5G air interfaces need channel measurements

-October 31, 2015

5G wireless communications should bring increased network capacity, higher peak data rates, and more reliable service in mobile communications systems. Many of the goals are 10x, 100x or 1000x today's performance but aren't achievable in the currently available spectrum below 6 GHz. Therefore, new air interfaces are being investigated in Centimeter (cm) and Millimeter-wave (mmWave) frequencies up to 100 GHz. Characterization of the radio channel at mmWave frequencies presents many new challenges for engineers. Here are some of these challenges and some considerations.

To define new air interface standards, researchers will need to characterize the radio channel so they can understand how the signal will propagate. Researchers are using channel sounding techniques to collect the CIR (channel impulse response) data so they can extract channel parameters by using channel parameter estimation algorithms. The extracted data are then used for developing new channel models as shown in Figure 1. Sounds straightforward and easy, right?

Well, not quite.

Figure 1. The model of a wireless transmission channel consists of channel sounding, estimating of channel parameters, and statistics.

Channel sounding measurement systems can range from simple to complex depending on parameters being estimated. When measuring time-varying channels with multipath propagation, you need to understand the complex impulse response with time and phase information. In addition, one of the key challenges is being able to duplicate or validate the measurements with different measurement systems in similar conditions.

Key technical challenges include:

  • Signal generation and analysis at mmWave frequencies with greater than 500 MHz bandwidth and with multi-channel support
  • Data capture and storage
  • Channel parameter estimations
  • Calibration and synchronization

Now let's discuss considerations to help you address these challenges.

Signal generation and analysis
To meet the high demands for 5G, the air interface standards will likely include mmWave frequencies up to 100 GHz, with 500 MHz to 2 GHz bandwidth, and with multichannel support. That's a lot to consider. The requirements will place great demand on the channel sounding measurement system. The measurement system needs to support these core requirements and provide repeatable measurements. Key components for this measurement system will be a wideband DAC (Digital to Analog Converter) in the form of a baseband AWG (arbitrary waveform generator) and an ADC (analog-to-digital converter) that will take the form of a wideband digitizer or oscilloscope to support the needed bandwidth and provide enough resolution to support the dynamic range needed to capture the signal. Also, because 5G is not yet defined, the test equipment should be flexible so that it can be configured and reconfigured as the test requirements and standards evolve.

Data capture and storage
When you consider the raw data that needs to be collected with a wideband measurement system that also has multi-channel capability, an eight-channel, 1 GHz bandwidth measurement can consume gigabytes of data in just one second, quickly filling disk drives. In addition, consider how to get this data from the ADC to a storage device. It's nearly impossible for the data to be captured and streamed in real-time. Disk drive manufacturers might like this because they will sell more storage, but it's just not practical. Instead, there are two other data capture-methods to consider that can reduce the amount of collected data:

  • If the sounding signal is less than one transmitting period, you can capture only the effective data or only the data needed to perform the CIR calculations. This method can greatly reduce the data collection.
  • Taking this one step further, you can perform the wideband measurement with real-time autocorrelation and signal processing with an onboard FPGA to produce the effective CIR data within the measurement system, now only the CIR results need to be saved. Thus, significantly saving storage space and providing the CIR results much faster.

Channel parameter estimations
Much of the research to date has focused on a single channel. MIMO (multiple input, multiple output) channels, however, introduce spatial and correlation information. The key issue with MIMO channels is how to estimate the spatial parameters. This includes parameters such as AoA (angle of arrive), AoD (angle of departure), and AS (angular spread). There are several channel parameter estimation algorithms that can be considered including beamforming based, subspace based, and ML (Maximum Likelihood) based. For consistency, coherence, and estimation performance, the ML-based estimation algorithms provide the very good performance for MIMO channel parameter estimations. Specifically, the SAGE (space-alternating generalized expectation-maximization) algorithm (ML based) with relativity low computing, is widely accepted by the research community.

Calibration and synchronization
Calibration and synchronization are paramount to getting accurate, repeatable results. Synchronization of the transmitter and receiver subsystems can be achieved using two Rubidium clocks to provide a stable, high precision synchronized 10MHz reference clock to the transmitter and receiver as shown in Figure 2. In addition, triggering must be used to synchronize the sounding stimulus signal generation and acquisition. With-in a mmWave measurement system shown in Fig. 2, the following calibrations need to be considered:

  • System calibration, also called "back-to-back" calibration, involves physically connecting the transmitter to the receiver to align the frequency reference and system clocks. This can then be used to get accurate amplitude, phase and time of arrival estimates.
  • The differential IQ outputs of a baseband AWG can have timing, gain and quadrature errors, which can impact signal quality. An IQ mismatch calibration is used to address any imbalance of the in-phase and quadrature-phase signal that is output from the AWG.
  • The multichannel, wideband digitizer or oscilloscope can have inter-channel time and phase variants that will impact the measurement results. Various methods can be used to measure the cross-channel skew. One method is to measure the magnitude and phase differences across a large frequency range for each channel and apply a wideband correction filter.
  • Antenna and power calibration also need to be considered. Check with the antenna manufacturer for calibration data. If not provided, antenna array phase pattern measurement could be performed in a microwave chamber and compared with the theoretical performance of the antenna array.


Figure 2. Measurement System includes Rubidium clocks for precise Tx and Rx synchronization and acquisition trigger to align signal generation and data capture.

Conclusion
There are many challenges in the characterization of new 5G mmWave air interfaces. When considering time-varying channels with multipath propagation, the measurement systems can be complex. The measurement system should support mmWave, wideband signals and multiple channels, calibration, and synchronization. Such support can produce accurate and repeatable measurements with good channel parameter estimation algorithms for the characterization and development of realistic and accurate channel models.

Also see:

Loading comments...

Write a Comment

To comment please Log In

FEATURED RESOURCES