The proposed scheme relies on an optimal pilot design for time-domain channel estimation, which can either completely eliminate pilot contamination (PC) or significantly reduce it, depending on the channel's coherence time. This is achieved by designing an optimal pilot set that allows beneficial grouping of users in all cells and assigning a time-shifted pilot transmission to different groups. Unlike existing PC elimination schemes, which require an excessively long channel coherence time, the proposed scheme can completely eliminate PC under a much shorter coherence time. Moreover, existing schemes can no longer be used if the channel coherence time is insufficiently large. By contrast, the proposed scheme can still be implemented to significantly reduce PC even for extremely short channel coherence time, which is particularly beneficial for high-velocity scenarios. The simulation results demonstrate the efficiency of the proposed scheme. The scheme is designed for large-scale multiple-input multiple-output (MIMO) aided orthogonal frequency division multiplexing (OFDM) systems in multi-cell environments. The proposed scheme is an improvement over existing schemes that require a central controller to manage the staggering of pilot-aided training in all cells, which is a challenging large-scale pilot assignment task. The proposed scheme does not require knowledge of the second-order channel statistics or the users' angles of arrival (AOAs), making it more efficient and appealing. The scheme can be implemented in time-division duplexing (TDD) systems, where the downlink channel state information (CSI) can be estimated at the base station (BS) with the aid of uplink training by exploiting the channel's reciprocity. The signaling overhead imposed by the acquisition of uplink CSI scales linearly with the number of mobile stations (MSs), which is typically much lower than the number of BS antennas. The proposed scheme is designed to address the problem of pilot contamination in large-scale MIMO systems, which is a critical issue that limits the system's achievable performance. The scheme is an improvement over existing schemes that require a much longer training duration or a much longer channel coherence time. The proposed scheme can be implemented in high-velocity scenarios, where the channel coherence time is extremely short. The simulation results demonstrate the efficiency of the proposed scheme in reducing pilot contamination. The scheme is designed for multi-cell MIMO systems, where the number of orthogonal pilot sequences available is limited due to the limited channel coherence time. The worst-case scenario is associated with the pilot reuse factor of one, when all cells use the same set of orthogonal pilot sequences at the same time, causing severe pilot contamination. The proposed scheme is designed to address this issue by designing an optimal pilot set that allows beneficial grouping of users in all cells and assigning a time-shifted pilot transmission to different groups. The scheme can be implemented in time-division duplexing (TDD) systems, where the downlink channel state information (CSI) can be estimated at the base station (BS) with the aid of uplink training by exploiting the channel's reciprocity. The signaling overhead imposed by the acquisition of uplink CSI scales linearly with the number of mobile stations (MSs), which is typically much lower than the number of BS antennas. The proposed scheme is an improvement over existing schemes that require a central controller to manage the staggering of pilot-aided training in all cells, which is a challenging large-scale pilot assignment task. The proposed scheme does not require knowledge of the second-order channel statistics or the users' angles of arrival (AOAs), making it more efficient and appealing. The scheme can be implemented in high-velocity scenarios, where the channel coherence time is extremely short. The simulation results demonstrate the efficiency of the proposed scheme in reducing pilot contamination. The scheme is designed for large-scale MIMO systems, where the number of orthogonal pilot sequences available is limited due to the limited channel coherence time. The worst-case scenario is associated with the pilot reuse factor of one, when all cells use the same set of orthogonal pilot sequences at the same time, causing severe pilot contamination. The proposed scheme is designed to address this issue by designing an optimal pilot set that allows beneficial grouping of users in all cells and assigning a time-shifted pilot transmission to different groups. The scheme can be implemented in time-division duplexing (TDD) systems, where the downlink channel state information (CSI) can be estimated at the base station (BS) with the aid of uplink training by exploiting the channel's reciprocity. The signaling overhead imposed by the acquisition of uplink CSI scales linearly with the number of mobile stations (MSs), which is typically much lower than the number of BS antennas. The proposed scheme is an improvement over existing schemes that require a central controller to manage the staggering of pilot-aided training in all cells, which is a challenging large-scale pilot assignment task. The proposed scheme does not require knowledge of the second-order channel statistics or the users' angles of arrival (AOAs), making it more efficient and appealing. The scheme can be implemented in high-velocity scenarios, where the channel coherence time is extremely short. The simulation results demonstrate the efficiency of the proposed scheme in reducing pilot contamination.
