information, i.e. for testing the downlink HARQ operation, the tester has to adjust the downlink transmission in accordance to the received ACK/NACK information. To test these features efficiently, the tester has to provide special test features as built-in interference or channel models.

Figure 1 shows the block diagram of a test system for functional testing of an LTE UE. The grey blocks mark the special test behavior – for instance, functionality that is not specified within the 3GPP set of specifications. Most notable are the channel models which apply static flat channels on a per-subcarrier basis to enable the test engineer to test channel estimation and CQI reporting in a deterministic and repeatable environment.

Other special test features are the downlink scheduling which can be used to test HARQ processes; the uplink (UL) scheduling to stimulate uplink transmission; the timing control to verify that the UE obeys the timing commands sent within the PDSCH payload; and the UL power control to verify the correct behavior of the power control algorithms in the UE. In addition, extensive logging and measurement capability is needed to validate the UE implementation. Figures to be measured are for example uplink power per subframe, uplink transmission quality, and UL/DL timing offset and throughput. OCNG is required to simulate transmissions to others UEs that are typically present on the air interface as well.

PERFORMANCE TESTING
After a satisfactory functional testing, the engineers have to check the performance of the UE in a next step. This is to validate the requirements of 3GPP specification 36.101, which defines performance requirements for the user equipment. This step is divided into two parts: first is the performance of individual blocks such as receiver and transmitter performance, second is the system performance (i.e., the performance including closed loop operation and UE procedures).

Figure 2 shows the block diagram of a test system for performance tests of a LTE UE. The engineers have to substitute the static channel models with a fading channel simulator. Additionally, they have to add neighboring channels, blockers and interferers to the downlink signal in order to simulate an environment that the UE would experience in the field. While doing the block level performance test for the PDSCH, the engineer wants to see block error rate (BLER) curves varying the signal power, type and level of interference, the chosen transport format and the fading channel profile. During block level tests on the PDCCH, the engineer is interested in false detection ratios to characterise the performance of the blind detection algorithms in various interference scenarios.

Once the block level testing is done, testing moves on to a system level testing in order to validate the interaction of various system components. Required tests are throughput measurements with enabled HARQ operation (similar to the test described for HSDPA in Section 9.2 of 3GPP specification 25.101), CQI validation tests (similar to the tests described in Section 9.3 of 3GPP specification 25.101) or PDCCH detection tests (similar to Section 9.4 of 3GPP specification 25.101). In parallel, the engineers can verify the compliance of the uplink transmission with 3GPP specification 36.101, checking whether it fulfils the required spectral emission masks and the minimum quality requirements of the uplink transmission such as EVM and spectral flatness.

PRODUCTION TESTING
After the UE has finished all the described R&D testing phases, it enters production. Some of the tests that have been developed by the design team are ported to the production environment. Therefore, it is a desirable feature when the same test platform and environment are usable for development as well as production testing. This guarantees a high re-use – teams can transfer knowledge among each other and thus assist in solving problems. The requirements for production tests, however, differ from those for development tests. For instance, while it is less important to observe the internal workings, clear pass/fail criteria are crucial as well as keeping the test times as short as possible.

The LTE protocol testing solution based on the R&S CMW500 from Rohde & Schwarz is designed to meet the LTE testing challenges described in this article. With its modular concept and the built-in signal generating and analysing facilities, it is well suited to cover current testing needs for LTE. It is designed to cover a number of future test challenges, which include physical layer testing for MIMO and other advanced physical layer technologies which will be defined in the upcoming 3GPP releases. Moreover, the R&S CMW500 can be configured for testing of higher layers such as medium access control (MAC), radio link control (RLC), packet data convergence protocol (PDCP) and radio resource control (RRC). Inter-RAT handover to other radio access technologies such as GSM, W-CDMA are on the roadmap as well.

About the author
Moritz Harteneck has been working at Rohde & Schwarz headquarters in Munich since 2007. After completing his MSc and PhD at the University of Strathclyde in 1994 and 1998, respectively, he has been working in the area of wireless communication systems, focusing mainly on the development and testing of physical layer implementations.

Click here for the illustrations:

Figure 1, Figure 2, Figure 3