Multi-Connectivity for Multicast Video Streaming in Cellular Networks (Extended Abstract)

Sadaf ul Zuhra, Prasanna Chaporkar, Abhay Karandikar, H. Vincent PoorarXiv 2023 · DOI · arXiv

Introduction

In cellular multicast video streaming, network congestion often dictates the quality of experience. While multicast allows multiple users to receive the same data stream simultaneously, it suffers from a "weakest link" problem: the transmission rate is constrained by the user with the poorest signal quality, typically those at the cell edge. This paper explores Multi-Connectivity (MC) as a solution, enabling users to aggregate signals from multiple base stations to improve reception reliability and overall system throughput.

Problem Formulation and Complexity

The authors model the challenge as a resource allocation problem ($K^\star$) aimed at maximizing the number of successfully served users in a multi-cell environment. They demonstrate that finding an optimal solution is NP-hard, as it maps to a variation of the Maximum Coverage Problem.

Methodology: The CGA Algorithm

To address the computational complexity, the paper introduces the Centralized Greedy Approximation (CGA) algorithm. This iterative approach selects the base station and physical resource block (PRB) that provides the maximum incremental coverage—serving the highest number of currently unserved users at each step.

The Centralized Greedy Approximation (CGA) algorithm for resource allocation.

Theoretically, this greedy strategy is shown to achieve a $(1 - 1/e)$ approximation ratio (roughly 63%), which is the best possible performance for any polynomial-time algorithm for this class of problem, assuming $P \neq NP$.

Evaluation and Key Results

The proposed method was evaluated through numerical simulations of a seven-cell system with varying user densities and cell sizes. The results highlight two major advantages of Multi-Connectivity:

  1. Scalability with User Density: As the number of users per cell increases, the gap between single-connectivity and multi-connectivity widens. For instance, at 240 users per cell, MC reduces the number of unserved users from approximately 50 to roughly 20.
  2. Robustness to Cell Size: In larger cells (e.g., 400m radius), signal degradation at the edge becomes severe. While unserved users spike in single-connectivity setups, the MC approach maintains a significantly lower and flatter growth curve, demonstrating its efficacy for cell-edge users.
Performance comparison of single vs. multi-connectivity across varying user densities (a) and cell radii (b).

Conclusion and Future Directions

Multi-connectivity effectively mitigates the cell-edge bottleneck in multicast streaming by treating neighboring base stations as "helpers." While the CGA algorithm provides a strong theoretical foundation, future work is needed to address the signaling overhead of coordinating multiple base stations and to develop distributed versions of the algorithm suitable for dynamic, real-world cellular environments.

Summary updated

BibTeX
@misc{zuhra2023multi,
  author = {Sadaf ul Zuhra and Prasanna Chaporkar and Abhay Karandikar and H. Vincent Poor},
  title = {{Multi-Connectivity for Multicast Video Streaming in Cellular Networks (Extended Abstract)}},
  year = {2023},
  eprint = {2305.15252},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2305.15252},
}
Figure