In various embodiments, techniques are provided to determine channel characteristics of various communication systems such as OFDM systems or systems using a plurality of transmit antennas by using various sets of training symbols that produce zero cross-correlation energy. Channel communication can accordingly be simplified as the zero cross-correlation property allows for channel estimation without a matrix inversion.

The present invention relates to a method and apparatus for channel estimation between a transmitter and a receiver in a wireless communications system. In one arrangement, the method comprises: receiving at the receiver a first sequence of bits representing a first sequence of coded symbols transmitted over the communications channel; decoding the first sequence of coded symbols using maximum-likelihood based decoding including: generating traceback outcomes by tracing backwards the first sequence of bits through a maximum-likelihood based traceback path, the traceback outcomes including a first portion associated with a first traceback depth and a second portion associated with a second traceback depth that is deeper than the first traceback depth; generating a channel estimate of the communications channel based on the first portion of the traceback outcomes; and generating an estimate of at least some information bits coded in the first sequence of coded symbols based on the second portion of the traceback outcomes.

A communication method includes generating a first sequence, where the first sequence is based on a complete complementary code set, the complete complementary code set is based on a Golay pair and a Hadamard matrix through a Kronecker product operation, and the first sequence is used for channel estimation; and sending the first sequence. In the method, use of a P-matrix is avoided in a process of constructing a plurality of streams of zero correlation sequences.

The present disclosure provides signal detection methods and apparatuses. A method executed by a network side device includes: obtaining a channel matrix by performing a channel estimation on a signal received from a terminal device; performing a data preprocessing on the channel matrix to obtain a first matrix and a first vector; and inputting the first matrix and the first vector into a trained deep learning model to obtain an estimated detection signal, wherein the trained deep learning model includes N trained convolutional network models, and N is a positive integer equal to a number of transmitting antennas.

A method in a receiver includes receiving a signal that traversed a multipath channel having a channel response, so as to produce a received sequence of samples that are indicative of the channel response. Using a predefined similarity measure, multiple metrics between the received sequence of samples and multiple respective candidate sequences of samples are calculated by the receiver. Each candidate sequence includes a combination of one or more dominant signal components having respective sample delays, and each candidate sequence corresponds to a channel multipath candidate that has one or more channel paths at the respective sample delays. The channel response is estimated by the receiver, by selecting the candidate sequence that best matches the received sequence, the selecting being based on the similarity measure.

A method for associating signal sources and paths includes determining secondary paths of a signal received at a reception point, wherein the signal reflects off one or more reflective surfaces before being received at the reception point, determining mirror sources of the secondary paths in accordance with locations of the one or more reflective surfaces and a main source of the signal, determining associations between the secondary paths and the mirror sources based on cross points at which the signal reflected off the one or more reflective surfaces, thereby obtaining path-source associations, and instructing use of the path-source associations in multi-source channel estimation.

In various embodiments, techniques are provided to determine channel characteristics of various communication systems such as OFDM systems or systems using a plurality of transmit antennas by using various sets of training symbols that produce zero cross-correlation energy. Channel communication can accordingly be simplified as the zero cross-correlation property allows for channel estimation without a matrix inversion.

The present invention relates to a frequency region equalization method and an apparatus therefor in a cellular environment on the basis of a QAM-FBMC system. A method for receiving a signal by a receiver according to one embodiment of the present invention may include: performing channel estimation; comparing a minimum channel gain with a predetermined first threshold value; and performing layered detection when the minimum channel gain is smaller than the first threshold value. According to one embodiment of the present invention, a gain can be acquired in terms of a BER even while a structure of a symbol level equalizer is maintained.

A method for associating signal sources and paths includes determining secondary paths of a signal received at a reception point, wherein the signal reflects off one or more reflective surfaces before being received at the reception point, determining mirror sources of the secondary paths in accordance with locations of the one or more reflective surfaces and a main source of the signal, determining associations between the secondary paths and the mirror sources based on cross points at which the signal reflected off the one or more reflective surfaces, thereby obtaining path-source associations, and instructing use of the path-source associations in multi-source channel estimation.

Methods and systems for obtaining improved channel estimates for frequency-multiplexed data transmissions such as OFDM, OFDMA, or SC-FDMA transmissions overcome the limitations of the static assumption by using a polynomial or other fitting function to fit and model the frequency dependence of the channel coefficients, so that estimates can be applied to larger subcarrier groups. Some embodiments provide channel estimates for a singular signal source, while other embodiments provide joint channel estimates for a plurality of signal sources. In embodiments, selection of the fitting functions is influenced by all previously determined channel estimates. In some embodiments, a tracking algorithm allows use of the lowest possible order of polynomial or other fitting function to model the frequency dependence of the channel coefficients, whereby the channel estimate is continually shifted in frequency while the order of the polynomial or other fitting function remains low.