Cost Saving Analysis in Capacity-Constrained C-RAN Fronthaul

Cloud radio access network (C-RAN) is envisaged as a potential enabler of 5G and future generation mobile networks to meet the diverse and stringent requirements of 5G use cases and application scenarios due to its promising advantages. For practical deployment, the proposed C-RAN architecture should not only be cost effective but also energy efficient. Common public radio interface (CPRI) based fronthaul (FH) demands huge bandwidth due to new radio access technologies (RATS), which necessitate a large number of high capacity transceiver (TRX) equipment to be deployed. In a wavelength division multiplexing-passive optical network (WDM-PON) system, TRX are deployed to serve at a peak load, but, because of variations in the traffic demands, this situation occurs with a very low probability. Therefore, allowing a reasonable outage probability, FH can be dimensioned requiring a lower capacity, and consequently fewer WDM TRX, thus giving rise to twofold advantages: cost saving (by deploying fewer TRXs), and energy saving (by intelligently putting the unused TRXs in sleep mode). In this paper, we develop an efficient FH TRX cost model, and show through the numerical results FH cost saving up to 50% at a moderately low traffic density of 200 Gbps/km(2) by exploiting traffic randomness using queuing theory and spatial traffic models.