Autor: |
Jeyakumar, P., Vishwa, S., Prakash, V., Dinesh, P., Rajesh Ram, P., Muthuchidambaranathan, P. |
Předmět: |
|
Zdroj: |
Wireless Networks (10220038); Jan2024, Vol. 30 Issue 1, p67-76, 10p |
Abstrakt: |
Massive multiple-input multiple-output (MIMO) systems at millimeter wave have been touted as a possible technology for 5 G wireless communications. The massive MIMO systems, gathering channel information using a beam tracking approach is crucial for tracking user movement. Conventional beam tracking algorithms for narrowband systems using the classic hybrid precoding framework suffers from the beam split effect and hence cannot be effectively adapted to wideband massive MIMO systems. To address this problem, we describe a long-short term memory (LSTM) network for detecting user movement and calibrating beam direction based on received signals from earlier beam training to improve noise resilience. Simulation findings demonstrate that the proposed approach may reach near-optimal achievable sum-rate performance with little beam training overhead. [ABSTRACT FROM AUTHOR] |
Databáze: |
Complementary Index |
Externí odkaz: |
|