NCSA: A New Protocol for Random Multiple Access Based on Physical Layer Network Coding

This paper introduces a random multiple access method for satellite communications, named Network Coding-based Slotted Aloha (NCSA). The goal is to improve diversity of data bursts on a slotted-ALOHA-like channel thanks to error correcting codes and Physical-layer Network Coding (PNC). This scheme can be considered as a generalization of the Contention Resolution Diversity Slotted Aloha (CRDSA) where the different replicas of this system are replaced by the different parts of a single word of an error correcting code. The performance of this scheme is first studied through a density evolution approach. Then, simulations confirm the CRDSA results by showing that, for a time frame of $400$ slots, the achievable total throughput is greater than $0.7\times C$, where $C$ is the maximal throughput achieved by a centralized scheme. This paper is a first analysis of the proposed scheme which open several perspectives. The most promising approach is to integrate collided bursts into the decoding process in order to improve the obtained performance.

💡 Research Summary

The paper introduces a novel random multiple‑access protocol for satellite communications called Network Coding‑based Slotted Aloha (NCSA). Unlike traditional Slotted Aloha or its enhanced version Contention Resolution Diversity Slotted Aloha (CRDSA), which rely on transmitting several identical replicas of a packet, NCSA encodes each user’s data with an (n, k) linear error‑correcting code and distributes the n coded symbols across n randomly selected slots. When multiple users transmit in the same slot a collision occurs, but instead of discarding the collided signal, NCSA applies Physical‑layer Network Coding (PNC) to extract the XOR of the overlapping symbols.

At the receiver, slots that contain a single uncoded symbol (singletons) are first decoded directly. The recovered symbols are then used to cancel their contribution from other slots, similar to successive interference cancellation (SIC). For the remaining collided slots, the PNC‑derived XOR information is combined with the already decoded symbols to form a system of linear equations. Solving this system yields the original codewords for each user, thereby recovering all transmitted data without needing multiple identical replicas.

To evaluate performance, the authors extend the density‑evolution framework traditionally used for SIC‑based protocols. The new analysis incorporates the expected amount of useful PNC information that can be harvested from a collision, leading to revised state‑transition probabilities for “empty”, “singleton”, and “collision” slots. This analytical model predicts the evolution of the decoding graph as the frame progresses and identifies a stable operating region where the decoding process converges with high probability.

Simulation results for a frame of 400 slots confirm the theoretical predictions. NCSA achieves a total throughput exceeding 0.7 × C, where C denotes the maximum throughput attainable by an ideal centralized scheduler. This represents a substantial improvement over CRDSA, which typically reaches about 0.55–0.60 × C under the same conditions. The paper also explores the impact of key design parameters: the code rate (k/n), the number of slots per frame, and the user population. A higher redundancy (lower code rate) reduces collision probability and improves decoding success, but consumes more slots per user; a lower redundancy increases slot efficiency but places greater reliance on the PNC extraction capability.

The authors acknowledge that the current implementation treats collided slots only as sources of XOR information and does not fully exploit joint multi‑user decoding. Future work is proposed to integrate collided bursts directly into a more sophisticated network‑coding‑based multi‑user detection algorithm, which could further raise throughput and reduce latency. Additional research directions include adapting the scheme to realistic satellite channel impairments (e.g., non‑Gaussian noise, Doppler shifts), handling asynchronous transmissions, and optimizing power allocation under satellite payload constraints.

In summary, NCSA merges physical‑layer network coding with error‑correcting coding to transform collisions from a detrimental event into a valuable source of information. The density‑evolution analysis and simulation evidence demonstrate that this approach can significantly outperform existing random‑access schemes, offering a promising avenue for high‑efficiency, low‑latency access in bandwidth‑limited satellite and other wireless networks.