Today’s spectrum is very crowded and complex. The demand for more data is driving research to potentially utilise centimetre-wave and millimetre-wave frequency bands for 5G high data throughput applications. These frequency bands offer the potential for contiguous blocks of additional spectrum, which is not available in today’s sub-6 GHz frequency band. How­ever, there is also interest in utilising today’s sub-6 GHz spectrum more efficiently for 5G applications. This is driving research for new 5G candidate waveforms to impro­ve out-of-band spectrum suppression, and to utilise spectrum more efficiently.

Operating in the already crowded sub-6 GHz spectrum poses questions regarding how new 5G candidate waveforms and legacy 3G and 4G waveforms will interact and coexist with one another. While some scenarios can be explored with field testing, it can be beneficial for researchers and engineers to explore different potential coexistence scenarios in a research and development (R&D) lab environment before field testing with actual hardware. Flexibility is key for exploring many scenarios in the R&D lab environment with different waveform types, frequencies and parameters before field testing, in which the test scenarios may be more limited.

This article talks about a flexible R&D coexistence test-bed, which combines de­sign simulation with test equipment to evaluate potential 5G coexistence scenarios. The scenario under consideration shows a sub-6 GHz coexistence case study with a notched filter bank multicarrier (FBMC) 5G candidate waveform and a 4G long-term evolution (LTE) waveform centred in the FBMC notch. The FBMC notch width is then varied to evaluate LTE error vector magnitude (EVM) as a function of decreasing notch width.

5G candidate waveforms

There are many waveforms being resear­ched for 5G applications, some of which are discussed briefly below.

The first one is orthogonal frequency division multiplexing (OFDM), which is curr­­­­­ently used in 4G and, for that reason and more, it is under consideration for 5G through the use of filtered OFDM (F-OFDM).

The second one is FBMC which applies filtering on a per-subcarrier basis to provide improved out-of-band spectrum characteristics. It uses a flexible approach to baseband filtering, using either a polyphase network or an extended inverse fast Fourier transform (IFFT) filter.

The third one is universal filtered MC (UFMC), which applies filtering on a per-sub-band basis, instead of a per-subcarrier basis as with FBMC. A potential benefit of UFMC is reduced complexity of the baseband algorithms.

 

Fig. 1 shows how the steeper spectrum roll-out of some of the 5G candidate waveforms is related to spectrum efficiency. This simulation case study in Figure 1 shows that the proposed new waveform tech­nologies FBMC, UF-OFDM (UFMC) and others potentially have improved out-of-band performance relative to OFDM, and thus may offer the potential for a smaller frequency guard band to achieve higher spectrum efficiency.

The following section evaluates the coexistence case study with an FBMC 5G candidate waveform combined with a 4G LTE waveform, to see how out-of-band spectrum emissions can impact the coexistence and guard band requirements.

5G candidate FBMC waveform with 4G LTE waveform

For this case study, a sub-6 GHz coexistence test-bed was used. The LTE and FBMC signals are generated using the design software schematic shown in Fig. 2. The design software is a system-level simulator that can be used to construct system-level designs, evaluate design performance and perform design trade-offs.

The FBMC signal source and the LTE signal source are shown on the left side of the schematic. They generate complex waveforms, which are then modulated on to a carrier using IQ modulators. The FBMC and LTE waveforms are fed into a Signal Combiner element, which resamples the two waveforms and combines th­em into a single composite waveform that is downloaded to an arbitrary waveform generator (AWG). The output of the AWG is analysed using a radio frequency (RF) signal analyser and vector signal analysis (VSA) software.

One of the challenges with combining multiple types of waveforms is that they can have different centre frequencies, bandwidths and sample rates. This can make it challenging to combine many different and disparate waveforms.

Fortunately, recent breakthroughs in simulation arbitrary resampling technology have enabled multiple input signals, with different centre frequencies, bandwidths and sample rates, to be combined. This is made possible with the SystemVue Signal Combiner element, which effectively resamples multiple input signals and combines them to create one composite output waveform with a user-defined sample rate, which can be set for the AWG. This enables signal scenario flexibility to create complex waveforms in simulation, which can then be downloaded to the AWG to create coexistence test signals for the R&D lab environment.

To evaluate the coexistence of the LTE signal in the presence of the FBMC signal, the test scenario is modified in simulation to notch out some of the active subcarriers in the FBMC signal. The LTE centre frequency is then set so that it is centred in the notch.

The upper-right display in Fig. 3 shows the measurement result using the VSA software on the RF signal analyser. The LTE EVM result is approximately 0.6 per cent, which indicates that the FBMC out-of-band characteristics are having a minimal impact on the LTE signal for this particular notch configuration.

The FBMC notch width was then va­ried to evaluate LTE EVM as a function of decreasing notch width. This type of scenario could be useful in determining how much guard band is required for coexistence between the two waveforms to maximise spectral efficiency. EVM is used in this case study as the metric for waveform quality and as a measure of the interference impact. Bit error rate or throughput could be another metric of interest for a receiver.