Traffic management of a satellite communication network using stochastic optimization

Area of Science:

• Computer Science
• Telecommunications Engineering
• Network Optimization

Background:

• Nonhierarchical circuit-switched networks benefit from alternate routes at moderate loads.
• Alternate routes can cause instability and performance degradation in circuit-switched networks under heavy or overloaded conditions.
• Existing dynamic routing methods with fixed configurations may not adapt optimally to fluctuating network traffic.

Purpose of the Study:

• To develop and evaluate a novel traffic management scheme for mesh-connected, circuit-switched satellite communication networks.
• To enhance network performance and stability across diverse load conditions.
• To dynamically adapt network configuration based on real-time traffic measurements.

Main Methods:

• Implemented a traffic management scheme involving dynamic reconfiguration of network topology and link capacity allocation.
• Reserved a fraction of link capacity for direct routed calls to mitigate instability.
• Employed two neural network-based optimization techniques, simulated annealing and mean field annealing, for optimal capacity reservation.
• Conducted a comparative simulation study of the proposed scheme against pure dynamic routing with fixed configurations.

Main Results:

• The proposed traffic management scheme significantly improved network performance compared to pure dynamic routing with fixed configurations.
• Dynamic reconfiguration and optimal capacity reservation effectively managed network load and enhanced stability.
• Simulated annealing and mean field annealing proved effective in optimizing network resource allocation.

Conclusions:

• The developed traffic management scheme offers a superior approach for enhancing the performance of circuit-switched satellite networks.
• Dynamic adaptation of network configuration and intelligent capacity reservation are crucial for robust network operation under varying loads.
• Neural network-based optimization techniques provide a powerful tool for achieving optimal network resource allocation in complex communication systems.