At present, after 3G, how various communication technologies will evolve is a focus of the industry, especially for TD-SCDMA, whether the smooth evolution to the next generation of communication technologies can be achieved, which determines how long TD will have. the vitality of China, and how far my country’s independent innovation strategy can go. In November 2007, the 3GPPRAN151 meeting passed the proposal of LTE TDD fusion frame structure jointly signed by 27 companies, unifying the two frame structures of LTE TDD. The integrated LTE TDD frame structure is based on the frame structure of TD-SCDMA, which lays the foundation for the successful evolution of TD-SCDMA to LTE and even 4G standards.
TDD-LTE technical features
The LTE system supports two duplex modes of FDD and TDD. In these two duplex modes, most of the design of the system, especially the high-level protocols, are consistent. On the other hand, in the design of the bottom layer of the system, especially the design of the physical layer, due to the inherent differences in the physical characteristics of the two duplex modes of FDD and TDD, the LTE system has carried out a series of special designs for the working mode of TDD. These designs refer to and inherit the design idea of TD-SCDMA to a certain extent. We briefly describe and discuss these designs below.
radio frame structure
Because TDD uses time to distinguish uplink and downlink, resources are discontinuous in time, and a guard time interval is required to avoid transmission and reception interference between uplink and downlink, so LTE designs their own frame structures for FDD and TDD, namely Type1 and Type2, where Type1 is used for FDD and Type2 is used for TDD.
In FDD Type1, a 10ms radio frame is divided into 10 subframes with a length of 1ms, and each subframe consists of two slots with a length of 0.5ms. In TDD Type2, a 10ms radio frame consists of two half-frames with a length of 5ms, and each half-frame consists of 5 subframes with a length of 1ms, including 4 normal subframes and 1 special subframe. The normal subframe consists of two 0.5ms slots, and the special subframe consists of 3 special time slots (UpPTS, GP and DwPTS).
The most significant difference between TDD and FDD frame structures in LTE is that there is a 1ms special subframe in the TDDType2 frame structure, which consists of three special time slots: DwPTS, GP and UpPTS, whose meanings and functions are the same as those of TD- The SCDMA system is similar, in which DwPTS is always used for downlink transmission, UpPTS is always used for uplink transmission, and GP is used as the guard time interval for downlink to uplink conversion in TDD. , the total length of the three special time slots is fixed at 1ms, and their respective lengths can be configured according to the actual needs of the network.
Up and down time allocation
Another physical feature of TDD that is significantly different from FDD is that FDD relies on frequency to distinguish uplink and downlink, so its unidirectional resources are continuous in time; while TDD relies on time to distinguish uplink and downlink, so its unidirectional resources are in time. is discontinuous, and time resources are allocated in both directions.
The following figure shows 7 different uplink and downlink time ratios supported in LTE TDD, from “9:1” which allocates most resources to downlink to “2:3” which occupies more uplink resources. , the network can be configured flexibly according to the characteristics of the traffic. In this way, the inherent difference between TDD and FDD in resource composition becomes the reason why another part of LTE is specially designed for TDD. This part of the design mainly includes “physical layer HARQ related mechanism” and “random access channel using frequency division”.
Allowing multiple random access channels (frequency division) to exist at the same time is another design result formed by the structure of TDD uplink and downlink time division. In the design of LTE FDD, only one random access channel is allowed to exist at the same time, that is, the number of random access channels is only changed in the time domain. In TDD, time resources have been allocated in the uplink and downlink. At the same time, due to the existence of different uplink and downlink ratios, there may be a small number of uplink subframes (such as DL:UL=9:1), so in TDD It is necessary to support the random access channel of frequency division, that is, to provide multiple random access channels at the same time position by different frequencies, so as to provide sufficient random access capacity for the system.
In the case of FDD, the uplink and downlink resources are continuous in one direction, and the number of subframes is equal. Therefore, in the following example, when performing HARQ at the physical layer, a one-to-one correspondence can be established between downlink data and uplink ACK/NAK. In contrast to this, in the case of TDD, the resources in one direction are not continuous, so the corresponding time resources may not be obtained. In addition, the setting of the uplink and downlink ratios may cause the number of uplink and downlink subframes to be unequal, so a one-to-one correspondence cannot be established, so these need to be designed in a targeted manner. In LTETDD, in order to solve the above problems, the concept of MultipleACK/NAK is introduced, that is, one ACK/NAK is used to complete the feedback of several previous downlink data, which solves the feedback problem caused by the asymmetry of uplink and downlink time slots. On the other hand, the data transmission delay is also reduced, and the data does not need to wait for the next uplink time slot for feedback. Of course, the unnecessary excessive retransmission that this scheme may cause also needs attention.
The synchronization channel is another design that embodies a different duplexing scheme. The synchronization channel used for cell search in LTE includes “primary synchronization signal” and “secondary synchronization signal”. In the two frame structures, the synchronization signals have different positions: in FDDType1, the two synchronization signals are connected together and located in the middle of subframes 0 and 5; while in TDDType2, the secondary synchronization signal is located at the end of subframe 0, and the main synchronization signal is located at the end of subframe 0. The synchronization signal is located in the special subframe, that is, the third symbol of DwPTS. In the two frame structures, the absolute positions of the synchronization signals in the radio frame are different, and more importantly, the relative positions of the primary and secondary synchronization signals are different: in FDD, the two signals are connected together, while in TDD, the two signals are connected together. There is a time interval of two symbols between the signals. Since the synchronization signal is the first signal detected by the terminal during cell search, the design of different relative positions enables the terminal to detect the duplex mode of the network, ie FDD or TDD, at the very beginning of accessing the network.
random access preamble
The design of the random access preamble (Random Access preamble) is another special design of LTE for TDD. In LTE, the random access sequence adopts five random access sequence formats as shown in the following figure. The last random access sequence format is unique to TDD, and is also called “short RACH” because its length is significantly shorter than the other four formats. The reason for using the short RACH is also related to the design of the special time slot by TDD. As described in the figure, the short RACH is sent in the last part of the special time slot (ie UpPTS), so that the resources of this part are used to complete the uplink random connection. input operation to avoid occupying the resources of normal subframes. When using short RACH, one of the main issues that needs to be paid attention to is the coverage radius that its link budget can support. Since its length is much smaller than that of other formats of RACH sequences, its link budget is relatively low, which is suitable for coverage. Scenarios with a small radius (about 700m to 2km depending on the network environment).
R&S LTE TDD test solution
3GPP LTE is very different from the previous system in the air interface, so new requirements are put forward for testing. Based on the rich experience and leading position in the field of 3G testing, Rohde & Schwarz has been tracking and researching UMTS LTE since the early research and development stage, and has accumulated rich experience. R&D offers a complete line of testing products. These products include power meters, spectrum analyzers, signal sources, wireless comprehensive testers, protocol testers and RF compliance test systems. Equipment manufacturers can always rely on Rohde & Schwarz’s products and expert support. Rohde& Schwarz’s worldwide support network with professionally trained application engineers provides comprehensive customer support.
Since the development of the 3GPP LTE standard is not yet finalized, R&S maintains a high degree of flexibility in developing LTE options, and the software is regularly updated to ensure that the test instruments are based on the standard and the latest developments, making them compliant with 3GPP LTE requirements for future development. The following is an introduction to some important test items in the early research and development of LTE:
How to flexibly generate and analyze LTE RF and baseband signals,
How to test different MIMO modes,
How to test the protocol stack in the early stage of development to make it meet the requirements of consistency.
LTE signal generation
The test of LTE first needs to simulate the LTE radio frequency signal and study its statistical characteristics. For LTE downlink, researchers can refer to the RF characteristics of OFDMA from WiMAX and WLAN technologies. But for uplink, the SC-FDMA technology used in LTE uplink is not used in other standards. Therefore, the upstream signal characteristics need to be specially studied. Some common settings in LTE signal simulation include frequency, bandwidth, number of resource blocks contained in an LTE signal, antenna configuration, reference signal sequence configuration, downlink synchronization channel configuration, cyclic prefix length, allocation of user data and modulation schemes, and L1/L2 control Parameters such as the configuration of the channel.
The option R&S SMx-K55 is used for R&S company signal sources, such as R&S SMU200A, R&S SMJ100A and R&S SMATE200A, which can generate LTE FDD and LTE TDD uplink and downlink RF signals according to the TS36.211 standard for component performance testing And receiver testing of base stations and mobile terminals. The figure below shows the setup of the LTE TDD signal along with the graphical Display of the resource allocation diagram.
In addition, R&S also provides high-performance dual-channel baseband signal sources AMU200A and AFQ100A, and with the addition of AMU-K55 or AFQ-K255 options, it can simulate LTE baseband signals for early LTE R&D baseband signal simulation. And through an EX-IQ-BOX provided by R&S, users can generate digital baseband signal formats to suit their needs.
These meters and their options provide channel coding, MIMO precoding for up to four transmit antennas, and real-time fading simulation for 2×2 MIMO. This software option is directly installed on the instrument, providing users with a variety of configuration possibilities. Users can not only call the pre-defined test scenarios to quickly set up the test, but also flexibly set various parameters according to their own needs. Carry out customized tests: parameters such as reference symbols, control channels, synchronization channels and data channels, in addition, each subframe can be configured independently.
At present, R&S’s LTE signal simulation solution fully complies with the 3GPP V8.40 standard, including PRACH, sounding reference signal, PUCCH coding for uplink, PHICH and PCFICH coding for downlink, and E-Test model channel specified by the 36.141 standard.
LTE Signal Analysis
Secondly, in terms of radio frequency analysis of LTE signals, because LTE signals use a new access method OFDMA, the signal bandwidth can reach up to 20MHZ, which puts forward higher requirements for signal frequency domain analysis and modulation domain analysis. The R&S FSQ and R&S FSG signal analyzers can analyze 3GPP LTE base stations or transmitter modules of mobile phones. The signal analysis options R&S FSQ-K101 and R&S FSQ-K105 support the measurement of LTE FDD and TDD RF modulated signals and display the results in a graph or table: e.g. EVM, frequency error, spectral flatness, I/Q offset, eye diagram , constellation diagram and group delay measurement results. The options R&S FSQ-K100 and R&S FSQ-K104 can be used to analyze 3GPP LTE downstream signals. Similar to the upstream signal option, this option can analyze 3GPP LTE FDD in the frequency, time and modulation domains for all channel bandwidths specified by the standard. and TDD signals are measured.
To measure LTE baseband signals, whether balanced or unbalanced, the R&S FSQ’s analog (R&S FSQ-B71) and digital (R&S FSQ-B17) baseband input options can be used. At the same time, R&S also provides an EX-IQ-BOX that can adapt to the user’s own digital baseband format. By using it with the B17 interface on the FSQ, it can analyze the LTE digital baseband signal.
In addition, if you want to analyze OFDM signals, R&S has developed the FSQ-K96 option on the high-end signal analyzer FSQ, which can meet the needs of early LTE development and analysis of arbitrary OFDM signals.
LTE MIMO testing
R&S’s RF signal generator SMU200A, or baseband signal generator AMU200A, can use a single instrument for MIMO receiver testing. Both signal generators are equipped with two signal sources. After installing the R&S SMU-K74 or AMU-K74 option, the 4 fading channels required by the 2×2 MIMO system can be simulated in real time, so that the 2×2 MIMO system can be simulated in real time. The receiver is tested. Both meter solutions support various fading modes defined by the ITU for 3GPP LTE, including correlation characteristics between fading paths.
By connecting two or four R&S signal analyzers FSQ or FSG, R&S can provide 2×2 and 4×4 MIMO signal analysis, in this case, only one master FSQ/G needs to be configured with K100 (or K104) and K102 options, it can support three MIMO modes in LTE FDD and LTE TDD: transmit diversity, spatial multiplexing and cyclic delay diversity.
LTE protocol testing
The tests of the LTE protocol stack are used to verify some signaling functions such as call setup and release, call reconfiguration, state handling and mobility. Interoperability testing with 2G and 3G systems is another requirement for LTE. In addition, in order to ensure that the terminal’s protocol stack and applications can handle high data rate data, it is necessary to test and verify the terminal throughput requirements. In the early stage of LTE implementation, the R&D department needed multiple test scenarios including various parameter configurations to test the LTE protocol stack. In addition, the LTE physical layer has many important functions, including cell search, HARQ protocol, scheduling, link adaptation, uplink time control, and power control. And these processes have very strict timing requirements. Therefore, the physical layer also needs to be fully tested to ensure the performance of LTE.
Based on Rohde & Schwarz’s leading position in the field of UMTS LTE protocol stack testing, R&S has launched the LTE protocol tester CMW500. Its powerful hardware solution can provide frequencies up to 6GHz and bandwidth of 40MHz. It can not only be used for conformance testing, performance testing and interoperability testing, but also extends its benefits to the subsequent stages of the product life cycle, thereby giving chip and wireless device manufacturers the opportunity to develop UMTS LTE protocol conformance at all stages brings multiple benefits. Moreover, it also has an optional software solution for PC, which can support individual developers to carry out protocol development work at an early stage, thereby effectively reducing the cost of the entire research and development process of UMTS LTE wireless equipment. Therefore, the use of CMW500 enables the collaborative development, testing and optimization of software and hardware in parallel, thereby accelerating the time-to-market of products.
By configuring the CMW-KP500 MLAPI and CMW-KP501 LLAPI on the CMW500, R&S provides two different programming interfaces, the low-level and high-level required for protocol stack testing, so that developers can flexibly test the protocol stack at an early stage, and such The test is fully compatible with the later conformance test, which can save the time and cost of the later test.