Exploring real-world scenarios using Multi-Robot Systems (MRS) or Multi-Agent Systems (MAS) requires strategies that enable autonomous performance, including distributed coordination, scalability, robustness, and formation maintenance. Maintaining reliable formations under communication constraints and cyberattacks is critical for effective exploration and reconnaissance. While many studies focus individually on communication, security, or formation control, few provide a unified analysis of these interdependent challenges. This review addresses that gap by synthesizing existing coordination, cybersecurity, and formation strategies, highlighting how they collectively support robust formation maintenance. Communication limitations in dynamic or cluttered environments have been addressed using reinforcement learning (RL), epistemic planning, and event-triggered policies. Cybersecurity threats such as denial-of-service (DoS), false data injection (FDI), and masquerade attacks are prevented through decentralized filtering, redundancy-based detection, and consensus methods, including countermeasures for internally compromised agents. Formation control approaches, such as leader–follower models and finite-time consensus, are evaluated for stability, collision avoidance, and resilience under adversarial influence. We identify remaining gaps, emerging trends, and open research questions to guide the development of resilient MRS capable of sustaining formation and mission performance under complex, adversarial conditions.