State Transition Analysis of Time-Frequency Resource Conversion-based Call Admission Control for LTE-Type Cellular Network

Hangguan Shan, Zhifeng Ni, Weihua Zhuang, Aiping Huang, Wei WangarXiv 2013 · arXiv

Abstract

This paper provides a rigorous mathematical framework for a novel LTE resource management strategy called Time-Frequency Resource Conversion (TFRC). By exploiting "context information"—specifically, whether a user is actively engaging with a specific data stream—the system can dynamically reallocate bandwidth to reduce congestion without dropping active calls. We analyze the state transitions of a cellular system employing a double-threshold guard channel policy, utilizing a multi-stair Markov model to derive steady-state probabilities and performance metrics essential for network optimization.

Introduction

As mobile devices become more advanced, they often maintain multiple simultaneous data connections, such as a video stream in the foreground while an application updates in the background. Standard LTE networks treat these connections with rigid resource requirements, which can lead to unnecessary network congestion. Call Admission Control (CAC) is the process of deciding when to accept a new or "handed-off" user into a cell. Traditional policies are often too conservative or too lenient. This study analyzes a "double-threshold guard channel policy" that uses TFRC to "squeeze" bandwidth from background tasks to make room for new high-priority connections, effectively increasing the network's carrying capacity.

Methodology

The core of this research is the State Transition Analysis of the system. To model how a cellular base station moves between different loads and user states, we identify five categories of trigger events:

  1. New Call Arrivals: Categorized by whether the user is currently idle or already has an active connection.
  2. Handoff User Arrivals: Users moving into the cell from a neighboring one.
  3. Handoff User Departures: Users leaving the cell.
  4. Call Terminations: Natural ending of a data session.
  5. Periodic TFRC: Scheduled moments when the system checks user context and reduces the bandwidth of "background" connections.

The researchers utilize a multi-stair Markov model to approximate the complex, mixed continuous-discrete process of user state changes. This allows for the calculation of steady-state probabilities, which are essential for predicting performance metrics like drop rates and blocking probability. The mathematical symbols and definitions used in this analysis are detailed in the table below.

Table I: Definitions of symbols used in the state transition analysis.

Key Insights and Theoretical Foundations

A significant mathematical insight involves the memoryless property of call holding times. It is proven that even if a connection's bandwidth is reduced and then later restored, its remaining "call holding time" still follows the original exponential distribution. This simplifies the mathematics significantly, allowing standard Markovian analysis to hold true even in a highly dynamic environment. Furthermore, the paper provides granular handoff logic, breaking down handoffs into six sub-scenarios based on the user's connection state and the level of TFRC already applied.

Results and Discussion

The analysis successfully derives the transition rates for every possible state change in the system. These rates serve as a foundation for engineers looking to implement this CAC policy. Key quantitative formulations include:

  • Handoff Rates: Derived as a function of the average user number ($\bar{K}_A$), the cell residual time ($\eta$), and the steady-state probability of a user occupying specific subchannels.
  • Service Rates under TFRC: The authors provide a formula for the adjusted service rate when a connection is partially withdrawn: $\mu_1(1 - r_I^{(i)}/r_1)$.

Limitations and Future Work

The model relies on the assumption that connection durations and cell residual times follow an exponential distribution. While standard in networking models, real-world traffic can be "bursty," potentially requiring more complex power-law models. Additionally, the analysis assumes a fixed periodic feedback interval ($\tau$) for context information; future research should explore the impact of feedback latency and signaling overhead on system performance in high-mobility scenarios.

Conclusion

Context-aware resource management through TFRC offers a promising path toward more efficient LTE networks. By distinguishing between foreground and background tasks, networks can significantly improve their carrying capacity without compromising user experience. This state transition analysis provides the theoretical groundwork necessary for deploying such dynamic admission control policies.

Summary updated

Abstract

To address network congestion stemmed from traffic generated by advanced user equipments, in [1] we propose a novel network resource allocation strategy, time-frequency resource conversion (TFRC), via exploiting user behavior, a specific kind of context information. Considering an LTE-type cellular network, a call admission control policy called double-threshold guard channel policy is proposed there to facilitate the implementation of TFRC. In this report, we present state transition analysis of this TFRC-based call admission control policy for an LTE-type cellular network. Overall, there are five categories of events that can trigger a transition of the system state: 1) a new call arrival; 2) a handoff user arrival; 3) a handoff user departure; 4) a call termination; and 5) a periodic time-frequency resource conversion. We analyze them case by case in this report and the validation of the analysis has been provided in [1].
BibTeX
@misc{shan2013state,
  author = {Hangguan Shan and Zhifeng Ni and Weihua Zhuang and Aiping Huang and Wei Wang},
  title = {{State Transition Analysis of Time-Frequency Resource Conversion-based Call Admission Control for LTE-Type Cellular Network}},
  year = {2013},
  eprint = {1312.0333},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1312.0333},
}
Figure