On 5/28/20 7:14 AM, Hassam Mughal wrote:
> Hello Dr. Tom,
>
> A bundle of thanks for your explanation. So, what would be the best way
> to calculate the available bandwidth at each node before transmitting a
> packet? What I think is if I use the total channel capacity and the
> previous interval throughput value, and based on that calculate the
> available bandwidth. Or, is there some other proper way, that will
> provide me with the exactly available bandwidth on each wireless
> interface. In my simulation, I am using two wifi interfaces, one with
> IEEE 802.11b, DsssRate11Mbps and ConstantRateWifiManager with Data Mode.
> While on the other Wifi Interface, I am using IEEE 802.11n_5GHZ with
> ErpOfdmRate36Mbps and MinstrelHtWifiManager with NonUnicastMode.
Please keep in mind that Wi-Fi is a shared channel, so if you talk about
available bandwidth or total channel capacity in this context, it is
more complicated. The available bandwidth that you have at any given
time depends on who you are sending to, and who else in your radio range
is also sending at the same time. Also, channel capacity is an
information theory concept and not something typically related to
simulation.
In short, I do not think you can just use a term like 'available
bandwidth' without explaining exactly what you mean by it in your
context. Rather than try to come up with definitions on the mailing
list, I suggest instead that you survey the literature and see how
others have used these terms, and adapt one or more of the definitions
to align with your scenario.
For instance, if you go to
https://scholar.google.com and search for
'wifi available bandwidth', you will find many papers that use the term,
and other terminology can be similarly searched.
- Tom