论文学习----URLLC benefit from NOMA (1)

Title"How URLLC can Benefit from NOMA-based Retransmissions"
Author:Radosław Kotaba, Carles Navarro Manchon, Tommaso Balercia, ´and Petar Popovski
Note:通过这篇论文，初步了解了结合NOMA的HARQ以及不同描述CSI情况下的优化调度问题。

BS Operation

Symbols used

b:probability of the new packets generating
N:device numbers
L: index of slot
W:number of resources
K:number of channel uses,i.e. symbols, the same for packets of each UE
B: same amount of B information
R:rate =B/K(bits/symbole)

${\mathbf{y}}_i={\sum }_{j\in {\Gamma }^i}\sqrt{P_{i,j}}{g}_{i,j}{\mathbf{x}}_{i,j}+{\mathbf{n}}_i$
$P_{i,j}\in \mathbb{N}$ :transmit power of user j in TF-block i
$g_{i,j}\in \mathbb{C}$ and $g_{i,j}=\frac{h_{i,j}}{\sqrt{d_j^{\alpha }}}$ ：channel between j-th UE and BS over the i-th TF-block,where $h_{i,j}$ is Rayleigh fading component, i.i.d zro mean circularly symmetric complex BGaussian random variable with unit variance, $d_j\sim U(D_{min},D_{max})$ is uniformly distributed
${\mathbf{x}}_{i,j}\in {\mathbb{C}}^K$ :complex additive white Gaussian noise with zero mean and variance ${\sigma }^2$

HARQ with Statistical CSI

QMA-HARQ

1. The SNR of the packet received from user j is distributed as follow
   

2. the decoding failure probability after l-th attempt (counting from 0 as the initial one) is given
   
   When IR-HARQ is used, then
   
   rearrange the equation above with the last packet realization since all the previous SNRs are already known by introducing two terms ${\gamma }_{c{c}_j}^l$ and ${\gamma }_{i{r}_j}^l$
   

3. error probability
   
   mininum power required,

4. optimization problem
   BS’s goal is to spend (on average) as little power on a packet as possible while providing certain reliability guarantees
   
   for different mechanism ,it can be simplified
   Chase Combinine:
   
   Incremental Redundancy
   

NOMA-HARQ

1.defining the error probability of the transmission over shared
channel resources.

Term1:error probability with interferer’s signal decoded and canceled
Term2:the case when SIC is not successful
Closed form:

2. Optimization problem