Frequency-offset immunosensors based on acoustic wave devices are known to provide extremely high sensitivity and selectivity where the target is detected and identified based on the amount of frequency shift. We propose a new method to further classify chemically similar molecules extrapolating on the concept of in-phase (I) and quadrature (Q) domain used for the detection of orthogonal M-ary signals in digital telecommunication systems. We performed a series of detection experiments using samples of explosives such as cyclotrimethylene trinitramine [or royal demolition explosive (RDX)] and trinitrotoluene (TNT), containing nitrous oxide (NO2) groups and chemically analogous substances (e.g., musk oil). This detection scheme involves the use of semi-orthogonal monoclonal anti-TNT and anti-RDX antibodies immobilized onto two separate sensor surfaces. The term semi-orthogonal represents the co-option of a term used heavily in digital radio for the purpose of describing chemical orthogonality. The antibody to TNT which we use has some reactivity with RDX, and other nitrous oxide compounds. This feature of an antibody is referred to in the literature as antibody promiscuity. The antibody for RDX which we use shows very little cross reactivity with other molecules and, hence, the chemical responsiveness of the two antibodies is not quite orthogonal. Their responses are then chemically semi-orthogonal. The two semi-orthogonal immunosensor responses were then monitored and the baseline frequency shifts were recorded. After remapping the measured frequency data of the analytes onto a new 2-D domain by setting the TNT-specific sensor as the (I) or real component and the RDX-specific sensor as the (Q) or imaginary component, we could observe that all the substances were detected and mapped out to distinct regions on the I-Q plot. We assert that there is a strong resemblance between digital radio system quadrature detection techniques and our I-Q mapping scheme of the semi-orthogonal immunosensor signatures.