The gains afforded by cloud radio access network (C-RAN) in terms of savingsin capital and operating expenses, flexibility, interference management andnetwork densification rely on the presence of high-capacity low-latencyfronthaul connectivity between remote radio heads (RRHs) and baseband unit(BBU). In light of the non-uniform and limited availability of fiber opticscables, the bandwidth constraints on the fronthaul network call, on the onehand, for the development of advanced baseband compression strategies and, onthe other hand, for a closer investigation of the optimal functional splitbetween RRHs and BBU. In this chapter, after a brief introduction to signalprocessing challenges in C-RAN, this optimal function split is studied at thephysical (PHY) layer as it pertains to two key baseband signal processingsteps, namely channel estimation in the uplink and channel encoding/ linearprecoding in the downlink. Joint optimization of baseband fronthaul compressionand of baseband signal processing is tackled under different PHY functionalsplits, whereby uplink channel estimation and downlink channel encoding/ linearprecoding are carried out either at the RRHs or at the BBU. The analysis, basedon information-theoretical arguments, and numerical results yields insight intothe configurations of network architecture and fronthaul capacities in whichdifferent functional splits are advantageous. The treatment also emphasizes theversatility of deterministic and stochastic successive convex approximationstrategies for the optimization of C-RANs.
展开▼
