Abstract
Service-Oriented Architecture (SOA) is employed in a wide range of applications because it promises to expose computation-intensive tasks as services and combine them with new applications to accelerate their development process. For service-oriented multimedia applications, the performance of multicasting transmission services under multimedia traffic must be evaluated. Multiview Video (MVV) is a promising and emerging type of multimedia traffic that consists of multiple video streams, allows stream switching, and requires strict Quality-of-Service (QoS). In this study, we investigated multicast transmission of MVV in wireless cellular networks with Partial Frequency Reuse (PFR) and focused on two challenging problems: (1) multicast group formation and (2) subchannel and power allocation. Initially, we propose a novel Content-based Partial frequency reuse Diversity Grouping (CPDG) scheme to allocate users to multicast groups on the basis of Partial frequency reuse Diversity (PD) in wireless networks with PFR. After group formation, an optimization problem for subchannel and power allocation is formulated to minimize network power consumption under QoS constraint. Thereafter, the optimization problem is divided into two subproblems and solved using the simplex method and Karush-Kuhn-Tucker (KKT) conditions. Finally, to solve the optimization problem, a suboptimal scheme referred to as Partial frequency reuse Diversity Based Energy-efficient Multicasting (PDEM) is proposed by dynamically allocating wireless resources under QoS constraint. The simulation results show that the proposed PDEM scheme can achieve close-to-optimal performance.
Introduction
As cellular networks transition toward Ultra-High Definition (UHD) and Augmented Reality (AR), Multiview Video (MVV) has emerged as a high-bandwidth challenge. MVV consists of multiple synchronized video streams from different camera angles, allowing users to switch views dynamically. This flexibility requires strict Quality-of-Service (QoS) and high data rates. In traditional cellular setups, Frequency Reuse 1 often leads to significant interference for users at the cell edge. Partial Frequency Reuse (PFR) mitigates this by dividing the cell into center and edge regions with distinct frequency bands. However, multicasting MVV in a PFR environment introduces complexities in grouping users across different regions and allocating resources efficiently to minimize power consumption.
Methodology
Traffic Modeling and Stream Popularity
The study utilizes a traffic model where MVV consists of 8 camera streams. To simulate realistic demand, a Zipf distribution is employed to model stream popularity. This distribution reflects how a small number of views typically attract the majority of users, while others are requested less frequently.
Content-based Grouping (CPDG)
The researchers proposed the Content-based Partial Frequency Reuse Diversity Grouping (CPDG) algorithm. This scheme groups users based on their requested video stream and their physical location (Cell Center vs. Cell Edge). A core innovation is the exploitation of Partial Frequency Reuse Diversity (PD): since cell-center users can utilize both center and edge frequency bands, they provide a flexible resource pool that can be balanced to optimize overall network efficiency.
Resource Allocation and the PDEM Algorithm
To minimize total network power consumption while meeting QoS requirements, the authors formulated an optimization problem. Given that the original problem is NP-hard, it was decomposed into two sub-problems: subchannel allocation (solved via the Simplex method) and power allocation (solved via KKT conditions). For real-time deployment, the Partial Frequency Reuse Diversity Based Energy-efficient Multicasting (PDEM) scheme was developed as a fast, suboptimal heuristic that dynamically manages resources.
Key Results and Analysis
Simulation results indicate that the PDEM scheme achieves power consumption levels nearly identical to the theoretical optimal solution. Furthermore, the grouping strategy effectively manages the number of multicast groups as the number of available camera streams increases.
Key findings include:
- Energy Efficiency (EE): EE improves as the user density increases, as more users share the same multicast resources.
- Throughput Gains: The proposed scheme maintains higher system throughput compared to models that ignore the integer constraints of subchannel allocation.
- Transmit Power Limits: Transmit power eventually reaches a ceiling ($P_{max}$), at which point throughput per group begins to decline, providing a critical operational guideline for network providers.
Conclusion
By leveraging Partial Frequency Reuse Diversity, the proposed CPDG and PDEM frameworks offer a robust solution for energy-efficient MVV multicasting. The approach balances the flexibility of cell-center users with the strict requirements of high-definition video, achieving near-optimal performance with manageable computational complexity. Future research may extend this model to high-speed mobility scenarios and ultra-dense 5G/6G network architectures.
Summary updated