@phdthesis{,
    author = {Hälsig, Tim},
    title = {Synchronization and Equalization for Multi-Transceiver Systems with a Focus on Line-of-Sight},
    editor = {},
    booktitle = {},
    series = {},
    journal = {},
    address = {},
    publisher = {},
    edition = {},
    year = {2020},
    isbn = {},
    volume = {},
    number =  {},
    pages = {},
    url = {},
    doi = {},
    keywords = {synchronization, equalization, MIMO, LoS, spatial multiplexing, frequency offsets, symbol timing, parameter estimation, adaptive filters, mmWave, 60 GHz, multi-gigabit per second, measurement},
    abstract = {This work considers the problem of synchronization and equalization for frequency-selective multiple-input and multiple-output (MIMO) systems. The focus is on line-of-sight MIMO transmission at millimeter-wave frequencies for backhaul-like scenarios, but the approach is generally applicable to MIMO systems using spatial multiplexing. It is shown that two timing impairments are typically present in a wireless transmission system, namely carrier frequency offset and sampling frequency offset. They exist due the desire for low-complexity hardware, and in particular oscillator, implementations. Both of them can be characterized with a phase offset process with respect to their ideal nominal value. These impairments cause the observed channel characteristic, between transmitter and receiver baseband, to be time varying in general. For MIMO systems, where multiple transceivers are used, it is discussed that a different number of these phase processes will be observed, depending on the hardware setup. For example, when the transceivers are widely separated on either side of the link, sharing a time reference between them may be infeasible, meaning that as many independent phase processes as transceivers exist in the system. Two training signal based channel estimation techniques are proposed, in order to identify the phase processes, and the time-varying channel characteristics in general. One of them is a correlation method, which is a standard technique known from the literature, but will in this work be firstly investigated for the case of multiple timing impairments, especially multiple sampling frequency offset processes, in a MIMO system. As an alternative, adaptive filtering is proposed for direct identification and tracking of the time-varying system. Parameter selection recommendations for both approaches are given. In particular, some new results for the step size selection of the adaptive filter for MIMO systems with multiple timing impairments are derived. Standard and adaptive equalization are discussed for dealing with the channel impairments. It is seen that standard equalization greatly simplifies, depending on the oscillator setup. However, for the most general case, the equalizer needs to be fully fractionally spaced. Simulation and measurement results corroborate that the proposed estimation and equalization strategies are viable, showing that multi-gigabit per second wireless transmission is feasible with a spatial-multiplexing line-of-sight MIMO system, even when independent oscillators are used for each transceiver chain.},
    note = {},
    
    school = {Universität der Bundeswehr München},
    
}