Channel-dependent scheduling (CDS)

[Mangoman20]

I was in a presentation recently when someone mentioned the word CDS and its implementation within South African cellular data networks. I googled it and it sounded interesting and made some sense.

Is anybody familiar with this concept ? any idea if and when it will be implemented in SA ?

 

[guest2013-1]

June 2023

 

[UnUnOctium]

I am familiar. It's not implemented in networks rather than in standards. Simply put, the better a mobile user's channel, the more power (CDMA) or sub-channels (OFDMA) that will be allocated to it to exploit their potential for increased throughput. The opposite for users with worse channel conditions. I know of people in my research group whose research focuses in optimising scheduling.

 

[Mangoman20]

June 2023

Lol, thanks, I've set a calendar note

I am familiar. It's not implemented in networks rather than in standards. Simply put, the better a mobile user's channel, the more power (CDMA) or sub-channels (OFDMA) that will be allocated to it to exploit their potential for increased throughput. The opposite for users with worse channel conditions. I know of people in my research group whose research focuses in optimising scheduling.

Interesting, thanks for the feedback !

 

[UnUnOctium]

Ugh, just saw an error: with OFDMA it can be a mix of power and sub-channels (but mainly power) that are allocated.

I wouldn't be surprised (in fact I'd be shocked if it hasn't) if it's already enabled here in SA. Would be too wasteful and only ups efficiency compared to unbiased scheduling. Of course, it does have the trade-off that users at the edge of the cell would get a slightly degraded service with CDS whereas unbiased scheduling would be a little bit better for them, but a lot worse for the ones nearer to the BS.

This is especially true for LTE where OFDMA power loading is done by waterfilling type algorithms so it inherently has CDS.