Change Rattles Radio Access Networks

Last week we conducted a webinar called “Change Rattles Radio Access Networks,” in which we discussed the recent advancements of radio access networks (RAN). The main topic was how operators can take advances in radio architecture to build a highly flexible and scalable network to meet the growing demands of their business. Some of the key points covered—cloud-based RAN, baseband pooling, effects of virtualization, and compact base stations—generated several interesting questions for us. Since there wasn’t enough time to address all these questions during the webinear, we decided to continue the dialogue in this post.

To view the original webinar, please click here.

Can you please define the “multi-mode” RAN more clearly?
Many radio technologies such as LTE FDD, LTE TDD, WCDMA, and HSPA + co-exist in the current eco-system. These technologies differ in several aspects, including band utilization, modulation techniques, interfaces, and protocols.

“Multi-mode” RAN base stations are capable of simultaneously supporting these multiple radio technologies. This can be effectively addressed using Software Defined Radio (SDR) where a common radio hardware platform is able to support all these radio technologies simultaneously.

Please elaborate on how cloud-based controllers are implemented?
In a cloud-based controller, the guest operating system, the underlying middleware, and all the controller applications are hosted in a virtualized environment. A virtual machine manager (VMM) provides the hardware abstraction that encapsulates and isolates a given virtual machine from the other virtual machines on the same physical hardware.

Controller applications run on one or more virtual machines. The VMs are dedicated for each role pertaining to cell and UE specific control plane and traffic plane.

If we are able to create compact (or very compact) base stations, then why do we need multi-mode BTS?
Both distributed and compact base station architectures entail SDR-based baseband unita (BBU), as well as multi-mode radio frequency monitoring (RFM). Therefore, both types of base transceiver stations (BTS) provide the capability to migrate from Technology-A to Technology-B without any HW upgrade. That’s why in the current scenario multi-mode support is a mandatory requirement for all base station architectures. The key difference between the two types is in the deployment and overall cost reduction of compact BTS due to reduced size, form factor, etc.

How practical is it to provide optical fiber links between the antennas and baseband pools? What is the likely distance of these fiber links?
For a 100MHz LTE system, CPRI line rate of 6144Mbps should suffice taking into account I/Q sample width, control and management channels. Typical optical transceiver bandwidth available today should be able to support more than one such channel. They also have low noise figure and high spur-free dynamic range, which is good enough to provide transport distances in the order of few tens of kilometers.

Is the RRU & BBU connectivity implemented over “Radio over Fiber”?
There are three possible methods to transport wireless signals over optical link: RF-over-fiber, IF-over-fiber and baseband-over-fiber. RF-over-fiber transport has the benefits of simpler base-station designs, independence of the air-interface and multi wireless band operation. To meet the requirement of connecting large number of RRUs to central BBU, DWDM RoF technique would be a strong candidate.

What role will microwave backhaul play in centralized RAN (C-RAN) using centralized BB, if any at all?
Traditional microwave systems operating in 6-38 GHz Bands have limited transmission bandwidth far below the typical aggregated bandwidth required for a 4G cell site. Traffic capacity required for 4G cell site could be up to 1Gbps. To scale beyond the data rate with traditional microwave systems, multiple RF channels are required, adding to cost of electronics and spectrum licensing. A more cost-effective alternative is to move these backhaul applications to the new 70-80 GHz Eband spectrum, capable of providing multi-gigabit-per-second data rates using only a single, up to 5 GHz wide, radio channel for distances up to 10 km.

What are the trade-offs between the compact BTS versus the distributed BTS with baseband pooling in terms of capacity, coverage and backhaul?
Distributed BTS is more tailored towards baseband pooling concept as we can have centralized pool of all BBUs, enabling utilization of resources across the common pool to address resource requirements.

Will “cloudification” first happen at the controller (RNC/BSC) or base station?
The evolution to the cloud is happening at both the entities—controller and base station. Cloud-enabled controller applications and baseband pooling at base stations is expected to occur as the first step. Movement to cloud platform and virtualization of base stations shall occur in second step. As of now the base stations are taking precedence over the controllers. This is mainly because the advances in the hardware architecture and evolution of multi-standard BTS based on distributed BTS architecture is already taking place, whereas similar innovations for controllers are still in the pipeline. In addition, to supporting LTE Rel-11 and Rel 12 features, it is also essential to have co-operative radio processing which is also ensuring the evolution of cloud concept at base-stations.

In a cRAN, how will network operators handle the challenge of transporting large amounts of I/Q data from the baseband to the RRU?
The evolution towards the cloud-based architecture is based on distributed BTS architecture. The connectivity between BBU and Remote radio head (RRH) is provided using optical fiber links. The additional advantage of using optical links is that it eliminates feeder loss and can provide connectivity over long distances without affecting latency.

In a cloud based controller, how do you resolve time-critical processing as required for LTE latency?
As mentioned earlier, cloud based RAN employs optical fiber links to support baseband pooling. This provides sufficient speed and bandwidth eliminating the latency issues. Over very large distances, we can utilize several IP transport techniques to ensure that the required latency figures are met.

In what timeframe do you envisage cloud based core and/or RAN architectures to become reality?
As per the current market trends, the cloud-based architecture could become reality starting Q3/Q4 2012.

Each technology will have its own performance. How do we optimize each technology given the use of common antenna systems?
Active Antenna Concept helps in addressing the optimization required for each technology. One of the advantages of this approach is that it is possible to electronically tilt elevation beams by having independent control of the phase and amplitude on each of the mini-radios which make up the active antenna. When we consider a multi-standard RAN, it is possible to control the orientations of different carriers (Tx/Rx) in the same frequency band, and also control different air interfaces, while helps in optimizing each technology.

What is the interface between the baseband pool and the remote radio? Is it standardized?
CPRI/OBSAI is the current standard that is used on the interface between BBU and RRU. ETSI standardization activity is going on currently with respect to Open Radio Interface (ORI).

What are the current, commercially available SDR (SW defined radio) capabilities in terms of supportable RF bands bandwidth? Is "tuning" from 900 MHz to 1.8 GHz, or from 1.8 GHz to 3.5 GHz possible?
The current generation of multi-mode BTS don’t support single RFM that can be “tuned” to support all frequency bands. This is one of the challenges being addressed currently to make the technology ready for deployment.

What pooling benefits are achievable if cloud RAN resources are driven by a TDM based CPRI interface?
CPRI defines point-to-point link layer interface (layer 1 and layer 2) and does not define any PHY topology between BS and RAU. Such a link has all the features necessary to enable any topology, including direct interconnection of multiple remote antenna units to the central base station. However the connectivity should meet critical requirements such as BER and latency.


Leave a Reply

Your email address will not be published. Required fields are marked *