Aggregated Throughput Prediction for Collated Massive Machine-Type Communications in 5G Wireless Networks.

Area of Science:

• Wireless Communications
• Network Optimization
• Mobile Computing

Background:

• Increasing data demands in dense wireless networks necessitate advanced control schemes.
• Fifth-generation mobile communications (5G) support massive Machine Type Communications (mMTC), requiring optimized throughput.
• Carrier Sense Multiple Access (CSMA) transmission rates need optimization for mandatory Quality of Service (QoS) metrics.

Purpose of the Study:

• To optimize the carrier sense multiple access (CSMA) transmission attempt rate for enhanced wireless network throughput.
• To address the slow convergence issues of gradient descent algorithms in dense network environments.
• To present a practical and efficient method for achieving optimal throughput in 5G networks.

Main Methods:

• An approximation technique is introduced, partitioning dense global networks into less complex local neighborhoods.
• Newton's method of optimization is applied to achieve rapid convergence rates.
• Polynomial interpolation is utilized to estimate average network throughput based on node count and service rates.

Main Results:

• The proposed method demonstrates faster convergence rates compared to traditional gradient descent algorithms.
• Convergence rate is shown to depend on local network size, not global complexity.
• Three-dimensional planes effectively visualize network performance, illustrating average throughput as a function of nodes and service rates.

Conclusions:

• The developed approximation technique offers a more practical and efficient solution for optimizing wireless network throughput.
• The model's fast convergence and reduced complexity make it suitable for real-world 5G applications.
• The study provides a profound understanding of network performance through detailed throughput analysis.