The performance of a coherent telescope array (CTA) as an optical communications receiving system is analyzed with regards to the benefits and disadvantages in comparison to monolithic large-aperture single telescopes and conventional photon bucket systems, especially in terms of background photons. The bit error rates for differential phase-shift keying (DPSK) and binary pulse position modulation (BPPM) schemes are derived, the former using self-homodyne interferometry as a novel demodulation technique, in which the background photons interact incoherently. The possibility of further dividing the receiver's aperture into N>2 smaller subtelescopes is explored, and its adaptability for implementing the N binary phase-shift keying (N-BPSK) technique with optical code division multiplexing (OCDM) is discussed. An N-CTA receiver system with an N-BPSK/OCDM technique is envisioned not only to achieve multiterabit per second data transfer capability, but also to adapt a novel noise suppression technique based on photon correlations, which would render a need for expensive and complicated adaptive optics obsolete. Further combined with significant reduction in weight and overall costs, an N-CTA system with an N-BPSK/OCDM technique could serve in deep space communications as well as in defense to achieve information dominance in the battlefield.