Cross-Layer Design for Downlink Multihop Cloud Radio Access Networks With Network Coding

Liang Liu, Wei Yu

Abstract

Cloud Radio Access Network (C-RAN) architectures typically utilize either data-sharing or compression-based strategies for downlink communication. While compression-based strategies are often considered superior in simple star topologies, this paper demonstrates that a data-sharing strategy combined with network coding is more effective in multi-hop fronthaul networks. By jointly optimizing beamforming, user-RRH association, and network coding, the proposed approach leverages the inherent efficiency of information multicasting over complex network graphs.

Introduction

In a C-RAN, the baseband processing is centralized at a Central Processor (CP), while Remote Radio Heads (RRHs) perform basic radio frequency functions. The critical bottleneck in this architecture is the fronthaul—the network connecting the CP to the RRHs. Traditionally, research has focused on single-hop "star" topologies. However, practical deployments often involve multi-hop networks consisting of various routers and switches. This paper investigates how these multi-hop topologies influence the choice between the two primary fronthaul strategies: data-sharing and compression-based transmission.

System Model and Strategies

The study considers a downlink C-RAN where the CP is connected to multiple RRHs through a network of intermediate routers.

Architecture of a multi-hop Cloud Radio Access Network (C-RAN) with a central processor (CP), intermediate routers, and remote radio heads (RRHs) serving mobile users.

Data-Sharing Strategy

In this approach, raw user messages are multicast from the CP to a set of serving RRHs. These RRHs then cooperatively transmit the signals to users via joint beamforming. This strategy requires efficient multicasting, which the authors facilitate through network coding. Network coding allows different data flows to be combined at routers, transforming the traditionally NP-hard Steiner tree packing problem into a manageable optimization task.

Compression-Based Strategy

Here, the CP performs beamforming and then quantizes the resulting signals. These compressed signals are unicast to specific RRHs, which decompress them for wireless transmission. While effective in direct-link scenarios, this strategy relies on unicast routing, which can be less efficient in multi-hop environments as it forces the network to carry unique, high-bandwidth quantized streams to every RRH.

Optimization Framework

The authors propose a cross-layer design to maximize network throughput. The optimization problem is non-convex and involves several interdependent variables across the physical and network layers. To solve this, the paper employs Successive Convex Approximation (SCA) to iteratively transform the non-convex problem into a series of solvable convex sub-problems.

Furthermore, sparse optimization techniques are used to handle the discrete nature of user-RRH associations. By using a continuous approximation of the discrete "on/off" association function, the algorithm can effectively "prune" unnecessary links and optimize the network's sparsity. The resulting two-stage algorithm first identifies the best associations and then refines the beamforming and network coding vectors.

Key Insights and Results

The core finding of the research is a "topology flip" in strategy performance based on the fronthaul architecture:

  • Single-Hop Advantage: In star topologies, the compression-based strategy typically outperforms data-sharing by approximately 25% because it avoids the overhead of multicasting raw messages.
  • Multi-Hop Advantage: In multi-hop scenarios, the data-sharing strategy with network coding becomes superior, offering up to a 30% throughput gain over non-optimized schemes.

The efficiency of multicasting is the primary driver. In a multi-hop network, sending the same information to multiple nodes (multicast) is fundamentally more resource-efficient than sending unique quantized streams to each node (unicast). The proposed iterative algorithms are also computationally efficient, typically converging to a locally optimal solution in fewer than 10 iterations.

Conclusion

This research highlights that the optimal C-RAN fronthaul strategy is highly dependent on the underlying network topology. For modern, complex multi-hop deployments, a data-sharing strategy integrated with network coding and cross-layer optimization provides significant throughput benefits. While challenges remain regarding perfect Channel State Information (CSI) and real-time computational complexity, this framework provides a clear path for improving efficiency in next-generation wireless networks.

Summary updated

Abstract

There are two fundamentally different fronthaul techniques in the downlink communication of cloud radio access network (C-RAN): the data-sharing strategy and the compression-based strategy. Under the former strategy, each user's message is multicast from the central processor (CP) to all the serving remote radio heads (RRHs) over the fronthaul network, which then cooperatively serve the users through joint beamforming; while under the latter strategy, the user messages are first beamformed then quantized at the CP, and the compressed signal is unicast to the corresponding RRH, which then decompresses its received signal for wireless transmission. Previous works show that in general the compression-based strategy outperforms the data-sharing strategy. This paper, on the other hand, points out that in a C-RAN model where the RRHs are connected to the CP via multihop routers, data-sharing can be superior to compression if the network coding technique is adopted for multicasting user messages to the cooperating RRHs, and the RRH's beamforming vectors, the user-RRH association, and the network coding design over the fronthaul network are jointly optimized based on the techniques of sparse optimization and successive convex approximation. This is in comparison to the compression-based strategy, where information is unicast over the fronthaul network by simple routing, and the RRH's compression noise covariance and beamforming vectors, as well as the routing strategy over the fronthaul network are jointly optimized based on the successive convex approximation technique. The observed gain in overall network throughput is due to that information multicast is more efficient than information unicast over the multihop fronthaul of a C-RAN.
BibTeX
@article{liu2017cross,
  author = {Liang Liu and Wei Yu},
  title = {{Cross-Layer Design for Downlink Multihop Cloud Radio Access Networks With Network Coding}},
  journal = {IEEE Trans. Signal Process.},
  year = {2017},
  url = {https://openreview.net/forum?id=vB_-9ddjgr},
}
Figure