A Novel Framework of K-repetition Grant-free Access via Diversity Slotted Aloha (DSA)

TL;DR

This work tackles the challenge of uncoordinated uplink access in massive mMTC by proposing α-IIC-DSA, an AP-centric MUD framework for K-GF-NOMA that exploits signal correlation across resource blocks to perform SIC-based decoding. The method iteratively cancels interference using MAI derived from previously decoded MTCD signals, with a simple two-block structure at the AP: interference cancellation (IC) and decoding with CRC (Dec_CRC); the depth of iteration is controlled by α. The paper provides a general model, a complexity analysis across three regimes (α=2, α with $N/2 \gg α$, and α=N), and a case study showing how α, K, and traffic intensity $\gamma$ influence access probability and computational load, revealing important trade-offs. The results suggest that α-IIC-DSA is a viable approach for scalable, AP-centric MUD in 5G/B5G mMTC deployments, where device-side complexity remains low while the AP handles increased processing demands.

Abstract

This article introduces a novel framework of multi-user detection (MUD) for K-repetition grant-free non-orthogonal multiple access (K-GF-NOMA), called $α$ iterative interference cancellation diversity slotted aloha ($α$-IIC-DSA). The proposed framework targets at a simple yet effective decoding process where the AP can intelligently exploit the correlation among signals received at different resource blocks (RBs) so as to generate required multi-access interference (MAI) for realizing the signal-interference cancellation (SIC) based MUD. By keeping all operation and hardware complexity at the access point (AP), the proposed framework is applicable to the scenarios with random and uncoordinated access by numerous miniature mMTC devices (MTCDs). Numerical experiments are conducted to gain deep understanding on the performance of launching the proposed framework for K-GF-NOMA.

Figures (5)

• Figure 1: Standard signaling of K-GFA (a) K-Repetition Grant-Free Transmission Procedure (b) resource structure of a time frame and MTCDs distribution over RBs ($R$ = 10, $N$ = 5, $K$ = 3).
• Figure 2: Flow chart of the decoding process for the proposed $\alpha$-IIC-DSA framework.
• Figure 3: The examples of the proposed $\alpha$-IIC-DSA module when $\alpha$ = 2, 3.
• Figure 4: Expected access probability on $\alpha$-IIC-DSA system when $\alpha$ = 1, 2, $\infty$ and $K$ = 2, 3, 4, 5 for different $\gamma$.
• Figure 5: Expected access probability and complexities of the blind IC process when $K$ = 2, 4 and $\gamma$ = 0.3, 0.6, 0.9 for $\alpha$ ranged from 1 to 5, where $C_{wr}$, $C_{dec}$ and $C_{sto}$ denote average complexities for memory write-read, decoding and storage, respectively