Dynamic Channel Allocation Control with thresholds in Wireless Cellular Networks using Simpy

Yang Cao, Cheul-Woo RoInternational Journal of Contents 2012 · DOI

Abstract

New and handoff calls control mechanisms are the key point to wireless cellular networks. In this paper, we present an adaptive algorithm for dynamic channel allocation scheme with guard channels and also with handoff calls waiting queue ensuring that handoff calls take priority over new calls. Our goal is to find better tradeoff between handoffs and new calls blocking probabilities in order to achieve more efficient channel utilization. Simpy is a Python based discrete event simulation system. We use Simpy to build our simulation models to get analytical data.

Introduction

In wireless cellular networks, service providers face a fundamental tension between two types of traffic: New Calls (NC) and Handoff Calls (HC). A handoff occurs when an active user moves from one cell's coverage area to another. From a user experience perspective, dropping an ongoing call (handoff failure) is significantly more disruptive than having a new call attempt blocked.

To prioritize handoffs, networks traditionally employ "guard channels"—a subset of available channels reserved exclusively for users transitioning into a cell. However, static guard channel policies often lead to inefficiencies. Reserving too many channels wastes resources and unnecessarily blocks new users, while reserving too few risks dropping active calls during peak traffic periods. This research proposes an adaptive mechanism that dynamically adjusts these reservations in real-time based on fluctuating traffic loads.

Methodology

The core of this study is the development of Policy 3: Adaptive Guard Channels with Thresholds. This approach integrates two primary prioritization strategies:

  1. Dynamic Guard Channels ($g$): The system continuously monitors the Handoff Dropping Probability ($P_{dh}$). If $P_{dh}$ falls below a minimum threshold ($T_{min}$), the system reduces the number of reserved channels to accommodate more new calls. Conversely, if $P_{dh}$ exceeds a maximum threshold ($T_{max}$), the number of reserved channels is increased to protect existing calls.
  2. Handoff Waiting Queue: When all channels are occupied, handoff calls are not immediately terminated. Instead, they enter a First-In-First-Out (FIFO) queue, allowing them to secure a channel as soon as one becomes available.

To evaluate this policy, the authors utilized Simpy, a Python-based discrete-event simulation framework. This allowed for the modeling of complex, parallel processes—such as thousands of independent call arrivals and departures—to collect robust analytical data on network performance.

Simulation Parameters

The simulation utilized synthetic traffic generated according to specific statistical distributions:

  • Call Arrivals: Modeled using an Exponential distribution, standard for random, independent events.
  • Call Duration: Modeled using a Pareto distribution (heavy-tailed), which accurately reflects real-world traffic where most calls are short, but a significant minority last a long time.
  • Network Configuration: The model assumed 50 channels per cell, a handoff queue size of 3, and a 5:1 ratio of new calls to handoff calls.

Key Results and Discussion

The adaptive policy (Policy 3) was compared against no-guard-channel (Policy 1) and various fixed-guard-channel policies (Policy 2.x).

Figure 3: Comparison of Handoff Dropping Probability (Pdh) across different policies.

As shown in Figure 3, Policy 3 (represented by the pink square markers) successfully maintains a low and stable handoff dropping probability ($P_{dh}$) even as the new call arrival rate increases. In contrast, Policy 1 (the baseline with no guard channels) shows a sharp exponential increase in dropped calls as traffic rises.

To quantify the overall performance, the authors used a Grade of Service (GoS) metric: $GoS = P_{dh} + W \cdot P_{bn}$, where $P_{bn}$ is the new call blocking probability and $W$ is a weighting factor (set to 0.1). This formula mathematically prioritizes preventing dropped calls over preventing blocked new calls.

Figure 2: Grade of Service (GoS) comparison, showing the superior performance of the adaptive Policy 3.

Figure 2 demonstrates that Policy 3 (dark blue diamonds) achieves the best overall Grade of Service. While fixed policies with high guard channel counts could achieve low dropping rates, they did so at the expense of excessively high new call blocking. The adaptive policy provides a "Goldilocks" solution, maintaining low dropping rates while keeping the new call blocking rate significantly lower than high-fixed-reservation models.

Conclusion

The study confirms that an adaptive guard channel algorithm with thresholds can effectively "self-correct," increasing protection for handoffs exactly when traffic spikes threaten call stability. This dynamic approach maximizes network utility without sacrificing user experience. Future work may extend this single-cell model to multi-tier network architectures and more granular radio resource utilization analysis.

Summary updated

Abstract

New and handoff calls control mechanisms are the key point to wireless cellular networks. In this paper, we present an adaptive algorithm for dynamic channel allocation scheme with guard channels and also with handoff calls waiting queue ensuring that handoff calls take priority over new calls. Our goal is to find better tradeoff between handoffs and new calls blocking probabilities in order to achieve more efficient channel utilization. Simpy is a Python based discrete event simulation system. We use Simpy to build our simulation models to get analytical data.
BibTeX
@article{cao2012dynamic,
  author = {Yang Cao and Cheul-Woo Ro},
  title = {{Dynamic Channel Allocation Control with thresholds in Wireless Cellular Networks using Simpy}},
  journal = {International Journal of Contents},
  year = {2012},
  doi = {10.5392/ijoc.2012.8.2.019},
}
Figure