[Paper Review] An Iterative Interference Cancellation Algorithm for Large Intelligent Surfaces

Jesús Rodríguez Sánchez, Fredrik Rusek+2 more|arXiv (Cornell University)|Nov 25, 2019

Advanced Wireless Communication Technologies6 references10 citations

TL;DR

This paper proposes an iterative interference cancellation (IIC) algorithm for Large Intelligent Surfaces (LIS) that enables decentralized uplink detection via panel-based processing, significantly outperforming a reduced matched filter (RMF) baseline. The algorithm maximizes sum-rate capacity through localized equalizer design using singular value decomposition and panel-to-panel data exchange, achieving higher spectral efficiency with fewer outputs per panel compared to conventional methods.

ABSTRACT

The Large Intelligent Surface (LIS) concept is a promising technology aiming to revolutionize wireless communication by exploiting spatial multiplexing at its fullest. Despite of its potential, due to the size of the LIS and the large number of antenna elements involved there is a need of decentralized architectures together with distributed algorithms which can reduce the inter-connection data-rate and computational requirement in the Central Processing Unit (CPU). In this article we address the uplink detection problem in the LIS system and propose a decentralize architecture based on panels, which perform local linear processing. We also provide the sum-rate capacity for such architecture and derive an algorithm to obtain the equalizer, which aims to maximize the sum-rate capacity. A performance analysis is also presented, including a comparison to a naive approach based on a reduced form of the matched filter (MF) method. The results shows the superiority of the proposed algorithm.

Motivation & Objective

• To address the high computational and backhaul load of centralized processing in Large Intelligent Surfaces (LIS) with massive antenna arrays.
• To develop a decentralized architecture based on panels that reduces inter-connection data rates and enables pipelined equalizer computation.
• To maximize sum-rate capacity in the uplink by designing an optimal equalizer using localized processing and iterative interference cancellation.
• To provide performance trade-offs between panel size, number of outputs per panel, and total system performance.

Proposed method

• The LIS is partitioned into P square panels, each with M_p antenna elements and N_p outputs, connected via a daisy-chained network for inter-panel data exchange.
• Each panel performs local linear processing using a precoded channel matrix and interference cancellation based on the previous panel's effective interference matrix.
• The equalizer is derived by maximizing the sum-rate capacity expression, using singular value decomposition of the effective channel matrix scaled by interference covariance.
• The algorithm iteratively updates the interference cancellation matrix using eigen-decomposition of the accumulated interference, ensuring monotonic sum-rate increase.
• A backplane aggregates processed outputs from all panels and forwards them to the CPU for final user symbol estimation.
• The method avoids full channel state information at the CPU, relying instead on local processing and limited inter-panel feedback.

Experimental results

Research questions

• RQ1 How can uplink detection in Large Intelligent Surfaces be decentralized to reduce computational load and backhaul requirements at the central processor?
• RQ2 What is the optimal equalizer design that maximizes sum-rate capacity under a decentralized panel-based architecture?
• RQ3 How does the number of outputs per panel (N_p) and panel size affect the achievable sum-rate capacity and system performance?
• RQ4 What performance gain does the proposed iterative interference cancellation (IIC) algorithm provide over a reduced matched filter (RMF) baseline?
• RQ5 What trade-offs exist between panel size, number of panels, and total number of outputs for achieving target spectral efficiency?

Key findings

• The IIC algorithm achieves higher sum-rate capacity than the reduced matched filter (RMF) baseline for all tested values of N_p and both panel sizes.
• The small-panel configuration (20cm×20cm) converges faster to the channel capacity limit, requiring fewer outputs per panel for the same performance target.
• For the same total number of outputs (N), the large-panel configuration (1m×1m) achieves better sum-rate performance than the small-panel configuration.
• The IIC algorithm monotonically increases sum-rate capacity at each iteration, proven via eigen-decomposition of the interference covariance matrix.
• The performance gain of IIC over RMF allows for a reduction in the number of outputs per panel by up to 50% for a given target sum-rate, depending on configuration.
• The system design enables pipelining of equalizer computations across physical resource blocks due to the daisy-chain inter-panel connection, supporting real-time operation.