2. Inter-cell Coordination Evolution Strategies

 

CA improves network capacity by broadening frequency bandwidth, whereas inter-cell coordination technologies do the same task by enhancing frequency efficiency. Inter-cell coordination is designed to manage radio resources more efficiently by having cells in different sites share user and/or cell information with each other.

 

Inter-cell coordination can also be used both in small cell-introduced HetNet and a legacy homogeneous network. But, more complicated and refined coordination is required because in HetNet, where both high power and lower power cells are deployed together, UEs at cell edges are likely to experience different interference situations.

 

Inter-cell coordination technologies applicable to a macro cell network include ICIC and DL CoMP (proprietary) commercialized in 2012. In MWC 2014, inter-site CA was demonstrated showing how cell sites can cooperate with each other for optimized CA. CoMP operates in a centralized way based on C-RAN (A-SCAN) introduced in 2012 along with some of SK Telecom's proprietary technologies like improved scheduling, energy efficiency, etc. Since CoMP commercialization, ICIC has been replaced by CoMP.

 

SK Telecom began to use small cells in its networks in 2014. The more small cells are used, the higher frequency reuse ratio is achieved. However, there have been some drawbacks, like higher handover rates, stronger interference, increased control overhead, etc. as more cells mean more cell edges. So, to overcome these issues, and to maximize the effect of network capacity increased by small cells, an appropriate method of inter-cell coordination should be chosen depending on how densely small cells are deployed.

 

To enhance network capacity efficiently depending on the degree of small cell deployment in a macro cell, SK Telecom presented SUPER Cell concept, and a 3-phase evolution plan in MWC 2013. In line with the plan, the company is gradually moving forward to its strategic destination, successful commercialization of SUPER Cell in 2016. In October 2013, it conducted a demonstration of Virtual One Cell (SUPER Cell 1.0).

 

Table 3 provides a brief overview of SK Telecom's 3-phase plan for evolving SUPER cell.

 

Table 3. SUPER Cell Evolution Strategies

 

Below, we will discuss some inter-cell coordination technologies: inter-site CA for a macro cell-based homogeneous network, and SUPER Cell for HetNet. Inter-site CA can also be used in HetNet, and will be discussed under SUPER Cell 2.0 section.

 

2.1 Inter-Site CA in Macro Cell Networks

 

CA, designed to increase speeds (by n times) by combining different frequency bandwidths, may slow down UE's speeds or interrupt CA communication if coverages of the aggregated frequency bands do not match. Coverage mismatches are usually caused near cell site edges.

Inter-site CA lets BBUs cooperate with each other to ensure frequency bandwidth aggregation is performed not only between the bands in the same cell site, but also between ones in different sites. If UE moves out of the coverage of one band while still within the coverage of the other band, the UE's aggregated speed naturally drops. In such case, inter-site CA allows the UE to replace the "out-of-coverage" band with the same band, although operated by the neighbor cell site, that has better channel condition, so that the UE can continue to benefit from the band aggregation.

 

Figure 4 provides an example of the effect of inter-site CA that can be gained in a macro cell network that supports 2-band CA. In the example, there are two cell sites (Cell site 0 and 1), and they each have two cells that are operated by two bands (F1₀ and F2₀ from Cell site 0, and F1₁ and F2₁ from Cell site 2). A CA-enabled device, by connecting to a serving cell in each band (PCell in one band and SCell in the other), receives data from both serving cells.

 

Figure 4 (a) shows a case where inter-site CA is not supported. Before or after handover, if the quality of one frequency band becomes degraded at cell site 0 or 1, CA performance becomes downgraded too. On the other hand, Figure 4 (b) illustrates a case where inter-site CA is supported. Both cell sites, through mutual cooperation, ensure optimal CA at cell edges any time, even in case coverages between the bands are mismatched, by dynamically combining frequency bands to increase the aggregated transmission rate.

 

 

Figure 4. Inter-site CA in Macro Cell Networks

 

2.2 SUPER Cell 1.0: Virtual One Cell 

 

UPER Cell 1.0 architecture is used during initial stages where small cells are introduced, and works best when there are not many small cells in a macro cell yet. Small cells use the same frequency and Physical Cell ID (PCI) as the macro cell's, and both small and macro cells work as Virtual One Cell.

 

Because this architecture causes no handover when devices switch from one cell to another, it prevents call drops, and improves transmission quality during communication near cell edges. Unfortunately, however, the effect of network capacity enhancement is minimal because the effect of frequency reuse, a benefit of introducing small cells, is not expected.   

 

To accommodate this issue, transmission mode 9 (TM9) newly defined in 3GPP Release 10 is employed in this architecture. TM9 allows a device to receive the same signal from more than one cell while staying near cell edges, and to communicate at maximum speeds while staying at cell centers, thereby effectively enhancing network capacity.

 

In October 2013, in a demonstration showing how Virtual One Cell works, SK Telecom proved the transmission speeds of devices at cell edges increased by 1.5 ~ 2 times, and network capacity by 5~10%. The company is now preparing for its commercialization in late 2014.

 

Figure 5. Virtual One Cell based on TM9

 

2.3 SUPER Cell 2.0: Elastic Cell and Inter-site CA

 

SUPER Cell 2.0 can be best used when there are a moderate number of small cells in a macro cell after initial stages of introducing small cells. In this phase, unlike phase 1 Virtual One Cell, small cells have cell IDs (PCIs) different from the macro cell's, and work as independent cells. Compared to phase 1, more signal interference is caused. So, this architecture controls interference through inter-cell coordination, rather than having the cells work as one cell, and additionally assigns different frequencies to some small cells.

 

Macro and small cells improve cell edge performance by supporting CoMP and inter-site CA through mutual cooperation. Cells that use the same frequency cooperate with each other using CoMP while those that use different frequencies cooperate using inter-site CA.

 

Both CoMP and inter-site CA are performed by the central scheduler located at a BBU pooling site, but they use different ways to improve cell edge performance: In CoMP, the central scheduler dynamically selects a groups of cells that experience good channel condition with UE, lets the cells send the same data to the UE, and turns off the cells that cause interference. On the other hand, in inter-site CA, it dynamically combines frequency bandwidths in different sites

 

HetNet that works based on CoMP is called Elastic Cell by SK Telecom. In July 2014, SK Telecom, in a demonstration showing how Elastic Cell technology works, confirmed the data transfer rates at cell edges actually were improved by 50%. The company aims to commercially launch the technology by 2016.

 

Elastic Cell is the core technology of SUPER Cell, and helps the paradigm of data transfer between base stations and users to shift from cell-centric to user-centric. Previously, UE could communicate with only one cell at a time, and had to search for a cell that has the strongest signal strength, itself. But, now Elastic Cell allows UE to communicate with more than one cell at once, and enables the network to select a transmission cell group to communicate with the UE. Cells that are experiencing excellent channel conditions with UE are dynamically selected as a transmission cell group.  

 

Figure 6 is an illustration of UE communication in SUPER Cell 2.0, and shows how the number of transmission cells change as UE moves from one place to another: 1 → 2 → 3 → 2 (CoMP used in t2 and t3, and inter-site CA in t4). As such, SUPER Cell 2.0 guarantees UE is always served by the best cells, using CoMP and inter-site CA.

 

Figure 6. CoMP-based Elastic Cell and Inter-site CA

 

2.4 SUPER Cell 3.0: Hierarchy Cell

 

SUPER Cell 3.0 can be effective once small cells become densely deployed in a macro cell. At this stage, UE is usually within the coverage of some small cells no matter where it is, and thus would inevitably experience frequent handovers every time it moves. Moreover, transmission efficiency of the small cells would drop because of drastic increase in control overhead (e.g. handover control info., neighbor cell measurement reports, broadcasted system info., etc.), and mobile batteries would not last long because of frequent handovers.  

 

In the legacy architecture, cells are designed to deliver both control signalings and user data together, and the size of available transmission bandwidth is limited because conventional frequency bands are usually at lower frequencies below 3.5 GHz. So, the legacy architecture has not been very effective in improving network capacity when small cells are densely deployed in a macro cell. On the other hand, in Hierarchy Cell, control signalings (control plane) and user data (user plane) are separated and delivered through different radio paths according to their QoE parameters. Small cells become high-capacity cells as they get to use much broader bandwidths in higher frequency bands (e.g. 3.5 GHz or 30 GHz) than macro cells. Control signalings and VoLTE data that require broader coverage are delivered by a macro cell, while user data which requires fast transmission is delivered by a higher-capacity small cell.

 

The key idea of this architecture is dual connectivity. This means that UE can be connected to both macro and small cells at the same time. The macro cell, with broader coverage, always serves as a primary cell, delivering control signalings and working as mobility anchor. So, even when UE keeps moving from one small cell to another, no handover is caused, and mobile batteries last long. High-capacity small cells always serve as secondary cells, taking care of data delivery. Thanks to the dual connectivity, little control overhead is caused to small cells. Moreover, small cells, now that broader bandwidths are secured, can focus on fast transmission, achieving higher transmission efficiency even in a highly-dense cell environment, and thereby improving network capacity effectively. This architecture is scheduled to be commercialized in 2016.  

 

Figure 7. Hierarchy Cell based on dual connectivity