TY - GEN
T1 - Optimizing Over-The-Air Computation in IRS-Aided C-RAN Systems
AU - Yu, Daesung
AU - Park, Seok Hwan
AU - Simeone, Osvaldo
AU - Shitz, Shlomo Shamai
N1 - Publisher Copyright:
© 2020 IEEE.
PY - 2020/5
Y1 - 2020/5
N2 - Over-The-Air computation (AirComp) is an efficient solution to enable federated learning on wireless channels. Air-Comp assumes that the wireless channels from different devices can be controlled, e.g., via transmitter-side phase compensation, in order to ensure coherent on-Air combining. Intelligent reflecting surfaces (IRSs) can provide an alternative, or additional, means of controlling channel propagation conditions. This work studies the advantages of deploying IRSs for AirComp systems in a large-scale cloud radio access network (C-RAN). In this system, worker devices upload locally updated models to a parameter server (PS) through distributed access points (APs) that communicate with the PS on finite-capacity fronthaul links. The problem of jointly optimizing the IRSs' reflecting phases and a linear detector at the PS is tackled with the goal of minimizing the mean squared error (MSE) of a parameter estimated at the PS. Numerical results validate the advantages of deploying IRSs with optimized phases for AirComp in C-RAN systems.
AB - Over-The-Air computation (AirComp) is an efficient solution to enable federated learning on wireless channels. Air-Comp assumes that the wireless channels from different devices can be controlled, e.g., via transmitter-side phase compensation, in order to ensure coherent on-Air combining. Intelligent reflecting surfaces (IRSs) can provide an alternative, or additional, means of controlling channel propagation conditions. This work studies the advantages of deploying IRSs for AirComp systems in a large-scale cloud radio access network (C-RAN). In this system, worker devices upload locally updated models to a parameter server (PS) through distributed access points (APs) that communicate with the PS on finite-capacity fronthaul links. The problem of jointly optimizing the IRSs' reflecting phases and a linear detector at the PS is tackled with the goal of minimizing the mean squared error (MSE) of a parameter estimated at the PS. Numerical results validate the advantages of deploying IRSs with optimized phases for AirComp in C-RAN systems.
KW - C-RAN
KW - Over-The-Air computation
KW - intelligent reflecting surface
UR - http://www.scopus.com/inward/record.url?scp=85090384956&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85090384956&partnerID=8YFLogxK
U2 - 10.1109/SPAWC48557.2020.9154243
DO - 10.1109/SPAWC48557.2020.9154243
M3 - Conference contribution
AN - SCOPUS:85090384956
T3 - IEEE Workshop on Signal Processing Advances in Wireless Communications, SPAWC
BT - 2020 IEEE 21st International Workshop on Signal Processing Advances in Wireless Communications, SPAWC 2020
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 21st IEEE International Workshop on Signal Processing Advances in Wireless Communications, SPAWC 2020
Y2 - 26 May 2020 through 29 May 2020
ER -