GNU Radio GFDM Modulator/Demodulator

The gr-gfdm project is a Free Software Package which aims to provide an implementation of Generalized Frequency Division Multiplexing (GFDM) in the GNU Radio framework. GFDM is a proposed waveform for use in 5G.

This project was initiated as a Bachelor thesis at the Communication Engineering Lab (CEL) at the Karlsruhe Institute of Technology (KIT), Germany, http://www.cel.kit.edu.

Concept

Tackling current issues with OFDM several new waveforms are proposed for consideration in 5G. FBMC, UFMC, BFDM and GFDM are the names of the proposed waveforms and they are filtered multicarrier systems. Orthogonality of neighboring carriers is no constraint.

GFDM was proposed by the Vodafone Chair of TU Dresden and accurately described in [1]. Due to the high complexity of a Transmit-Matrix-based approach a low coplexity receiver [2] and transmitter [3] are proposed.

GFDM can be described by parallelizing several SC-FDE streams on subcarrier. The transmit symbols are localized in a time/frequency-grid and pulshaping is applied subcarrier-wise. After pulshaping and localizing on the correct subcarrier-frequency the symbolstreams are superpositioned and can be transmitted. On receiver side the symbols on the subcarrier can be extracted by applying a MF, ZF or MMSE-filter of the previous pulseshaping filter. Non-orthogonality of neighboring subcarrier introduces ICI if demodulating with MF. A successive interference cancellation algorithm is proposed to remove interference.

Due to its block-nature a block synchronisation with improved Schmidl & Cox - Symbols can be achievd.

Capabilities

For generating a transmit symbol stream gr-gfdm provides following blocks:

• resource mapper
• modulator
• cyclic prefixer

For reception of a GFDM symbol stream gr-gfdm provides following blocks:

• synchronization
• channel estimator
• prefix remover
• simple demodulator
• advanced demodulator with IC
• demapper

Signal processing in the blocks is performed in kernels which have no dependency on GNU Radio. Thus inclusion in interfaces to other programming languages is easier.

The modulator and demodulator are implemented using the low complexity approach with Sparse Frequency Domain processing and heavy use of FFTW and VOLK to accelerate signal processing on modern GPPs.

Tests of optimized C++ code are performed with a reference implementation in Python. Additionally the C++ code is wrapped with cython.

Requirements

• GNU Radio OOT

  • GNU Radio 3.7.9 (verified)
  • GR 3.7 API
  • GR-FFT
  • GR-FILTER
  • VOLK
  • UHD (examples)

• PyGFDM

  • Numpy
  • Scipy
  • Commpy
  • Matplotlib

Build/Install instructions

1. Install/Build GNU Radio with at least support for FFT,FILTER and VOLK

2. Get gr-gfdm from github - git clone https://github.com/kit-cel/gr-gfdm.git

3. Configure gr-gfdm - mkdir build && cd build && cmake ../

4. Build and install gr-gfdm - make && sudo make install (default install target: /usr/local)

5. Configure custom blocks path in GNU Radio Companion to use /usr/local/share/gnuradio/grc/blocks

Troubleshooting/Bugs

In case you encounter bugs not related to GNU Radio core functions in gr-gfdm open an issue at https://github.com/kit-cel/gr-gfdm/issues. Otherwise consider reporting the issue to the GNU Radio project https://www.gnuradio.org.

References

1. N. Michailow et al. “Generalized Frequency Division Multiplexing for 5th Generation Cellular Networks”. In: Communications, IEEE Transactions on 62.9 (2014), S. 3045–3061. doi: 10.1109/TCOMM.2014.2345566.

2. I.S. Gaspar et al. “Low Complexity GFDM Receiver Based on Sparse Frequency Domain Processing”. In: Vehicular Technology Conference (VTC Spring), 2013 IEEE 77th. IEEE, 2013, S. 1–6. doi: 10.1109/VTCSpring.2013.6692619.

3. N. Michailow et al. “Generalized frequency division multiplexing: Analysis of an alternative multi-carrier technique for next generation cellular systems”. In: Wireless Communication Systems (ISWCS), 2012 International Symposium on. IEEE, 2012, S. 171–175. doi: 10.1109/ISWCS.2012.6328352.

4. J. Demel, C. Bockelmann, A. Dekorsy "Evaluation of a Software Defined GFDM Implementation for Industry 4.0 Applications". In: IEEE International Conference on Industrial Technology (ICIT 2017), Toronto, Canada, 22. - 25. March 2017

5. J. Demel, C. Bockelmann, A. Dekorsy, Andrej Rode, Sebastian Koslowski, Friedrich K. Jondral "An optimized GFDM software implementation for future Cloud-RAN and field tests". In: GNU Radio Conference 2017, San Diego, USA, 11. - 15. September 2017