Design of Coded Slotted ALOHA With Interference Cancellation Errors

Coded Slotted ALOHA (CSA) is a random access scheme based on the application of packet erasure correcting codes to transmitted packets and the use of successive interference cancellation at the receiver. CSA has been widely studied and a common assumption is that interference cancellation can always...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on vehicular technology 2021-12, Vol.70 (12), p.12742-12757
Main Authors: Dumas, Charles, Salaun, Lou, Hmedoush, Iman, Adjih, Cedric, Chen, Chung Shue
Format: Article
Language:English
Subjects:
Access protocols
Asymptotic properties
Cancellation
Coded slotted ALOHA
Codes
Computer Science
Decoding
decoding failure
Density
density evolution
Errors
Evolution
Interference
Interference cancellation
Networking and Internet Architecture
Packet transmission
Parameters
Performance evaluation
Probability distribution
Random access
SIC failure
Throughput
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Coded Slotted ALOHA (CSA) is a random access scheme based on the application of packet erasure correcting codes to transmitted packets and the use of successive interference cancellation at the receiver. CSA has been widely studied and a common assumption is that interference cancellation can always be applied perfectly. In this paper, we study the design of CSA protocol, accounting for a non-zero probability of error due to imperfect interference cancellation (IC). A classical method to evaluate the performance of such protocols is density evolution, originating from coding theory, and that we adapt to our assumptions. Analyzing the convergence of density evolution in asymptotic conditions, we derive the optimal parameters of CSA, i.e., the set of code selection probabilities of users that maximizes the channel load. A new parameter is introduced to model the packet loss rate of the system, which is non-zero due to potential IC errors. Multi-packet reception (MPR) and the performance of 2-MPR are also studied. We investigate the trade-off between optimal load and packet loss rate, which sheds light on new optimal distributions that outperform known ones. Finally, we show that our asymptotic analytical results are consistent with simulations obtained on a finite number of slots.
ISSN:0018-9545
1939-9359
DOI:10.1109/TVT.2021.3120069