One area involving cybersecurity and AI issues that is very much on the rise is Federated Learning. There are some open gaps for these AI architectures.

Federated Learning and its challenges such as non-IID, communication-efficiency, device-heterogeneity, and privacy are possible topics.

You can refer to the following papers:

Article Fast Federated Learning by Balancing Communication Trade-Offs

Preprint Fast Federated Learning by Balancing Communication Trade-Offs

@ Felipe Andre Zeiser

@Milad Khademi

Thank you both.. Will look it up.

There are number of research area in Cloud Computing itself like cloud security, Managing the resources in such a way that the cost of exaction of the user's task got minimized at the provider end. But now a days Fog Computing, which is an extension of Cloud Computing is emerging. And the major research issue in Fog computing is to determine how much part of the user's tasks needed to be executed on the Cloud and how much to be executed on the Fog Layer.

As for AI is concerned, there are multiple areas where AI can be used like in Image Processing, Image de-hazing, Satellite Images, Bio-Medical Images. Many More research areas are there.

Get a paper that is published not more than 3 years ago and it is in your area of interest and start reviewing it. That is a literature review. Look up for more papers that falls within that same topic. Now try to identify the strength and weakness of those proposals. Those weaknesses are your research gaps and identify ways to address those limitations.

Faisal A. Garba Thank you.

The hot topic related to AI is federated learning which can used to design algorithm for decentralized edge devices. The topic may be related to UAV control using federated learning.

Another hot research topic in IOT is cyber security attack detection of Distributed Denial of Service (DDoS)

Thank you @Grace U Nneji

@Grace U Nneji please can I request for a bit more information on the recommended research topic?

Olumayowa Ajayi Another direction you can consider is imaging and signal processing for AI applications. The world is becoming data-driven. More and more

data are generated in different formats. Most of the data are images and audio.

Federated learning IoT applications is very challenging due to computational issues. This can be a very good research topic for PhD. I have attached an article regarding this.

Article A Digital Twin-Based Drone-Assisted Secure Data Aggregation ...

Article FBI: A Federated Learning-Based Blockchain-Embedded Data Acc...

Thank you