Abstract
Nowadays, radio access networks (RANs) are moving toward a small cell-based paradigm, where macro-cell antennas are aided by additional ones with lower coverage capabilities. This paradigm shift introduces a significant challenge: the need for a backhaul link to carry traffic from small cell base stations to the network gateway. While wired and wireless solutions exist, Power Line Communications (PLC) offers a promising broadband access alternative. This paper explores the feasibility of using existing electrical grids as a high-speed backhaul solution. By combining physical channel modeling with realistic traffic simulations, the research demonstrates how PLC technology can be assessed for this application, guiding the reader through small cell networks, power line infrastructure, and traffic generation modeling.
The Backhaul Challenge in Small Cell Networks
As mobile data traffic continues to grow exponentially, telecommunications providers are shifting from a few large macro-cells to a dense "small cell" paradigm involving micro, pico, and femtocells. These smaller base stations improve local capacity and battery life but introduce a massive logistical headache: backhaul. Every small cell requires a high-speed connection to the network gateway. Laying new fiber-optic cables to every building is prohibitively expensive, and wireless backhaul often suffers from interference or line-of-sight issues. This research investigates leveraging the ubiquitous copper wires already delivering power to buildings as a cost-effective, high-performance "last mile" link.
Methodology: Modeling the PLC Channel
The study employs a multi-layered simulation approach to evaluate PLC's potential as a backhaul medium.
Stochastic Geometry and Network Topology
To model the network realistically, the authors use stochastic geometry to randomly deploy base stations across a territory, accounting for realistic overlaps and densities typical of urban environments.
Bottom-Up Channel Modeling
Instead of simple statistical approximations, the authors apply transmission line theory to model the physical behavior of power cables. This accounts for signal attenuation and reflections caused by network topology and specific cable properties, such as the European-standard NAYY150SE and NAYY50SE cables. This approach captures the impact of medium discontinuities on signal propagation.
Traffic and Resource Allocation
The simulations incorporate a realistic traffic mix—3% voice (Poisson distribution) and 97% data (self-similar Pareto and Weibull distributions) to capture the bursty nature of internet usage. Furthermore, an optimized Time Division Multiple Access (TDMA) scheme is proposed to maximize aggregated throughput while ensuring a minimum Grade of Service (GoS) for all users.
Key Insights and Results
- Infrastructure as a "Free" Medium: PLC effectively eliminates the deployment time and cost of new backhaul infrastructure by utilizing existing residential wiring.
- ACG-Capacity Correlation: A strong correlation was found between Average Channel Gain (ACG) and Shannon link capacity. This suggests that network nodes can infer available capacity by measuring noise intensity, reducing computational overhead for real-time scheduling.
- Viability and Performance: Simulations confirm that PLC can achieve high-speed data transfers exceeding 100 Mbps per link in the 1–30 MHz range. The optimized resource allocation significantly outperformed standard TDMA, maintaining a high GoS even as user density increased.
Limitations and Future Work
While promising, the model is currently based on the European radial power grid paradigm; results may vary in North American or Asian grids. Additionally, the study assumes Gaussian noise, whereas real-world power lines face impulsive noise from appliances. Future research will explore higher frequency ranges (up to 300 MHz) and conduct field trials to validate these theoretical models in outdoor PLC environments.
Summary updated