Related Experiment Videos
A unified machine learning framework for intelligent resource allocation toward 6G wireless communications
Nishu Gupta1, Rupali Bhartiya2, Seema Babusing Rathod3
1Department of Information Technology, Indian Institute of Information Technology , Lucknow, India. nishugupta@ieee.org.
Scientific Reports
|July 13, 2026
Summary
This study introduces an AI and Deep Learning approach for efficient resource allocation in 6G wireless networks, improving energy efficiency and reducing bit error rates for better performance.
Area of Science:
- Wireless Communication
- Artificial Intelligence
- Deep Learning
Background:
- 6G wireless networks face challenges in resource allocation due to demands for low latency, high speed, and energy efficiency.
- Traditional methods are insufficient for dynamic network conditions and service requirements.
- Artificial Intelligence (AI) and Deep Learning (DL) offer promising solutions for intelligent decision-making in complex network scenarios.
Purpose of the Study:
- To propose an integrated AI and DL approach for efficient and intelligent resource allocation in 6G wireless networks.
- To address challenges in dynamic spectrum allocation, energy depletion, and signal attenuation.
- To optimize communication routes, power, and spectrum allocation.
Main Methods:
- An integrated AI and DL framework combining optimal path selection and efficient allocation mechanisms.
- Utilized input parameters: residual battery indicator (RBI), channel matrix (H), normalized spectrum availability (v), SINR, node pairs (s, d), service levels, and historical statistics.
- Employed a Recursive Hampel Filter-Based Estimation Model (ReHF-EM) for data quality and a Dual-Stage Multi-Time-Scale Temporal Attention-Based LSTM network (D-MTSTA-LSTM) for decision-making.
- Optimized model parameters using the Pied Kingfisher Optimizer (PKfO).
Main Results:
- Achieved a 23.6% increase in Energy Efficiency (EE).
- Demonstrated a 19.2% reduction in Bit Error Rate (BER).
- Evaluated performance based on Spectrum Efficiency (SE), SINR margin, Computational Time (CT), and Accuracy.
Conclusions:
- The proposed integrated AI and DL approach effectively enhances resource allocation in 6G networks.
- The D-MTSTA-LSTM network accurately predicts optimal routes and resource allocation, learning short- and long-term network trends.
- The PKfO fine-tuning improves model efficiency and reduces complexity, leading to significant gains in EE and reductions in BER.