Over last 3-5 years, with more than 70% traffic in wireless network’s coming from indoors, the network operators have been increasingly looking at small cells to complement macro cells. However, in the last 18-24 months, operators, particularly China Mobile and KT Telecom, have been working on having Centralized RAN (C-RAN) network; wherein the cell head is limited to being a Remote Radio Head (RRH) while the BaseBand Unit (BBU) of all cells are hosted together in a centralized location. Such a system results in a compact radio unit like in small cells, while offering multitude of other benefits. Operators are now working on power reduction in C-RAN in order to be used for indoor and enterprise segments, the forte of small cells. So now, the question arises, what will be the effect on small cells due to C-RAN? In order to compare the two, let’s first understand the details of each technology.
Characteristics of Small Cells
Small cells are primarily compact base stations that communicate with core network. Like in macro cell, small cells comprise of both radio and baseband unit. Also, they have multiple variants - Micro, Pico, Femto, etc. each with different transmission ranges, but all operating in licensed spectrum.
Small cell deployment could be planned or unplanned; deployed by both individuals in-home and/or by operators/3rd party agencies in campus-like environment, shopping malls, etc. While it enhances the overall capacity significantly, it also puts heavy burden on radio resource allocation and interference management. In order to circumvent this, 3GPP in Release 9, introduced SON (Self-Organizing Network). While SON has been able to address interference challenges to large extent, it has still not been able to optimize multiple objectives like throughput, network resource, energy and cost. Having said that, so far there have been more than 11 million deployments, by 47 operators globally.
Moving towards C-RAN
C-RAN primarily separates the RF and baseband functionalities. While RF would be handled by a compact RRH; the centralized BBU would be responsible for all operations, configurations and resource allocation across the coverage area. This could either be a data-centre or located in cloud (reason why C-RAN is alternatively called Cloud-RAN). Recently, China Mobile, in collaboration with Intel, carried out detailed investigation. C-RAN not only enables dynamic resource management across different cells, depending on real-time factors like - number of users in cell, traffic load, channel conditions, etc., but also provides host of other advantages like capex and opex savings, increased asset utilization, energy savings, etc.
As with most technological inventions, C-RAN comes up with its own challenges. Since RRH transmits oversampled I/Q symbols to BBU, there is a humongous amount of traffic that needs to be transmitted to BBU with very low latency (~1-2 micro-second). This is possible only if the communication between RRH and BBU takes place over high-speed optical fibre. While operators across the globe have been installing fibre for better connectivity, C-RAN makes fibre/very-high-speed-transmission (order of Gbps), a compulsive necessity. Countries in APAC region, particularly Japan, Korea and to certain extent China have large amount of fibre already deployed, especially in their big cities, while it is not the case in US, Europe, etc.
Impact of C-RAN on Small Cells?
C-RAN being in nascent stage; has no standards yet for resource management, energy savings, capacity optimization, etc. On the other hand, small cells have been existent for quite a while; and have in fact crossed the peak of ‘Hype-Cycle’, with good amount of success. In this context, it would be worth looking into the impact of C-RAN on small cells; and importantly, the difference between future version of small cells with segregated radio/baseband unit and C-RAN. This still remains unclear at this stage. While various operators, vendors and reports from Gartner, Fierce-Wireless, etc. have their own views, it looks like C-RAN will be highly effective in large, dense and fibre-present environment like train stations, airports, etc. while small cells might be a natural choice for homes and small offices.
Moreover, with the recent focus on cloudification and virtualization of telecom network functions (NFV), C-RAN is set to grow as a prominent technology. Strikingly, even SmallCellForum, in 2014, have started a detailed study on bringing the aspect of virtualization in small cells; and the impact of NFV and Software-Defined Networking (SDN) in small cells. I believe that, over the coming months, it is imperative to have a closer look on the implication of SDN and NFV on these two technologies, before we could arrive at C-RAN v/s small cell conclusion.