Diversity combiner with MLSE for digital cellular radio

Claim 1

An apparatus for combining signals Z 1 (k) and Z 2 (k) received on different antennas over a multipath fading channel comprising: first and second channel estimators, each receiving said signals Z 1 (k) and Z 2 (k) and providing estimates of each channel associated with said signals Z 1 (k), Z 2 (k) as f 1 (n, k) and f 2 (n, k); and an MLSE equalizer for receiving each of said signals Z 1 (k) and Z 2 (k), and said channel estimates f 1 (n, k), and forming a single estimate of symbol values contained in said signals Z 1 (k) and Z 2 (k) wherein said MLSE equalizer in forming said single estimate executes a Viterbi algorithm which computes a branch metric for each symbol I(k) contained in said signals Z 1 (k) and Z 2 (k), as ##EQU4## where σ 1 (k) and σ 2 (k) are standard deviations of n 1 (k), n 2 (k) representing the noise in each channel at the time of each symbol time (k).

Claim 2

An apparatus for combining signals Z 1 (k) and Z 2 (k) received on different antennas over a multipath fading channel comprising: first and second channel estimators, each receiving said signals Z 1 (k) and Z 2 (k) and providing estimates of each channel associated with said signals Z 1 (k), Z 2 (k) as f 1 (n, k) and f 2 (n, k); and an MLSE equalizer for receiving each of said signals Z 1 (k) and Z 2 (k), and said channel estimates f 1 (n, k), and forming a single estimate of symbol values contained in said signals Z 1 (k) and Z 2 (k) wherein said MLSE equalizer in forming said single estimate executes a Viterbi algorithm which computes a branch metric for each symbol I(k) contained in said signals Z 1 (k) and Z 2 (k), as ##EQU4## where σ 1 (k) and σ 2 (k) are standard deviations of n 1 (k), n 2 (k) representing the noise in each channel at the time of each symbol time (k).

Claim 6

A method for processing modulated symbol data I(k) transmitted via a carrier signal over a communications channel subject to multipath fading, comprising: receiving said symbol data over first and second antennas within said communications channel; demodulating said symbol data received from each antenna to form first and second streams of symbol data, Z 1 (k) and Z 2 (k); estimating from said first and second streams of symbol data an amplitude and phase error function f 1 (n, k) and f 2 (n, k) for each of said first and second streams of symbol data; estimating said symbol data I(k) with MLSE equalizer using said streams of symbol data Z 1 (k) and Z 2 (k) and said error functions f 1 (n, k) and f 2 (n, k); wherein each of said streams of symbol data Z 1 (k) and Z 2 (k) can be expressed as: ##EQU5## wherein n 1 (k) and n 2 (k) represent the estimated uncorrelated noise in each received signal.

Claim 7

A method for processing modulated symbol data I(k) transmitted via a carrier signal over a communications channel subject to multipath fading, comprising: receiving said symbol data over first and second antennas within said communications channel; demodulating said symbol data received from each antenna to form first and second streams of symbol data, Z 1 (k) and Z 2 (k); estimating from said first and second streams of symbol data an amplitude and phase error function f 1 (n, k) and f 2 (n, k) for each of said first and second streams of symbol data; estimating said symbol data I(k) with MLSE equalizer using said streams of symbol data Z 1 (k) and Z 2 (k) and said error functions f 1 (n, k) and f 2 (n, k); wherein each of said streams of symbol data Z 1 (k) and Z 2 (k) can be expressed as: ##EQU5## wherein n 1 (k) and n 2 (k) represent the estimated uncorrelated noise in each received signal.