Abstract
The quickly increasing data traffic and the user demand for a full coverage of mobile services anywhere and anytime are leading mobile networking into a future of small cell networks. However, due to the high-density and randomness of small cell networks, there are several technical challenges. In this paper, we investigate two critical issues: best signal quality and mobility management. Under the assumptions that base stations are uniformly distributed in a ring shaped region and that shadowings are lognormal, independent and identically distributed, we prove that when the number of sites in the ring tends to infinity, then (i) the maximum signal strength received at the center of the ring tends in distribution to a Gumbel distribution when properly renormalized, and (ii) it is asymptotically independent of the interference. Using these properties, we derive the distribution of the best signal quality. Furthermore, an optimized random cell scanning scheme is proposed, based on the evaluation of the optimal number of sites to be scanned for maximizing the user data throughput.
Introduction
As cellular networks evolve toward dense "small cell" deployments (e.g., femtocells), traditional mobility management strategies are becoming increasingly inadequate. In conventional macro-cell networks, base stations are strategically placed, allowing the network to provide mobile devices with a predefined "neighbor cell list" for monitoring.
In contrast, small cell environments are characterized by high density and random, user-driven deployments. Maintaining manual neighbor lists in such chaotic environments is computationally prohibitive. This research addresses two fundamental challenges: statistically characterizing the "best" signal a user receives in a random network and developing a "random scanning" approach to replace rigid neighbor lists.
Methodology
The study utilizes Stochastic Geometry and Extreme Value Theory (EVT) to model the network. Base stations are assumed to be uniformly distributed within a ring-shaped region, with radio propagation modeled using path-loss and lognormal shadowing.
Instead of analyzing average signal strengths, the authors focus on the maximum signal strength ($M_n$) among $n$ sites. By applying the Fisher-Tippett theorem, they demonstrate that as the number of sites increases, the distribution of this maximum signal converges to a Gumbel distribution. A significant technical contribution is the proof that this maximum signal strength becomes asymptotically independent of total interference in dense networks.
The model's accuracy was validated through numerical simulations calibrated with industry-standard parameters (e.g., WCDMA/WiMAX standards, $\beta = 3$, $\sigma = 8$ dB). The Jensen-Shannon Divergence (JSDIV) was used to quantify the fit between analytical and empirical distributions across various parameters.
Key Results and Insights
The research yields several critical insights for small cell network design:
- Gumbel Convergence: The "best" signal in a random network follows a predictable Gumbel distribution. This allows for performance prediction without precise knowledge of every small cell's location.
- Interference Independence: The asymptotic independence between the strongest signal and total interference significantly simplifies SINR calculations in dense networks.
- Analytical Accuracy: Derived formulas for signal quality distribution show exceptional agreement with empirical simulations, even for a moderate number of cells ($n=50$).
- Optimal Scanning Strategy: The authors proposed an "acceleration" metric ($\rho_n$) to balance throughput gains against scanning latency. In a simulated WCDMA environment, scanning approximately 42 to 43 cells was found to be optimal. Notably, the relative benefit of scanning additional cells diminishes significantly after the first 30, suggesting a practical balance for real-world systems.
Discussion and Future Work
While the model provides a robust framework, it assumes a uniform distribution of base stations. Real-world deployments may exhibit clustering, potentially requiring non-homogeneous Poisson process models. Future work could also address uplink interference patterns and the impact of fast fading, which was averaged out in this study. The proposed framework is particularly relevant for the development of "self-organizing networks" (SON), where network parameters are automatically tuned based on these statistical properties.
Summary updated