The LTE physical layer comprises two types of signals known as physical signals and physical channels. Physical signals are generated in Layer 1 and used for system synchronization, cell identification, and radio channel estimation. Physical channels carry data from higher layers including control, scheduling, and user payload. The uplink physical signals and channels.

Uplink Frame Structure There are two uplink frame structures, one for FDD operation called type 1 and the other for FDD operation called type 2. Frame structure type 1 is 10 ms long and consists of ten subframes, each comprising two 0.5 ms slots. Figure 1 shows how the DMRS and PUSCH map onto the frame structure.

The number of symbols in a slot depends on the CP length. For a normal CP, there are seven SC-FDMA symbols per slot. For an extended CP used for when the delay spread is large, there are six SC-FDMA symbols per slot. Demodulation reference signals are transmitted in the fourth symbol (that is, symbol number 3) of every slot. The PUSCH can be transmitted in any other symbol. Figure 2 shows the uplink frame structure type 1 in both frequency and time.













Figure 1. Frame Structure 1 for uplink showing mapping for DMRS and PUSCH

















Figure 2. Frame Structure 1 for the uplink showing one subframe vs.frequency.





Each vertical bar represents one subcarrier. Transmissions are allocated in units called resource blocks (RB) comprising 12 adjacent subcarriers for a period of 0.5 ms. In addition to the DMRS and PUSCH the figure also shows the PUCCH which is always allocated to the edge RB of the channel bandwidth alternating from low to high frequency on adjacent slots.

Note that the frequency allocation for one UE is typically less than the system bandwidth. This is because the number of RB allocated directly scales to the transmitted data rate which may not always be the maximum. The DMRS is only transmitted within the PUSCH and PUCCH frequency allocation-unlike the reference signals on the downlink which are always transmitted across the entire channel bandwidth even if the channel is not fully occupied.

If the base station needs to estimate the uplink channel conditions when no control or payload data is scheduled then it will allocate the SRS which is independent of the PUSCH and PUCCH. The PUSCH can be modulated at QPSK, 16QAM or 64QAM. The PUCCH is only QPSK and the DMRS is BPSK with a 45 degree rotation.

Analyzing an SC-FDMA Signal
Figure 3 shows some of the measurements that can be made on a typical SC-FDMA signal using the Agilent 89601A Vector Signal Analyzer software. The IQ constellation in trace A (top left) shows that this is a 16QAM signal. The unity circle represents the DMRS occurring every seventh symbol, which are phase-modulated using an orthogonal Zadoff-Chu sequence.

Trace B (lower left) shows signal power versus frequency. The frequency scale is in 15 kHz sub-carriers numbered from -600 to 599, which represents a bandwidth of 18 MHz or 100 RB. The nominal channel bandwidth is therefore 20 MHz and the allocated signal bandwidth is 5 MHz towards the lower end. The brown dots represent the instantaneous subcarrier amplitude and the white dots the average over 10 ms.

In the center of the trace, the spike represents the local oscillator (LO) leakage - IQ offset - of the signal; the large image to the right is an OFDM artifact deliberately created using 0.5 dB IQ gain imbalance in the signal. Both the LO leakage and the power in non-allocated sub-carriers will be limited by the 3GPP specifications. Trace C (top middle) shows a summary of the measured impairments including the error vector magnitude (EVM), frequency error, and IQ offset.

Note the data EVM at 1.15 percent is much higher than the DMRS EVM at 0.114 percent. This is due to a +0.1 dB boost in the data power as reported in trace E, which for this example was ignored by the receiver to create data-specific EVM. Also note the DMRS po.



wer boost is reported as +1 dB, which can also be observed in the IQ constellation because the unity circle does not pass through eight of the 16QAM points. Trace D (lower middle) shows the distribution of EVM by subcarrier. The average and peak of the allocated signal EVM is in line with the numbers in trace C. The EVM for the non-allocated subcarriers reads much higher, although this impairment will be specified with a new “in-band emission” requirement as a power ratio between the allocated RB and unallocated RB.

The ratio for this particular signal is around 30 dB as trace B shows. The blue dots in trace D also show the EVM of the DMRS, which is very low.Trace E (top right) shows a measurement of EVM by modulation type from one capture. This signal uses only the DMRS phase modulation and 16QAM so the QPSK and 64QAM results are blank.

Finally, trace F (lower right) shows the PAR — the whole point of SC-FDMA — in the form of a complementary cumulative distribution function (CCDF) measurement. It is not possible to come up with a single figure of merit for the PAR advantage of SC-FDMA over OFDMA because it depends on the data rate. The PAR of OFDMA is always higher than SC-FDMA even for narrow frequency allocations;

however, when data rates rise and the frequency allocation gets wider, the SC-FDMA PAR remains constant but OFDMA gets worse and approaches Gaussian noise. A 5 MHz OFDMA 16QAM signal would look very much like Gaussian noise.

From the white trace it can be seen at 0.01 percent probability the SC-FDMA signal is 3 dB better than the blue Gaussian reference trace. As every amplifier designer knows, shaving even a tenth of a decibel shaved from the peak power budget is a significant improvement.





Included in this comprehensive suite of LTE tools are solutions to design and simulate LTE signals, create and measure LTE encoded signals with sources and analyzers, and test mixed analog & digital signals – see figure 4. Just added to Agilent’s suite of LTE solutions is a one-box tester that provides the platform for protocol design and test solutions, in partnership with Anite.

This platform will provide RF and protocol conformance test systems when they are needed. And, the newly introduced signaling analyzer enables analysis of the new LTE/SAE network.