LMMSE Processing for Cell-free Massive MIMO with Radio Stripes and MRC Fronthaul

• Published in: 2022 IEEE Wireless Communications and Networking Conference (WCNC) pp. 1336 - 1340
• Main Authors: Yuan, Zhifeng, Ma, Yihua, Yu, Guanghui
• Format: Conference Proceeding
• Language: English
• Published: IEEE 10-04-2022
• Subjects: Cell-free massive MIMO; Complexity theory; Conferences; Costs; Hardware; limited fronthaul; Loading; Massive MIMO; quasi-LMMSE; radio stripes; Simulation
• ISSN: 1558-2612
• DOI: 10.1109/WCNC51071.2022.9771691

Cell-free massive multi-input multi-output (MIMO) provides ubiquitous connectivity for multiple users. Compared with co-located massive MIMO, the major cost includes fronthaul and access points (AP). This paper assumes an implementation using radio stripes. Maximum ratio combining (MRC) uses a low-loading fronthaul and low-cost APs, but the performance is bad. To solve these problems, this paper first proposes an SVD-based method named quasi-LMMSE (Q-LMMSE). Q-LMMSE is derived from standard linear minimum- mean-square-error (LMMSE) using only the MRC frounthaul signals. Employing the simple MRC fronthaul design, a novel processing method is derived from standard linear minimum- mean-square-error (LMMSE). Simulation results show that the proposed Q-LMMSE gains a higher SE with low fronthaul loading, AP complexity, and latency than existing distributed schemes and performs even better than centralized LMMSE in some cases.