Dear Rajkumar,

Greetings!

Each and every domain possesses the possibility of research. It's on your knowledge and the amount of literature survey you have gone through. If you want to pursue your research in fading, you need to study some emerging technologies (according to your interest). I recommend the following areas in wireless communication for you.

1) Effect of fading in the security of wireless network

2) NOMA and fading

3) wireless Channels for UAV

Best of luck

Wireless channels contain many propagation paths, with different amplitudes and phases. When a device moves around over distances proportional to the wavelength, the phases are changes rapidly and the paths might add constructively or destructively. If one models this statistically, you get a fading distribution.

Depending on the amplitude mix of the paths, different kinds of fading distribution appears. If all paths are equally strong, you will get Rayleigh fading. If one path is much stronger than the others, while the remaining ones are equally strong, you will get Rician fading.

If the fading doesn't match any of those categories, then the Nakagami distribution is a popular one that allows for fine-tuning a distribution to fit with measurements.

It can be of interest to evaluate the performance of communication systems in different fading scenarios, but one should avoid designing algorithms that are too reliant on a particular fading scenario because the distribution can change rapidly in practice. We talked about this in my podcast: https://ma-mimo.ellintech.se/2022/01/19/episode-25-what-models-are-useful/

Fading channel models are often used in wireless networks to model the effects of EM transmission. Therefore, it is always possible to work on parameters such as the change of signal strength according to location, antenna structure, wavelength in different environments and can also be created some alternatives for software and hardware support.

Well, these are three questions that have to be answered separately:

1) As Emil mentioned, Rayleigh and Rice fading models have been conventionally used as statistical models for fading in environments on which a relatively large (not so large in practice) number of paths are combined, in non-line-of-sight/line-of-sight scenarios, respectively. When either of these models did not work properly (for instance, did not fit well to field measurements), Nakagami-m model has been used because of its mathematical simplicity. The field of statistical channel modeling (or fading distributions) consists of providing a formulation of the resulting distribution that responds to a certain physical phenomenon, i.e., how electromagnetic waves interact with the environment. In the last two decades, several models (usually known as generalized fading models) have been proposed in the literature to model those propagation conditions for which Rayleigh/Rice/Nakagami-m models are not enough.

2) As a research area, statistical channel modeling is definitely a mature one. If there's one factor why it is somehow limited, it's because the resulting distributions of reasonably simple physical models become complicated. See for instance the TWDP model you mention, or just the sum of a number of waves (Article Stochastic Fading Channel Models With Multiple Dominant Spec...

3) Is there a chance to publish many papers in this area? In my opinion, the "many" should never be a target beforehand. As I said before, there are still scenarios on which formulating a new distribution which responds to a clear physical reality is needed. However, it is a challenging area because (1) it's not always possible to do this while keeping the model complexity limited -- models need to be reasonably tractable to be useful; and (2) there are already a number of existing physically-justified models with rather good performances. My advice is: don't go for the "many", go for the "good".

Traditionally path loss modeling is decomposed into a random fading factor and a deterministic distance-based component (often distance raised to some negative power). However, this approach ignores the fact that the distance is as unknown as the channel fading state and therefore should also be modeled as random (and channel estimation does not, in fact, estimate the fading component but both factors together). So fading models that combine the effects of multipath propagation and random distances are more comprehensive and relevant for wireless network analysis but not well researched yet. The distance part can be included in the fading, or vice versa, i.e., the small-scale fading can be viewed as a perturbation in the distance. With good fading, a transmitter appears closer, and with bad fading, it appears further away. Such a geometric interpretation of fading has led to some interesting new insights but is still in its infancy, especially when compared with the maturity of classical fading models.

Dear Martin Haenggi , please could you provide some references for those random distances models ?

Thanks

Sure, here are three:

M. Haenggi, "A Geometric Interpretation of Fading in Wireless Networks: Theory and Applications," IEEE Transactions on Information Theory, vol. 54, pp. 5500-5510, Dec. 2008.

X. Zhang and M. Haenggi, "The Performance of Successive Interference Cancellation in Random Wireless Networks," IEEE Transactions on Information Theory, vol. 60, pp. 6368-6388, Oct. 2014 (see Lemma 1, for instance).

Wireless networks appear Poissonian due to strong shadowing

B Błaszczyszyn, MK Karray, HP Keeler - IEEE Transactions on Wireless Communications, 2015 (this is about random distances combined with shadowing).

Dear Rajkumar Angamba Singh ,

In addition to valuable visions introduced by the respected colleagues, more study for channel modelling between fast moving objects in dispersive environment may be of interest.

Best wishes

Study of the statistics of channel that experiences multipath effect between TX and RX is relative mature. These models have been used by industry for quite a long time. You might looks at from multiple Txs not just a single Tx. By modeling the locations of Txs using stochastic geometry you will get some new directions. For example our work looks at the interference based on neighbor relationship. Discrete Exclusion Zone for Dynamic Spectrum Access Wireless Networks