As technology trends towards the ubiquitous presence of Internet of Things (IoT) solutions in everyday life, implementations to facilitate the communication and optimise functionality become more and more key. For this research, we have looked at Isolation Forests (iForests) as a Machine Learning (ML) solution and approach to the optimisation of communication efficiency and security. To this end, a dataset containing data that can be interpreted as a Directional Graph which can then be extrapolated and interpreted by this model to flag anomalous communications and to return a certain of accuracy. In this context, we then plot accuracy scores returned by our model to ascertain the accuracy of anomaly detection during legitimate communication, as well as some strained scenarios like bruteforce and flood attacks. Looking at anomaly scores as well as duration of communications, we can then establish that communication is healthy, or under attack, and even which sort of attack it is most likely to be.