Advanced space missions embrace constellation or swarm-like CubeSats, small reconfigurable satellite platforms, to achieve unprecedented spatial and temporal resolutions at a significantly lower investment. The realization of a high-speed inter-satellite optical communication link is essential to ensure the success of such CubeSat missions. CubeSats are a class of research spacecraft called nanosatellites. CubeSats are built to standard dimensions (Units or “U”) of 10 cm x 10 cm x 10 cm. Size, Weight, and Power-Cost (SWaP-C) have never become so crucial as it is in a typical CubeSat platform. The desired performance metrics are extremely challenging to achieve with the available resources defined by the satellite platform. Therefore, special design and optimization rules are indispensable to design a simple optical transceiver with full-duplex capability, fast-tracking speed, and a full field of regard. In this dissertation, I address the challenges pertinent to inter-satellite communication systems and present a potential omnidirectional communication method. In particular, I present transceiver design optimizations for omnidirectional optical communication, collaborative communication strategies between RF and optical for maximum reach, and CubeSat crosslink implementation based on the wavelength-selective-optical-transceiver design. I demonstrate the relations and dependencies among key transceiver design parameters such as scanning mirror’s smallest step angle, laser beam divergence, optics dimensions, and scanning area filling efficiency, etc. Additionally, the optimization challenges of the transmit laser beam size considering the interplay among beam divergence, beam clipping, and scattering are studied in detail. Besides, this dissertation presents the optical and mechanical design of the transceiver units that can ﬁt inside a state-of-the-art CubeSat to achieve an omnidirectional high-speed (more than 400 Mb/s) optical crosslink. Furthermore, a mathematical model is derived to investigate the link performance in the presence of angular pointing jitters for different receiver architectures. Alongside the statistical pointing error model, the derived model incorporates major receiver design parameters such as detector radius, receiver aperture size, F-number of the lens system, beam compression ratio, etc. It is shown that an optimum receiver design based on the presented model can achieve more than ﬁve orders of magnitude Bit Error Rate improvement even at large pointing errors.