TL;DR
This paper introduces a novel Call Admission Control (CAC) policy for wireless cellular networks that uses an "acceptance factor" to dynamically manage new call arrivals. This approach aims to reduce new call blocking probability without significantly increasing handoff call dropping probability, thereby improving overall Quality of Service (QoS).
What Problem Does This Paper Address?
Wireless cellular networks face a constant challenge in balancing resource utilization with maintaining a high Quality of Service (QoS). A key aspect of QoS is minimizing both call blocking (when a new call cannot be established) and call dropping (when an ongoing call is terminated prematurely, often due to handoffs). As the number of users and calls increases, call blocking probability tends to rise, impacting user experience. Existing Call Admission Control (CAC) schemes, particularly those prioritizing handoff calls, often lead to increased blocking of new calls. This paper addresses the need for a CAC policy that can effectively reduce new call blocking while keeping handoff call dropping rates stable, especially under varying traffic loads.
Methods
The authors propose a new guard-band CAC scheme that builds upon the existing "new-call bounding scheme." The core innovation is the introduction of an "acceptance factor" ($\alpha$) applied to a specific guard band of channels. This factor dynamically controls the probability of accepting new calls within this band.
Here's a breakdown of the approach:
- Guard Bands: The total channels (C) in a cell are divided into three conceptual regions:
- 0 to M channels: Shared by both handoff and new calls with the same acceptance probability.
- M to N channels: A special "acceptance band" where new calls are accepted with a probability determined by the acceptance factor ($\alpha$), while handoff calls are accepted with probability 1.
- N to C channels: Reserved exclusively for handoff calls (acceptance probability of 0 for new calls).
- Acceptance Factor ($\alpha$): This parameter, ranging from 0 to 1, allows for flexible control.
- If $\alpha = 0$, the scheme reverts to the conventional new-call bounding scheme.
- If $\alpha = 1$, it behaves like a non-priority scheme.
- Analytical Modeling: The performance of the proposed scheme is analyzed using a one-dimensional Markov chain, which is a standard technique for modeling such systems. This allows for the derivation of steady-state probabilities, blocking probability ($P_B$) for new calls, and dropping probability ($P_D$) for handoff calls.
Datasets
Not discussed in the paper. The performance evaluation is based on analytical modeling and simulation.
Key Insights and Ideas
- Dynamic Control with Acceptance Factor: The introduction of an acceptance factor provides a flexible knob to tune the trade-off between new call blocking and handoff call dropping. This is a significant improvement over fixed-priority schemes.
- Optimal Acceptance Factor Varies with Traffic: The paper highlights that the optimal value of the acceptance factor is not static. For lower traffic loads, an $\alpha$ of 0.9 yields the lowest blocking probability, while for higher traffic loads, $\alpha = 0.5$ is more effective. This dynamic adaptation is crucial for real-world scenarios.
- Near-Constant Handoff Dropping: A key finding is that the proposed scheme achieves a significantly reduced new call blocking probability while maintaining a call dropping probability that is "considerably constant" and comparable to the conventional new-call bounding scheme.
- Generalization of Existing Schemes: The proposed scheme is presented as a generalization of the new-call bounding scheme, offering a more nuanced approach to CAC.
Key Results
The paper demonstrates through analytical modeling and graphical comparisons that the proposed guard-band CAC scheme with an acceptance factor can achieve:
- Reduced Call Blocking Probability: For both lower and higher traffic arrival rates, the proposed scheme generally shows lower call blocking probabilities compared to the conventional new-call bounding scheme.
- Stable Call Dropping Probability: The call dropping probability for handoff calls remains largely unaffected and comparable to the conventional scheme, even when new call blocking is reduced.
- Performance Tuning: Figures 4 and 5 illustrate how different acceptance factors (0.2, 0.5, 0.9) impact blocking and dropping probabilities across various call arrival rates. For instance, at higher traffic rates, $\alpha=0.5$ shows the minimum blocking probability, while at lower rates, $\alpha=0.9$ is more beneficial.
Limitations and Future Work
- Limitations:
- The analysis is based on a one-dimensional Markov chain, which might simplify certain real-world complexities.
- The paper focuses on a single class of traffic.
- Future Work:
- The authors plan to extend their research to analyze CAC policies using a two-dimensional Markov chain for multiclass traffic. This would allow for a more sophisticated modeling of diverse user demands and service types.
Summary updated