5G trends – buzzwords reviewed

As the “5G” research and standardization is progressing, I would like to elaborate a bit on some concepts that are potentially serving as building blocks for the future systems. They are divided in this post into: Air interface and Networking concepts.

Selected air interface aspects

• millimeter Wave (mmW) – The high-frequency ranges are currently becoming very important as they offer large spectrum blocks to provide the high capacity required for the 5G system. The industry is currently speaking of frequencies ranging from 24GHz to up to 86GHz (targeting WRC-2019). Of course, those come at the expense of short ranges (propagation losses), lower diffraction and weaker NLOS operation.

• Massive MIMO (MMIMO) – A coherent Tx/Rx using tens to hundreds or thousands of antenna elements are envisioned to fit perfectly with the use of mmW, because it can maximize the mmW range/coverage by using very narrow beamforming. At the same time, mmW also reveals the possibility to use massive antenna panels consisting of small individual antenna elements. This concept assumes that the massive antennas are at the “site” providing multi-user MIMO to devices which on the other hand utilize a lower number of antennas.

• Unified Air Interface (UAI) – As 5G use cases defined by 3GPP pose very different and sometimes contradicting requirements along with KPIs (e.g. latency, throughput, reliability, transmission character), there’s no way that the good old “one-size-fits-all” design principle can be applied. Thus the air interface covering different access schemes and waveform parameters is needed. One such example is a PHY layer frame design, where different numerologies for waveforms are discussed together with scalable TTIs to cover the “three edges of the services triangle” from ITU-R, namely eMBB (enhanced mobile broad-band), mMTC (massive machine type communications) and URLLC (ultra reliable and low latency communications).

Selected networking aspects

• Network Functions Virtualization (NFV) – Incorporates an IT concept to decouple software from hardware, i.e. virtualize network node functions so that they can be placed in a commodity hardware in different places (e.g. edge, center clouds) and defined as building blocks that can be “chained” together to create a virtual/logical network or a service in an optimized manner. This can be combined with SDN to create a fully software-based dynamic system, where new services can be created in a very efficient and fast manner.

• Software Defined Networking (SDN) – Incorporates the concept from computer/IP networks to decouple the Control-Plane (decision of where the traffic is sent) from User-Plane (actually forwarding the traffic) enabling efficient and separate optimization of each Plane. Through this, we can dynamically manipulate and optimize the network functions placement and operation to tailor them to specific needs and utilize resources more efficiently compared to static designs.

• Network Slicing – Enables to utilize a single infrastructure to provide different and, to a certain extent, independent logical networks. Through this, a network operator can provide tailored and isolated services to different users/tenants/businesses.

• Tight Interworking with evolved LTE-Advanced – As further evolution of LTE-Advanced is to be a part of the 5G system (for the operators to leverage their previous investments), the tight interworking with the new radio (NR) is a natural requirement. To enable this, concepts such as “Multi-RAT Dual-Connectivity”, LTE (NR) connectivity anchoring and RAN-based Multi-RAT handover have been introduced.

• Ultra Dense Networks (UDN) – A system where the ISD (Inter-site distance) is in the range of 10-50 meters and small cells provide solid coverage in the area (instead of a spotty one as in HetNets). This of course would not be provided everywhere and in a nation-wide way, but rather in hotspot areas. Here, multiconnectivity, interference management and efficient energy saving mechanisms are the most crucial for the design.

• User Centric Networking (UCN) – Envisions a system where “the network follows the user” in multiple aspects, e.g. from tailoring the design such that the content is optimally distributed to provide a “zero touch” experience, through QoE concepts, down to “no-cell” where there’s no cell edge and multiple transmission points serve the user.

• Native Self-Organizing Networks (SON) – As the first “version” of SON was designed as an add-on to the already existing system, it was not that successful (especially for the small cells), with 5G, there’s a possibility to do that differently, i.e. provide automation within the design of the next generation system (not on top).

• Native Device-to-Device communications (D2D) – D2D was added after the LTE system was standardized and “set in stone” (i.e. as a new feature). In 5G, it is envisioned that the device could be natively treated flexibly acting either as an end-point, or as a part of the network relaying the communications to other devices for certain use-cases (of course not for all devices and it should depend on the capabilities and use cases).

Summary

This post should serve as a short survey of “what’s out there” in terms of building blocks and concepts related to the “5G”, pointing to some directions where the communications world is moving. Summarizing, the key 5G trends in my opinion include: more flexibility in the design, forward compatibility and software based networking.

Stay tuned to get more details on the above in the future!