In past articles [1, 2], I stressed that Cloud RAN is a disruptive technology. There are a few reasons for this, but I think that most critical reason is that Cloud RAN breaks open a pricing structure that’s been in use ever since the wireless industry was created. The current pricing model for base stations is based on a tight coupling between hardware and software that is impossible to separate. So, when a network operator buys a base station, the operator selects how many frequency carriers each sector would support and a corresponding number of remote radio heads. This model becomes obsolete in Cloud RAN where the hardware and software are decoupled. There is no longer a 1:1 relationship between baseband modules and RRHs due to pooling and virtualization. New pricing schemes are now possible as there is more room for operators to optimize the subsystems they need in the network. Cloud RAN not only decouples hardware from software, but also changes the coupling among hardware subsystems. This has profound implications on the future cost structure of wireless networks and operators have taken notice. OEMs looking to challenge the position of the primary entrenched incumbents are leading the charge in Cloud RAN development with a vision to increase their market share.
A recent development brings Cloud RAN a step closer to commercial realization with the first ever development of physical layer acceleration technology on silicon – a joint effort by eASIC and ASOCS. ASOCS acceleration technology allows computationally intensive physical layer functions to run on specially designed hardware that interfaces with Intel GPPs which run other functions that are not as sensitive to timing response and synchronization, and are not as demanding on computational intensity. eASIC will implement the technology in silicon to provide a cost effective and power efficient module.
Cloud RAN can shake up the wireless industry by enabling new business models. CRAN can be deployed as a service platform for coverage and capacity in venues similar to DAS and small cells. In this model a neutral host service provider offers a CRAN platform that is shared between multiple carriers running multiple technologies. The neutral host may even be an equipment vendor that leases their product as a service – a trend that is well established in the Cloud and data center space where SDN and NFV concepts have a firmer foothold than in telecom. The CRAN platform allows flexible configuration and provisioning such that carriers can maintain full independence from each other as well as have a layer of separation from the neutral host service provider as CRAN allows decoupling of hardware from software and allows multiple relationships between changing modules. This lowers the cost of equipment (over DAS) and provides better performance than small cells through superior interference management.
Fundamentally, while CRAN pushes NFV to the edge, the disruption that the technology brings center on allowing new business models and pricing schemes. These are some of the hottest areas in the SDN/NFV space today where vendors are looking to transition from being pure product and solution providers (capex model) to being service providers (opex model). While the telecom ecosystem is known for being conservative, and carriers still want to own their infrastructure, the pressure to improve operational and financial performance may well be too strong to hold off CRAN from making its impact by allowing operators to adopt an opex model, at least where possible.