Technical Overview of CoMP in Release 16

To cope with highly reliable and low latency communications’ stringent requirements and high demand for data traffic, the 3rd Generation Partnership Project (3GPP) standardization body and its main technical partners have made considerable efforts to promote many advanced cellular technologies—e.g., massive Multiple Input, Multiple Output (MIMO), dynamic spectrum sharing, and Coordinated Multipoint (CoMP) transmission and reception. Among these, CoMP, originally defined in the 3GPP’s Release 11, has been continuously developed in the newly completed 3GPP Release 16, which was frozen in July 2020. In this new standard release, CoMP, combined with Time-Sensitive Networking (TSN) technology, is positioned to accomplish reliable 5G communications and offset the performance degradation problem of cellular network operations in various frequency bands, including the unlicensed ones. On one hand, this documentation implies the 3GPP’s ambition to expand its cellular ecosystem to broader fields. On the other hand, it also exposes the performance shortage of network operations in certain frequency bands, especially for the expansion of industrial vertical applications.

Mobile Network Operators (MNOs) typically rely on increasing the number of base stations and antennas to improve spectrum efficiency, a process that is referred to as cell-site densification plus spatial domain radio resource allocation. This strategy works well for User Equipment (UE) located near the center of the cell. However, due to inter-cell interference, throughput improvement is insignificant for users near the edge of the cell. By enabling multiple base stations to cooperatively serve a single user, CoMP technology can help achieve better throughput by reducing/avoiding inter-cell interference. In general, depending on the required constraints on the backhaul link between base stations and the level of scheduling complexity, CoMP can be classified into three categories:

• Coordinated Scheduling/Beamforming (CS/CB)
• Joint Processing (JP)
• Dynamic Point Selection (DPS)

With these types of implementation scenarios, CoMP groups a set of base stations to form a distributed MIMO system to serve UE in a virtual cell. As we mentioned, CoMP is not a new technology, but a few years after the technology was introduced, the 3GPP documented it in its new standard release and expanded it to unlicensed frequency bands. Does that mean the technology is ready to bring immediate benefits to 5G?  

Before CoMP was introduced in the 3GPP’s specifications, two technologies, i.e., Inter-Cell Interference Coordination (ICIC) and enhanced ICIC (eICIC), already existed to solve cell edge UE inter-cell interference problems by allocating different frequency or time resources. Jointly considering the three types of CoMP scenarios above, we summarize the main differences below:

Compared to CoMP, the problems with ICIC and eICIC are that neither of them can improve system throughput because of their restriction of radio resource usage. On the other hand, CoMP allows the use of radio resources not only in the frequency/time domain, but also in the spatial domain via MIMO beamforming to improve spectrum efficiency. However, to fully exploit the advantage of CoMP, the information data of a UE and/or its local Channel State Information (CSI) need to be shared among the cooperative base stations, which requires high capacity backhaul with low transmission delay—e.g., to form ideal backhaul links, no fewer 10 Gbps capacity and up to 2.5 microseconds latency are required according to the Next Generation Mobile Networks (NGMN) Alliance. Apart from that, UE also needs to exploit complicated signal processing algorithms to be able to identify the cells’ IDs, especially for heterogeneous networks. Moreover, time and frequency synchronization errors also have a severe impact on the performance of CoMP technology. TSN can help to build the synchronized spectrum access, but the different base station/UE distances with respect to different cell locations will create an additional communication delay that is unavoidable. In this case, overhead signalling is required to offset the frequency shift and limit the loss caused by beamforming mismatch. This situation is even worse for CoMP-JP, since multiple base stations need to simultaneously serve single UE.

By promoting CoMP in its Release 16, the 3GPP aims to further compensate for the performance loss of network operations near cell edges. In this case, by enabling multiple Access Points (APs) with multiple antenna to communicate with single UE, the link reliability can be improved. However, this gain comes at the cost of stringent requirements, as discussed above. In consideration of the current stage of network deployments, CoMP technology has the potential to radically improve indoor wireless networks for enterprise verticals, where the cost/benefit ratio will be favorable to a more expensive CoMP deployment. On the other hand, the Centralized Radio Access Network (C-RAN) topology could help lower the threshold of CoMP deployment, where no stringent backhaul (fronthaul) capacity or latency is required. Moreover, to fully exploit the unique feature of this technology, especially in unlicensed bands, industrial enterprises—in collaboration with their partners, such as MNOs/Communications Service Providers (CSPs), equipment suppliers, and end-market Systems Integrators (SIs)—should be aware of the performance impacts of random interference introduced by external nodes and select network deployment locations where these external impact factors can be efficiently controlled.  

Traditionally, CoMP is a single vendor-specific solution due to the proprietary X2/Xn interface. The growing momentum for Open RAN and its established open interface specifications allow multi-vendors to interoperate and leverage a flexible network deployment environment to drive technological innovation and reduce deployment costs. In theory, with the open X2/Xn interface between two different base stations, multi-vendor interoperable CoMP is applicable. However, considering the stringent backhaul latency and tight time/frequency synchronization requirements, the current network deployment conditions can hardly make it a reality. These requirements may be stringent and not realizable at the moment, but if 5G enterprise cellular networks take off, market demand may create a critical mass of applications and innovations that can drive the supply chain to reduce CoMP deployment thresholds.

ABI Research sees CoMP as a very promising technology to not only enhance throughputs of UE at cell edge areas but also improve communication reliability for industrial enterprise networks operated in licensed, shared, dedicated, and even unlicensed bands. However, all these gains can only be achieved under the premise that backhaul capacity and latency requirements are fixed and the device ecosystem is broadening.