Adaptive Modulation for Federated Learning Parity Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current federated learning methods in wireless communication systems face challenges in efficiently scheduling user equipment (UEs) participation and transmitting parity parts, leading to suboptimal resource allocation and increased latency as the number of participating UEs increases.
Innovation Solution
The method involves UEs receiving a channel state information reference signal, calculating and transmitting channel state information, and receiving scheduling information based on a reference channel state to determine participation. UEs encode and modulate local parameters, with the parity part being modulated based on the number of retransmissions and maximum UEs participating, and transmit these parameters with controlled power, optimizing resource allocation and participation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If more UEs participate in federated learning, then learning accuracy is improved, but resource allocation efficiency deteriorates and latency increases
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the modulation order for parity bits based on the number of participating UEs. When more UEs participate, the system changes the modulation parameter (using lower-order modulation like QPSK instead of higher-order QAM) to ensure reliable transmission of parity information, thereby maintaining learning accuracy while managing resource allocation efficiency
Solution Approach 2:
The system implements dynamics by making the parity bit modulation adaptive to the number of participating UEs. The modulation order is not fixed but dynamically selected based on system conditions, allowing the system to optimize between reliability and productivity as UE participation varies
2Reliability
If more UEs participate in federated learning, then learning accuracy is improved, but latency increases
Solution Approach 1:
The patent changes the modulation parameter for parity bits based on the number of participating UEs. By using appropriate modulation orders (e.g., QPSK for larger numbers of UEs), the system ensures that parity information is transmitted reliably within acceptable time frames, preventing retransmissions that would increase latency while maintaining learning accuracy
3Reliability
If transmission power is increased for parity part, then reliability of parity transmission is improved, but power consumption increases
Solution Approach 1:
Instead of uniformly increasing transmission power, the patent changes the modulation order parameter for parity bits. By using more robust lower-order modulation schemes when needed, the system achieves reliable parity transmission without necessarily increasing power consumption, as the reliability improvement comes from the modulation scheme rather than power amplification
Data Source
AI summary
The present specification provides a method by which a terminal performs federated learning with a plurality of terminals in a wireless communication system. More specifically, the method performed by one terminal comprises the steps of: receiving, from a server, a channel state information reference signal (CSI-RS); transmitting, to the server, channel state information (CSI) calculated on the basis of the CSI-RS; receiving, from the server, (i) information about a global parameter for the federated learning and (ii) compression state information for determining a weight compression method of the one terminal on the basis of channel state information of each of channels between the server and the plurality of terminals; determining a weight compression scheme based on (i) a difference value between the global parameter and a global parameter received before receiving the global parameter and (ii) the compression state information; and transmitting, to the server, an updated local parameter on the basis of the determined weight compression scheme.