Digital Communications Thread

[UnUnOctium]

Anyone interested in learning the more advanced topics about digital communications?

I noticed that a lot of questions fly out on MyBB about signal reception, etc. and thought I would make a thread, if people want to read it, which would answer 95% of them

I guess it would cover everything from simplified channel models to link budgeting to GSM/WiMAX/UMTS/LTE standards.

Just want to know if I will be typing for nothing or if there are any people who are going to bother to read and ask questions

 

[UnUnOctium]

Meh, I guess no one.

 

[copacetic]

Meh, I guess no one.

I'm sorry man.

 

[MickeyD]

Great idea - why not blog it and then we can point folks to it whenever they post a query? Similar to what happened with "Ordering ADSL for Dummies"...

 

[UnUnOctium]

Or just link this thread. Either way, problem is that I'd rather answer questions than just start spamming away everything I know because it will take ages (not to boast). Then questions can be put in appropriate order and answered by editing the first post. Thing is, if no one really wants to know then it's kind of pointless to do it

 

[MickeyD]

Question 1:
What is cell breathing and how can it impact me?

 

[UnUnOctium]

Question 1:
What is cell breathing and how can it impact me?

Lol, you know the answer to that one But I'll give an answer anyways, out of principle.

Cell breathing is the reduction and increase in the coverage range/radius of a base transceiver station (BTS) mainly due to user load. This is very prevalent in UMTS-based BTSs which use W-CDMA because of the nature of direct sequence CDMA (DS-CDMA, the modulation scheme W-CDMA is based on) requires precise power control per user. When the number of users are increased, less power can be allocated to each user (because the BTS has a maximum transmit power by law, tinfoil hats!), therefore reducing the amount of power with which the signal can be transmitted to the device and reducing range in turn.

What this means to you:
Heavy load (large amount of users) will increase how often your device does cell-hopping (changing BTSs) and/or decrease your throughput (depending on whether your device hops in the first place). Vice-versa applies.

 

[MickeyD]

Question 2:
Why cannot Telkom or Neotel just start providing everyone with optic fibre connections to their homes? Surely I can just plug my phone and modem into the jack and it will still work?

 

[UnUnOctium]

Question 2:
Why cannot Telkom or Neotel just start providing everyone with optic fibre connections to their homes? Surely I can just plug my phone and modem into the jack and it will still work?

Lol, because they're lazy as sh*t and maybe because the fibres (made of plastic) aren't the best electrical conductors nevermind the fact that the equipment on the other side needs to be changed too

 

[battletoad]

ive studied stochastic calculus and coding theory, i hear they're useful in communications. Is stochastic modelling really that widespread in the industry?

Q2: has there been some sort of fusion between coding theory and stochastic modelling for the purposes of bettering communications? It somehow makes an awful lot of sense to me.

And finally, what are the industry (fixed line and/or mobile) standard codes, ie reed-solomon, bch, etc.?

 

[UnUnOctium]

ive studied stochastic calculus and coding theory, i hear they're useful in communications. Is stochastic modelling really that widespread in the industry?

In research, yes. Communications channels are usually assumed to be WSSUS, which is a stochastic process itself. So yes, channel modelling sees the most use of stochastic maths.

Q2: has there been some sort of fusion between coding theory and stochastic modelling for the purposes of bettering communications? It somehow makes an awful lot of sense to me.

Those two topics aren't that related. Coding theory has deep roots in information theory but doesn't borrow much at all from stochastic processes. Give me an idea of what sort of fusion you mean.

And finally, what are the industry (fixed line and/or mobile) standard codes, ie reed-solomon, bch, etc.?

Quite varied depending on your application but everyone is now shifting towards Turbo and LDPC codes. Big craze for LDPC codes now because of their fast computation and good code rate.

A good example of the evolution of inner FEC: DVB-T -> convolutional, DVB-T2 -> LDPC. In terms of wireline, there isn't much left to squeeze out as the rates achievable for the SNR & channel bandwidth are very close to the Shannon limit. DOCSIS v3 recently came out and even the working group for that standard admitted that it'll probably be the last DOCSIS standard. Most are focusing on wireless.