Reinforcement Learning for TDMA UAV Communication Throughput
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Establishing an unmanned vehicle system with high information transfer rates is challenging due to slot overheads in TDMA networks, which can degrade throughput, and conventional solutions like increasing bandwidth or power may be impractical, hindering size reduction and interference.
Innovation Solution
Applying reinforcement learning to design wireless communication systems for unmanned vehicles, optimizing information transfer rates through a learning model that adjusts transmission rates to meet performance requirements, allowing cooperation among multiple vehicles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If slots are reserved for each unmanned vehicle in TDMA network, then collision of radio waves is prevented, but overhead increases and throughput degrades
Solution Approach 1:
The patent applies dynamic slot allocation where the number of reserved slots is not fixed but adapts based on the number of unmanned vehicles in the system. The master unmanned vehicle dynamically adjusts the slot configuration in response to changing system conditions, allowing the network to maintain reliable collision-free communication while minimizing overhead when fewer vehicles are present.
Solution Approach 2:
The system changes the parameter of slot allocation from a static per-vehicle basis to a dynamic basis that depends on the total number of vehicles. This parameter change allows the network to optimize throughput by reducing the number of reserved slots when fewer vehicles are operating, while still preventing collisions through the TDMA structure.
2Productivity
If wireless bandwidth is increased to improve throughput, then information transfer rate increases, but interference with other stations may occur
Solution Approach 1:
The patent segments the wireless communication channel into time slots through TDMA, allowing multiple unmanned vehicles to share the same frequency band without interference. This time-based segmentation enables efficient use of available bandwidth while preventing interference with other stations operating on different time slots or frequency channels.
3Productivity
If transmission power is increased to improve throughput, then information transfer rate increases, but weight of radio communication device increases
Solution Approach 1:
The patent changes the operational parameters of the communication system by implementing adaptive transmission rates and dynamic slot allocation. This allows the system to achieve high throughput through efficient resource management rather than relying on high transmission power, thereby avoiding the need for heavier radio communication devices.
4Adaptability or versatility
If number of unmanned vehicles increases to tens or hundreds, then system capability increases, but slot overhead degrades throughput
Solution Approach 1:
The patent implements dynamic slot allocation where the master unmanned vehicle adjusts the number and configuration of time slots based on the actual number of vehicles in the system. This dynamic adaptation allows the network to support a large number of vehicles (tens or hundreds) while minimizing overhead by allocating slots only to active vehicles, thereby maintaining high throughput despite system expansion.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A learning device includes a setting unit configured to set a first value for a parameter of a communication device controlled by a computer using a learned model; a reinforcement learning unit configured to allow a learning model to learn; a model extraction unit configured to extract, as a learned model, the learning model; a model evaluation unit configured to determine whether performance of the learned model has reached first requirement; an updating unit configured to update the first value to a second value when the performance is determined to have reached the first requirement; and a model selection unit. The model evaluation unit determines whether the performance of the learned model updated to the second value satisfies second requirement. When the performance of the learned model updated to the second value is determined to satisfy the second requirement, the model selection unit selects that learned model.