Estimation of LTE Signals' Time of Arrival in a Multi-Band Environment

This thesis summarizes the research work on the estimation of the Time of Arrival (ToA) of signals in Orthogonal Frequency Division Multiplexing (OFDM) based communication systems. The estimated ToA values can be employed for positioning purposes, thus providing an alternative mean of localization to satellite-based technologies. By employing signals of oppor- tunity (SoP), this information can be obtained without the need to allocate transmission resources specifically for positioning purposes. The OFDM multiplexing technique is widely employed in modern telecom- munication standards and presents interesting properties with regard to ToA estimation. In particular, the Third Generation Partnership Project (3GPP) Long Term Evolution is interesting for its diffusion, geographical coverage, and wide transmission bandwidth, making it a prime candidate for research. The upcoming fifth generation (5G) mobile systems are also expected to employ an OFDM-based physical layer, leading to further possibilities of ap- plication and development. In the first part of the work, novel algorithms for the estimation of ToA in OFDM-based systems have been developed. The slope-based algorithm exploits the phase rotation of sub-carriers to obtain the estimation with sim- ple unwrap and linear fitting operations. It can be shown that if the Direct Path (DP) is also the stronger one, the ToA can be inferred from slope of the phase rotation even when secondary paths are presents. A piecewise variant with outliers removal is also introduced to reduce the effects of noise and phase jumps on the final estimation. The Difference-Based algorithm (DBTE) instead relies on a rough first estimate of the channel parameters to obtain an estimation of the first prop- agation path delay even when secondary paths present, on average, larger amplitudes than the direct one. Multiple consecutive symbols can be aggre- gated to achieve a more accurate and reliable estimate. The DBTE method achieves good performance at the cost of a greater computational complexity than the slope-based method. The second part of the work focused on exploiting multiple transmission bands at the same time for ToA estimation. The developed research is based on the assumption that the propagation environment is highly correlated be- tween the transmission bands, which is reasonable if the carrier frequencies are relatively close to each other. This allows one to make full use of the higher bandwidth occupied by the set of signals as a whole, rather than just the bandwidth of each signal on its own, to achieve improved precision and multi-path robustness. The Space-Alternating Generalized Expectation-Maximization (SAGE) algorithm has been chosen for its versatility and good performance in com- plex propagation environments. SAGE presents the advantage of being appli- cable to multi-band scenarios without the need of significant modifications to its basic formulation. A simulator has been implemented in Matlab to evaluate the possible benefits of dual-band usage. In the LTE network multiple transmitter can be allocated on the same physical base station mast, in order to decrease deployment costs and im- prove the network coverage and quality of service. To this end, a set of live measurements on downlink LTE signals has been performed in Monfalcone, Italy. The considered cellular mast carries 3 cell IDs for each operator and transmits on LTE band 20. The Cell-Specific Reference Signal (CRS) has been used as the reference signal of choice. The CRS is always transmitted, allowing it to be used in a fully opportunistic way. The ToA is derived from the gathered data, showing (in agreement with simulations) that the combi- nation of signals from multiple bands leads to a reduced range and standard deviation in the estimations.

This thesis summarizes the research work on the estimation of the Time of Arrival (ToA) of signals in Orthogonal Frequency Division Multiplexing (OFDM) based communication systems. The estimated ToA values can be employed for positioning purposes, thus providing an alternative mean of localization to satellite-based technologies. By employing signals of oppor- tunity (SoP), this information can be obtained without the need to allocate transmission resources specifically for positioning purposes. The OFDM multiplexing technique is widely employed in modern telecom- munication standards and presents interesting properties with regard to ToA estimation. In particular, the Third Generation Partnership Project (3GPP) Long Term Evolution is interesting for its diffusion, geographical coverage, and wide transmission bandwidth, making it a prime candidate for research. The upcoming fifth generation (5G) mobile systems are also expected to employ an OFDM-based physical layer, leading to further possibilities of ap- plication and development. In the first part of the work, novel algorithms for the estimation of ToA in OFDM-based systems have been developed. The slope-based algorithm exploits the phase rotation of sub-carriers to obtain the estimation with sim- ple unwrap and linear fitting operations. It can be shown that if the Direct Path (DP) is also the stronger one, the ToA can be inferred from slope of the phase rotation even when secondary paths are presents. A piecewise variant with outliers removal is also introduced to reduce the effects of noise and phase jumps on the final estimation. The Difference-Based algorithm (DBTE) instead relies on a rough first estimate of the channel parameters to obtain an estimation of the first prop- agation path delay even when secondary paths present, on average, larger amplitudes than the direct one. Multiple consecutive symbols can be aggre- gated to achieve a more accurate and reliable estimate. The DBTE method achieves good performance at the cost of a greater computational complexity than the slope-based method. The second part of the work focused on exploiting multiple transmission bands at the same time for ToA estimation. The developed research is based on the assumption that the propagation environment is highly correlated be- tween the transmission bands, which is reasonable if the carrier frequencies are relatively close to each other. This allows one to make full use of the higher bandwidth occupied by the set of signals as a whole, rather than just the bandwidth of each signal on its own, to achieve improved precision and multi-path robustness. The Space-Alternating Generalized Expectation-Maximization (SAGE) algorithm has been chosen for its versatility and good performance in com- plex propagation environments. SAGE presents the advantage of being appli- cable to multi-band scenarios without the need of significant modifications to its basic formulation. A simulator has been implemented in Matlab to evaluate the possible benefits of dual-band usage. In the LTE network multiple transmitter can be allocated on the same physical base station mast, in order to decrease deployment costs and im- prove the network coverage and quality of service. To this end, a set of live measurements on downlink LTE signals has been performed in Monfalcone, Italy. The considered cellular mast carries 3 cell IDs for each operator and transmits on LTE band 20. The Cell-Specific Reference Signal (CRS) has been used as the reference signal of choice. The CRS is always transmitted, allowing it to be used in a fully opportunistic way. The ToA is derived from the gathered data, showing (in agreement with simulations) that the combi- nation of signals from multiple bands leads to a reduced range and standard deviation in the estimations.