I think you need to explain your premises, to show what solution is the most advantageous. From the abstract, it sounds like the user equipment you are postulating can use both WiFi and LTE, and you are also postulating that the WiFi local network is connected to the Internet via some unknown method, but not via this same LTE macrocell. Furthermore, you are also postulating that there is no difference in the cost structure, for service to the user. The other thing I believe you are postulating is that there is a difference in how LTE and WiFi manage "priorities," where WiFi is best effort but LTE is not.

If I guessed correctly, I think these premises need to be specified, because they might not describe what the reader assumes, or what is the normal case.

I think that today, LTE is mostly used for best effort IP traffic, much like WiFi, but that the LTE service is capacity-constrained, for example, with monthly data caps. Whereas most of the time, the WiFi network is connected to the Internet over a fixed service ISP, with either much higher monthly data caps, or none at all, and usually also with higher bit rates up and down, than the LTE service at that location. In this common scenario, the WiFi is always given priority. It makes sense, right?

Other times, the WiFi local network uses LTE macrocells to reach the Internet. For example in mobile hotspots. In this case, there would be no point differentiating between the WiFi and the LTE service. WiFi would only be used as a local interface standard, for convenience. There's a lot of user equipment out there that has no LTE interface, only WiFi or Ethernet.

So I would start by defining your assumptions first, because your optimal solution depends very heavily on these initial assumptions.

Now, let's assume my guesses on your assumptions are correct. If a local user equipment needs access only to another local equipment, why would it not always choose the WiFi network, instead of going to an LTE macrocell, and then back to the local equipment? Given that WiFi is usually much faster service than LTE, even if LTE offered some kind of priority differentiation, it would make no practical difference. Faster speed has long been the easy answer to service differentiation. "Throw cheap bandwidth at the problem," and much of that fancy service differentiation becomes unnecessary.

Dear Albert,

Thank you so much for for your explanation. Yes. I agree with you bu my problem is with the CACA regarding the minimum bit rate. In my previous model, I had only LTE cell with different zones where I assumed that there is a maximum capacity in the cell and the user is admitted if the bit rate of the user is >= the minimum bit rate. Now my question is can I assume the same for this model but with different minimum bit rate to each zone (LTE & WIFI)? .

I believe that you will understand my model if I send you my model and assumptions (attached).

I prefer other solution rather than LTE-WiFi integration. LTE is a costly network as spectrum is not free. WiFi is free in term of spectrum. So, ITU or other standard body is expected to include more high speed frequencies to ISM (Industrial Scientific and medical) standard, and introduce that frequencies to WiFi standard, so WiFi will have higher bit rate..and cheaper price :)

Dear Suherman.

Thank you for your answer. Yes you are right that WIFI is cheaper however, my target is to keep the user up all the time with better QoS. It is easy for the WIFI to be disconnected from time to time. This is why I have suggested this integration. Please look at my suggested attached model and revert to me please with your further suggestions.