OFDM Precoding for Filter-Bank based Waveforms

Abstract : New use cases for wireless communications recently emerged ranging from massive sensor networks to connected cars. These applications highly differ from typical signals supported by already deployed mobile technologies, which are mainly high data rate pipes. The forthcoming generation of mobile technology, 5G New Radio, introduces the concept of signal numerology so as to properly serve the requirements of the diverse applications it will support. Indeed by considering different numerologies, the time/frequency signal allocation is made more flexible which allows to shape the transmitted signal according to its needs. However, multiplexing signals with different numerologies generates interference and therefore signal distortion. Spatial filtering, such as beamforming, is envisioned for 5G above 6-GHz communications to limit inter-user interference. However, this issue still holds for sub-6 GHz systems where spatial filtering is not considered in 5G.In this work, we consider side lobe rejection techniques to ease service multiplexing in sub-6 GHz bands. Not only it provides inter-user interference mitigation but it also improves the bandwidth use efficiency in bands where frequency is a scarce resource. A novel solution, mixing filter-bank for confined spectrum and complex orthogonality for a straightforward re-use of known-how 4G/5G techniques, is proposed. The complex orthogonality is restored thanks to an OFDM precoding substituting the commonly used Offset-QAM signaling which limits the orthogonality to the real field. Moreover, the proposed solution, named Block-Filtered Orthogonal Frequency Division Multiplexing (BFOFDM), relies on a simple 5G receiver scheme which makes it backward compatible with already deployed technologies.The BF-OFDM system model is fully described and adapted to cellular standards. Besides, different prototype filter designs methods are proposed to either improve the intrinsic interference attenuation or to better confined the spectrum of the transmitted signal. Last but not least, the proposed waveform will be compared with state-of-the-art solutions for both typical and 5G oriented evaluation scenarios such as multi-service coexistence.
Complete list of metadatas

Cited literature [117 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02275297
Contributor : Abes Star <>
Submitted on : Friday, August 30, 2019 - 3:53:07 PM
Last modification on : Sunday, September 1, 2019 - 1:17:20 AM

File

TheseDEMMER.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02275297, version 1

Collections

Citation

David Demmer. OFDM Precoding for Filter-Bank based Waveforms. Signal and Image processing. Conservatoire national des arts et metiers - CNAM, 2019. English. ⟨NNT : 2019CNAM1237⟩. ⟨tel-02275297⟩

Share

Metrics

Record views

91

Files downloads

95