Page 26 - Read Online

Phadke et al. Intell Robot 2023;3:453-78 https://dx.doi.org/10.20517/ir.2023.27 Page 471

specific or generalized. Multi-UAV cooperative task assignment problems by considering different UAV
[118]
nature types are considered in . Routing protocols strengthen data transmission and optimize networks
for FANETs. Biologically inspired algorithms often work on natural swarming and exchanges observed in
[119]
real-world animal gatherings such as for ants, bees, and wolves. Article uses a novel ant colony
optimization algorithm with fuzzy logic for improving UAV swarm routing performance. Fuzzy models can
make decisions in uncertain environments commonly found in swarm operations. The predicted advantage
of fuzzy models is their efficiency in performance when achieving high throughput in large network loads
for mobile agents. Efficient routing protocols form the basis of any resilient communication component.
Although every study usually has its own set of proposed or existing methods that it uses to validate
performance, perhaps via Monte Carlo methods or comparisons with existing state-of-art, there is a
necessity for universal scalable metrics for the assessment of UAV swarm resilience. The very nature of such
proposals is not to fit into a single application domain but rather to promote widespread usage of common
metrics. Using common assessment methods and metrics makes it easier to compare the multitude of novel
techniques with each other. Although such literature is sparse, methodologies such as those in introduce
[120]
baseline assessment methods for swarm resilience based on complex networks.

4. DISCUSSION
Our investigation has uncovered a multifaceted landscape of UAV swarm resilience that underscores the
complexity inherent in diverse application scenarios. The tailoring of resilience mechanisms to match
specific challenges within each scenario is evident, reinforcing the importance of a context-driven approach.
By examining the responses of individual components and their collective behavior, we have unraveled
emergent properties that enhance the overall resilience of the swarm. Using keyword analysis and network
visualization allows the researcher to access the range and spread of the research topic in question and
identify key points of entry for further research directions.

The significance of this study extends to the broader realm of research on UAV swarm applications and
resilience. Our findings emphasize the need for a holistic understanding of the interplay between
components, recognizing the potential for emergent behavior. This insight extends beyond the immediate
scope of resilience, highlighting the importance of considering system-level behavior when designing and
deploying UAV swarm systems.

While our study aims to contribute valuable insights, it is essential to acknowledge the limitations we
encountered in crafting resilient case studies for UAV swarm scenarios. Literature dataset curation for
surveys is often influenced by the authors’ perspectives on trends and how they are interpreted. This can be
remedied by creating up-to-date surveys on the research topic that offers multifaceted viewpoints. This
allows the reader to explore a wide range of possibilities and prevents biased outlooks. Additionally, it is
impossible for a single survey to accurately cover every single research article and methodology. Multiple
surveys offer a broader coverage of the topic, ensuring that all key literature is included.

The inherent complexity of real-world applications often defies complete emulation in controlled
environments. Overcoming this challenge calls for collaborative efforts between researchers, industry
partners, and regulators to create realistic and representative testbeds for comprehensive resilience
evaluations. When it comes to the individual agents that make up the swarm, they are often influenced by
ongoing changes in regulations that require certain changes to them. For example, the FAA has enforced
[122]
the RID (Remote Identification) rule that requires certain classes of UAV agents to have an open broadcast
module that will transmit the location and certain identifying information of the agent and operator at all
times. Regulations such as this are certain to influence the way this information is processed by swarms and