Dealing With Fractional Sampling Time Offsets for Unsynchronized Mobile Channel Measurements

Physical Layer (PHY) investigations, e.g., studying the reliability of communications systems, heavily rely on the considered radio channels. Therefore, using accurate channel models or measured channel data is the key to meaningful results. A measurement setup close to the considered application is hence important. In case of scenarios with a moving transmitter or receiver, wired synchronization of their Local Oscillators (LOs) becomes infeasible. This leads to differing sampling frequencies which translate into Fractional Sampling Time Offsets (FSTOs). If a tap in the channel impulse response (CIR) does not coincide with a sampling time, a peak in the CIR turns into a plateau. Thereby, measured CIRs do not reflect the expected amplitudes of the channel components. By post-processing measured CIRs using interpolation, information about the amplitudes is not lost and can be reconstructed efficiently. Thus, performing mobile channel measurements without synchronization becomes feasible. Since propagation delays are equally causing FSTOs, measurement setups with synchronization also benefit from the proposed procedure. This is shown by applying our solution to real-world data.