I do not consider IoT as a technology. IoT is a concept that involves many technologies. We have different technologies regarding devices (hardware), many technologies regarding wireless communications, many technologies regarding software (applied to devices - hardware), etc.
Considering what I said, we can explore trends in each of these mentioned technologies (devices, communications, software, ...) or we can explore trends in IoT applications considering other concepts such as AI/machine learning, data security/privacy, blockchain, access control, etc.
For instance, you can start some research considering:
some communication technologies, such as Lora and Sigfox;
some techniques to protect data in IoT applications (considering data in transit and rest);
some reliable architecture to efficiently transfer sensitive data for critical applications...
I think a good start point is to define an application scenario as a base case, exploring possible problems with it, and then propose some solutions by applying some existent technologies/techniques or proposing new ones.
One could also research the big picture, what happens to society when everything important and trivial is automated, "spies on", "talks to" and "controls" everything else, while automation components become obsolescent, fail, misbehave, can be "taken over," reprogrammed, "lie" actively engage in harm?
A trending research area in IoT is its application in industry, namely, industrial Internet of Things (IIoT). It is a hot topic currently and Its standardization is on the way. IIoT could be implemented in different areas: Mining, Smart Grid and Power Industry, Transportation, Food Industry, and so on. Challenges included but not limited to Energy Consumption and Management, Interoperability of Devices, Security and Privacy of Data and Workers, and Context and Semantics-Aware Service Provisioning.
You could refer to the following paper for more information:
1. Article Online Optimization of Wireless Powered Mobile-Edge Computin...
2. Article Data Age Aware Scheduling for Wireless Powered Mobile-Edge C...