DAS (Distributed Antenna Systems) are usually positioned as the very expensive competitor against the up-and-coming, low cost small cell products. Radisys have been at the forefront of small cell development, providing software solutions for many ODMs and OEM players , so it's something of a surprise to see them talking about DAS.
We asked their senior line product manager, Renuka Bhalerao, for more detail.
So what's changing?
We've seen DAS vendors moving forward to adopt and embrace new in-building wireless solution architectures. Today, there are many so-called Active DAS products using fibre rather than coax to distribute the RF signal around the building. DAS is traditionally multi-operator and connected to multiple macrocells, one per network operator. By contrast, small cells are low cost, compact and low power but usually tied to a single network.
Several DAS providers are taking an interest in combining both technologies, making the best of both elements. A number already separately offer both DAS and small cell products today; a few have plans for a combined product where the small cells rather than macrocells drive the DAS system.
Indeed, TE Connectivity already offers this kind of architecture. I believe other DAS vendors such as Rosenberger and Corning share this view of leveraging small cells with DAS. This isn't a centralised baseband architecture promoted by the macrocell vendors (like Ericsson Radio Dot, Huawei Lampsite etc.), instead it's just replacing the macrocell with lower cost, lower power small cells.
The market for this is primarily LTE only today but could expand into 3G and potentially become multi-mode (3G/LTE).
Building size is the primary factor to determine if a DAS system is viable. Standalone Small Cells can address the needs for smaller buildings where the higher costs of a full DAS solution wouldn't be justified.
How would the small cells interconnect with the DAS system?
Analog RF is commonly used today and would be quite straightforward and feasible to use. A more efficient way is to use CPRI, providing a direct digital interface into the active DAS system and avoiding an unnecessary stage of RF encoding/decoding.
This also saves power consumption and the associated HVAC.
CPRI has several optional features and proprietary extensions, and we'll be looking carefully at how those affect us. I'd expect Tier 2/3 vendors to be most active in this space.
It is possible to connect both analog RF and digital CPRI into the same DAS system, providing savings for operators who adopt the more innovative technology quickly while retaining compatibility for those who are more conservative and less concerned about cost.
Are there any implications for site planning?
Small cells have a lower capacity per sector, say 64 or 128 concurrent users per LTE cell rather than 400 or more for a macrocell. But this can be a benefit too because it makes it easier to add extra capacity and scale more easily – you simply allocate more sectors.
Large DAS systems are usually deployed in zones where several radio heads share the same sector and RF signals. If there is a fault, the entire zone can be affected. By allocating a larger number of smaller zones, the impact of any individual outage such as a cabling fault is more limited.
It would also be possible to deploy both public access small cells and DAS in the same building, say to augment capacity for one operator in a specific area.
Can't Small Cells support multi-operator without DAS?
There are two existing 3GPP standards which do make this technically feasible. Both are already in commercial use today.
MOCN (Multiple Operator Core Network) is a network sharing concept that is fairly straightforward to implement, similar to network roaming. MORAN (Multiple Operator Radio Access Network) is more involved, providing a deeper integration and requiring more software within the small cells.
Some of our key customers are already using MOCN, but I see MORAN being the long term future – sharing the RAN as an end-to-end resource rather than just connecting through the core network. I can see a major challenge as we move outdoors to the streets where practical deployment constraints won't allow multiple small cells to be strung from the same lamp-posts.