Using Cloud Radio to Deliver Promises in the 4G Era

In recent years, mobile broadband has become an essential part of daily life, and mobile communication has entered the cloud era. As mobile technology evolves towards LTE, many operators are having to manage 2G, 3G, and LTE networks simultaneously and are facing challenges such as interference between networks. To fulfill their promises to end users, operators need help to efficiently deploy and manage their complex networks.

The LTE Forum at MWC 2013 was sponsored by ZTE. At the forum, mobile operators, service providers, and providers of infrastructure solutions sat down together to discuss various LTE challenges and solutions. ZTE brought to the table its cloud radio solution, which involves using the cloud for mobile access, mobile coordination, and mobile management so that the operator’s network resources are used optimally. This allows the operator to adapt to complex transmission scenarios and keep their promises in the LTE era.

4G’s Promise

Confucius once said, “If a person cannot keep their promise, then I don’t know what they can do.” What does 4G promise? In the 2G era, the promise was voice calls anywhere, anytime. This promise was kept. In the 3G era, the promise has been mobile data anytime, anywhere. This promise has also been kept.

In the 4G era, what is the promise? Is it 100 Mbps LTE and 1 Gbps LTE-A? No. High peak download speeds are features, not promises. Promises of 100 Mbps or 1 Gbps cannot be kept. 3GPP claims that the promise of LTE is truly global mobile broadband. There are two aspects to this promise: First, LTE is global. UMTS in Europe, CDMA in the US, and even TD-SCDMA in China all converge to LTE. So LTE is truly global. Second, LTE is ubiquitous mobile broadband, and a certain amount of bandwidth is reserved for everybody. With mobile broadband, you need to provide speeds of at least 1 Mbps to 2 Mbps anywhere, anytime. 

Current Reality

People sometimes say that LTE networks are capable of 20 Mbps to 60 Mbps, and with speeds such as these, the LTE promise is kept. However, such speeds are only achieved in certain locations and under certain conditions. What if more and more people come onto the network or a network is expanded to another area? Currently, if you are near a base station or traffic is light, bandwidth is very high. However, if the load is high and you are near an edge, customer experience drops dramatically.

Network capacity does not change in proportion to customer traffic. Where there is no coverage, mobile data might be needed. More importantly, traffic moves as people move. How can we ensure that capacity changes in proportion to traffic?

Cloud Radio: Delivering the 4G Promise

This year, ZTE launched a cloud radio solution. Cloud radio is not a single technology; it is a combination of many technologies. One of these technologies is C-RAN (cloud pooling). If fiber is available, then BBUs can be centralized so that traffic can be handled in a centralized way.

Another of these technologies is cloud P-Bridge. If small cells have been deployed but there is not enough fiber, CAT5 copper lines can be used for cloud pooling. 

Because small cells have smaller radiuses, interference is severe. Therefore, a virtual “super cell” can be built that comprises many small cells or even one macro cell. With a super cell, the same cell ID is used for many small cells so that interference is reduced and customer experience is improved.

If an operator does not have enough fiber, a cloud coordinator and cloud scheduler can be used. 

Cloud Pooling

ZTE offered cloud pooling last year. We worked with China Mobile to build many trial networks. BBU sites were centralized, and at the remote site, only fiber was used to connect with the remote radio head. In this way, rooms did not have to be leased at each radio site, and capex and opex were reduced. With centralized BBUs, resources can be allocated in a centralized way to handle tidal traffic. Inter-site CoMP can be implemented easily with cloud pooling, and this greatly improves customer experience at the cell edge.

Cloud Super Cell

If small cells are deployed in a macro cell, CoMP does not work very well. Therefore, cloud super cell technology is introduced. With this technology, many low-power nodes are used as one cell so that there is no handover and interference between cells (even if the cells’ radiuses are very short). When super cells are combined with other new technologies, such as carrier aggregation, mobile traffic can be handled very smoothly. Small cells are used not only for hotspots but also for high-speed trains. Super cells can handle high-speed handovers without dropouts, even when the train is running at 250 km/h.

Cloud P-bridge

CAT5 copper lines are already present in many indoor environments. How can CAT5 be used with C-RAN technology? ZTE invented P-Bridge, which compresses the CPRI interface below 1 Gbps (usually around 800–900 Mbps). In this way, CPRI data can be transported through copper lines. With local C-RAN, capacity can be significantly increased, the network can be easily installed, and capex and opex can be dramatically reduced. ZTE has already tested this technology with SoftBank, and it is available now.

Cloud Coordination

Many operators, especially in Europe, do not have enough dark fiber. This means they have to share fiber with other operators or share fiber with fixed networks or even enterprise services. IP fiber has two problems. First, unlike dark fiber, IP fiber is usually below 1 Gbps (roughly 400–500 Mbps). This is not very fast. More importantly, the latency of the IP fiber can be as high as 4 ms, and C-RAN technology cannot really be used with this latency. ZTE therefore developed a new technology called cloud co-ordination that can be used for soft bit combining. To reduce the need for bandwidth, the after-process data (soft bits) is used, not the raw data (CPRI data). We combine the soft bits for two eNode Bs. With this kind of coordination, throughput and capacity for cell edge users can still be improved, even when there is only the IP backhaul network. 

Cloud Scheduling

Many operators still have very limited fiber and use microwave for LTE backhaul. The good news is that ZTE has developed a technology called centralized scheduler. Scheduler equipment is added at the central office side, and X2+ interface is used to connect the centralized scheduler and eNode Bs. Only scheduler information is sent to the centralized scheduler, so bandwidth needs are low. Moreover, there is no need for very short latency. The cloud scheduler allows resources from one cell to be added to another when a user moves from one cell to another. Throughput at the cell edge can also be greatly improved. This technology is also available now and has been simulated and tested in real networks.

Simulation and Trial Results

If an operator has dark fiber/OTN with very high bandwidth and very short latency (below 100 µs), cell-edge throughput can be increased by 94% compared to that in a traditional network. For IP-RAN or IP fiber, if the bandwidth is less than 1 Gbps and latency is below 4 ms, more than 83% increase in cell-edge throughput is obtainable. If an operator has no fiber but has microwave with latency below 100 ms, throughput can be increased by 30% using cloud scheduling. For about 5% of cell-edge users, in some circumstances, throughput can be tripled.

The Key Takeaway