With recent developments in global navigation satellite systems (GNSS), dual- and triple-frequency reception is becoming increasingly available. This can be used for ionospheric error removal, but further research efforts are required to successfully apply it to mitigate and detect errors caused by multipath and non-line-of-sight reception. To review the current state of the art on this topic, the previously published multi-frequency fusion methods for minimization of multipath and non-line-of-sight errors are summarized and classified. The main input is the pseudorange of each frequency, but in addition, three main categories of inputs are identified, i.e., signal-to-noise ratio, carrier-phase measurements, and correlator output. Existing works that use these parameters as input for frequency fusion mainly consider stationary applications, error detection, and the minimization of the carrier phase multipath error. To develop the state of the art to also suit dynamic scenarios and to reduce the errors further, the information contribution of each individual parameter needs to be studied, to prepare the way for a frequency fusion algorithm that combines these information contributions optimally.