Robust and Efficient Distributed Compression for Cloud Radio Access Networks

TL;DR

This paper proposes a robust, distributed compression scheme for uplink cloud radio access networks that accounts for imperfect statistical information and includes a joint base station selection method to improve energy efficiency.

Contribution

It introduces a robust compression method based on worst-case optimization and a joint BS selection algorithm for energy-efficient network operation.

Findings

Robust compression compensates for performance loss due to imperfect covariance information.

Proposed BS selection performs close to exhaustive search.

The algorithms converge to locally optimal solutions.

Abstract

This work studies distributed compression for the uplink of a cloud radio access network where multiple multi-antenna base stations (BSs) are connected to a central unit, also referred to as cloud decoder, via capacity-constrained backhaul links. Since the signals received at different BSs are correlated, distributed source coding strategies are potentially beneficial, and can be implemented via sequential source coding with side information. For the problem of compression with side information, available compression strategies based on the criteria of maximizing the achievable rate or minimizing the mean square error are reviewed first. It is observed that, in either case, each BS requires information about a specific covariance matrix in order to realize the advantage of distributed source coding. Since this covariance matrix depends on the channel realizations corresponding to other BSs, a robust compression method is proposed for a practical scenario in which the information about the covariance available at each BS is imperfect. The problem is formulated using a deterministic worst-case approach, and an algorithm is proposed that achieves a stationary point for the problem. Then, BS selection is addressed with the aim of reducing the number of active BSs, thus enhancing the energy efficiency of the network. An optimization problem is formulated in which compression and BS selection are performed jointly by introducing a sparsity-inducing term into the objective function. An iterative algorithm is proposed that is shown to converge to a locally optimal point. From numerical results, it is observed that the proposed robust compression scheme compensates for a large fraction of the performance loss induced by the imperfect statistical information. Moreover, the proposed BS selection algorithm is seen to perform close to the more complex exhaustive search solution.