Congmin Fan

Dynamic Nested Clustering for Parallel PHY-Layer Processing in Cloud-RANs

Featured by centralized processing and cloud based infrastructure, Cloud Radio Access Network (C-RAN) is a promising solution to achieving an unprecedented system capacity in future wireless cellular networks. The huge capacity gain mainly comes from the centralized and coordinated signal processing at the cloud server. However, full-scale coordination in a large-scale C-RAN requires the processing of very large channel matrices, leading to high computational complexity and channel estimation overhead. To tackle this challenge, we establish a unified theoretical framework for dynamic clustering by exploiting the near-sparsity of large C-RAN channel matrices. Based on this framework, we propose a dynamic nested clustering (DNC) algorithm that greatly improves the system scalability in terms of baseband-processing and channel-estimation complexity. With the proposed DNC algorithm, we show that the computational complexity (i.e., the computation time with serial processing) for the optimal linear detector is significantly reduced from O(N3) to O(N2), where N is the number of remote radio heads (RRHs) in the C-RAN. Moreover, the proposed DNC algorithm is also amenable to parallel processing, which further reduces the computation time to O(N 42/23 ).

Scalable coordinated uplink processing in cloud radio access networks

Featured by centralized processing and cloud based infrastructure, Cloud Radio Access Network (C-RAN) is a promising solution to achieve an unprecedented system capacity in future wireless cellular networks. The huge capacity gain mainly comes from the centralized and coordinated signal processing at the cloud server. However, full-scale coordination in a large-scale C-RAN requires the processing of very large channel matrices, leading to high computational complexity and channel estimation overhead. To resolve this challenge, we show in this paper that the channel matrices can be greatly sparsified without substantially compromising the system capacity. Through rigorous analysis, we derive a simple threshold-based channel matrix sparsification approach. Based on this approach, for reasonably large networks, the non-zero entries in the channel matrix can be reduced to a very low percentage (say 0.13% ~ 2%) by compromising only 5% of SINR. This means each RRH only needs to obtain the CSI of a small number of closest users, resulting in a significant reduction in the channel estimation overhead. On the other hand, the high sparsity of the channel matrix allows us to design detection algorithms that are scalable in the sense that the average computational complexity per user does not grow with the network size.

Randomized Gaussian message passing for scalable uplink signal processing in C-RANs

In Cloud Radio Access Networks (C-RANs), the high computational complexity of signal processing becomes unaffordable due to the large number of remote radio heads (RRHs) and users. This paper proposes a randomized Gaussian message-passing (RGMP) algorithm to reduce the complexity of uplink signal processing in C-RANs. Specifically, we first propose to use Gaussian message passing to reduce the computational complexity. In C-RANs, RRHs only need to detect signals from nearby users as the signals from distant users are very weak and can be ignored. Thus, in message-passing algorithms, messages only need to be exchanged among nearby RRHs and users. This leads to a linear computational complexity with the number of RRHs and users. Then, to improve the convergence of message passing, we propose to exchange messages in a random order instead of exchanging them simultaneously or in a fixed order. Numerical results show that the proposed RGMP algorithm has better convergence performance than conventional message passing. The randomness of the message update schedule also simplifies the analysis, which allows us to derive some convergence conditions for the RGMP algorithm. Besides analysis, we also compare the convergence rate of RGMP with existing low-complexity algorithms through extensive simulations.

Scalable Uplink Signal Detection in C-RANs via Randomized Gaussian Message Passing

Cloud radio access network (C-RAN) is a promising architecture for unprecedented capacity enhancement in next-generation wireless networks thanks to the centralization and virtualization of base station processing. However, centralized signal processing in C-RANs involves high computational complexity that quickly becomes unaffordable when the network grows to a huge size. First, this paper endeavors to design a scalable uplink signal detection algorithm, in the sense that both the complexity per unit network area and the total computation time remain constant when the network size grows. To this end, we formulate the signal detection in C-RAN as an inference problem over a bipartite random geometric graph. By passing messages among neighboring nodes, message passing (a.k.a. belief propagation) provides an efficient way to solve the inference problem over a sparse graph. However, the traditional message-passing algorithm is not guaranteed to converge, because the corresponding bipartite random geometric graph is locally dense and contains many short loops. As a major contribution of this paper, we propose a randomized Gaussian message passing (RGMP) algorithm to improve the convergence. Instead of exchanging messages simultaneously or in a fixed order, we propose to exchange messages asynchronously in a random order. The proposed RGMP algorithm demonstrates significantly better convergence performance than conventional message passing. The randomness of the message updating schedule also simplifies the analysis, and allows the derivation of the convergence conditions for the RGMP algorithm. In addition, we generalize the RGMP algorithm to a blockwise RGMP (B-RGMP) algorithm, which allows parallel implementation. The average computation time of B-RGMP remains constant when the network size increases.

Throughput Bounds for Training-Based Multiuser MIMO Systems

In this paper, we endeavour to maximize the throughput of training-based multiuser multiple-input multiple-output (MIMO) systems. In a multiuser MIMO system, users are geographically separated. So, the near-far effect plays an indispensable role in channel fading. The existing optimal training design for conventional MIMO does not take the near-far effect into account, and thus is not applicable to a multiuser MIMO system. In this work, we use the majorization theory as a basic tool to study the tradeoff between the channel estimation quality and the information throughput. We establish tight upper and lower bounds of the throughput. Due to the near-far effect, the optimal training design for throughput maximization is to deactivate a portion of users with the weakest channels in transmission. This observation shed light on the practical design of training-based multiuser MIMO systems.

Scalable Uplink Processing via Sparse Message Passing in C-RAN

Cloud radio access network (C-RAN) emerges as a promising solution to sustain the mobile data explosion with low cost and high energy efficiency. The centralized base band processing of C-RAN facilitates coordinated signal processing at the cloud server, which can potentially lead to huge capacity gain. However, full-scale coordination in a large-scale system inevitably results in high computational complexity that limits the scalability of the system. To address this issue, this paper proposes a scalable uplink signal processing algorithm based on message passing. By exploiting near-sparsity of large C- RAN channel matrices, we derive a sparse message- passing algorithm that reduces the computational complexity of signal detection to be linear with the number of RRHs and users. This implies that the average computational complexity per user does not grow with the network size, and hence the system is scalable. In addition, we discuss the convergence of the sparse message-passing algorithm and propose a block-wise message-passing algorithm that significantly improves the probability of convergence.