In orthogonal frequency division multiplexing systems, iterative maximum likelihood channel estimation and signal detection is preformed. The channel estimation gives the maximum likelihood estimates of time domain channel parameters. A cost function is defined. An iterative process estimates a joint channel impulse response and a transmitted signal that minimize the cost function, and thus solves the channel estimation and signal detection problem jointly without having specific knowledge of the channel.

During receiving of message signals, in particular radio signals, it is particularly important to free the receiving signal of interference, such as, e.g. noise or various kinds of interference. It would be desirable if the receiver had knowledge of the type of occurring interference in advance, before processing the received signal, in order therefore to be able to improve considerably signal processing, in particular equalisation and demodulation, of the receiving signal. A receiving method is here proposed, in which first the auto-correlation function of the spurious components contained in the received signal is estimated. Then the estimated auto-correlation function is compared with presettable sample functions and the sample auto-correlation function which most closely corresponds to the estimated auto-correlation function is chosen. Finally, by means of the selected sample function, further signal processing, in particular equalisation, of the received message signal is carried out. Therefore a matching sample function is determined in advance for the most strongly occurring spurious component.

A suboptimal method for searching for a symbol sequence, the method comprising the steps of: determining a channel impulse response; sampling a received signal; selecting at least one of the highest and/or most reliable impulse response values; determining a reference signal using at least one impulse response value and a symbol sequence assumed as transmitted; determining differential terms corresponding to the selected impulse response values for the signal sample and the reference signal; using the determined differential terms in a transition metric for searching for a symbol sequence; forming a survivor path by adding the symbol sequence provided by the transition metric to the survivor path formed so far.

A method of building a model for a physical plant in the presence of noise can include initializing the model of the physical plant, wherein the model is characterized by a parameter vector, estimating an output of the model, and computing a composite cost comprising a weighted average of an error between the estimated output from the model and an actual output of the physical plant, and a derivative of the error. The method further can include determining a step size and a model update direction. The model of the physical plant can be updated. The updating step can be dependent upon the step size. Another embodiment can include the steps of determining a Kalman gain and determining an error vector comprised of two entries weighted by a scalar parameter.

The invention relates to a method for channel equalization in a receiver, in which a signal is received from a communication channel, the signal containing symbols formed of binary information by phase shift keying. In the receiver, channel estimation is performed to estimate the properties of the communication channel, and samples are taken of the received signal at intervals. In the method, a determined number of samples is examined at the time, and a decision step is taken, in which, to find out the transmitted symbols, bit decisions are computed on the basis of said determined number of samples. After each decision step, it is examined, whether said decision step is to be iterated. Upon iteration of said decision step, at least some of the bit decisions of the preceding decision step are used in addition to the samples under examination at the time, for computing the bit decisions.