Crystal-gazing 5G and beyond
A Think Tank team selected stalwarts, business thinkers and successful leaders in their respective fields pen their thoughts exclusively for us, making gripping narratives.
When we talk about 5G, some immediate things that come to mind are high speeds, low latency, diverse use cases across many verticals, and so on and so forth. But on a macro level, it may be interesting to know the radical changes and transitioning that 5G and beyond will bring forth to the entire telecommunications infrastructure going forward. The evolution and transformation of the networks with 5G and other advanced technologies in the future, are not often discussed in general. As we gauge the possibilities, and crystal gaze into what the networks of tomorrow would comprise of, alongside the implications of the advanced technology and its various components coming into play, some intrinsic aspects of the same need to be understood.
For purposes of discussion, communications technology can be broadly segregated into four layers:
- Physical layer – which transports a bitstream from one machine to another. This layer consists of LAN, packet radio, spectrum, OFC, copper wire, etc.
- Network layer – which delivers packets where they are supposed to go via packet routing. Internet protocol (IP) belongs to this layer.
- Transport layer – is the end-to-end layer from the source to the destination, i.e., a program running on a machine in conversation with other similar program in the destination machine.
- Application (exploitation) layer – which will have various elements like applications, use of networks as a service, evolving customization of resources to meet delivery needs, among several others.
All these layers are impacted with the technological advancement and evolution of the industry to the next generation of technology, albeit some more than the others. Let us see how!
In the physical layer, the availability and optimum use of spectrum is the most vital element. Compared to 4G, the enhanced characteristics of 5G require spectrum with much higher capacities – making the mid-bands and millimeter wave bands more ideally suited for it. Availability of Spectrum in more efficient bands, like the 6 GHz band for example, would help enhance both coverage and capacity for 5G service across the country.
The network planning in 5G also requires more effort. The base transceiver stations (BTS) used in 4G consists of both the radio and the switching equipment, making them more mobile for purposes of relocation and adjustments in planning. In 5G though, the small cells contain radios, which are spread across a given area, and connected to a common base-band unit (BBU). This makes fiberization for 5G an extremely critical requirement, as the switch needs to be connected to the small cells in a robust manner. As underground fiber may not be possible to deploy for this purpose on all instances, use of over-the-air (OTA) or wireless fiber, such as E&V bands becomes a necessity for the industry. Fixed fiber connectivity to the switch may also pose challenges in terms of flexibility to relocate or tweak network planning for 5G in that case. Effectively, fiber optimization for the networks will have to be achieved through a combination of optical and RF (OTA or wireless fiber) forms, which is expected to increase going forward.
Further, new technologies and methods, such as beamforming and full duplex used in 5G differentiates it from 4G. Beamforming is a technology for applying directionally to cell transmissions, i.e., after a base station locates a specific user, it can transmit to them in a targeted way. Transmissions are aimed at the specific user rather than sent in all directions. Full duplex here means 5G’s ability to operate on multiple bands at the same time, i.e., 5G can use this ability to transmit and receive simultaneously.
As we come to the network and transport layers, the differences between 4G and 5G widen up. While public 5G users will require the high speeds and capacities offered by the technology for their usage, enterprises will be needing the ultra-low latency and ultra-high reliability features (aka URLLC) for their operations, which would rely heavily on IoT and sensor-based applications. The radio units in 5G are software-based (compared to hardware-based radios in 4G), which makes this sort of customization possible for different use cases or consumer groups and increases the aggregation capacity of the network to serve these needs aptly. This will be a major determinant toward achieving Industry 4.0 for the Indian industrial ecosystem.
Artificial intelligence (AI) is slated to play a major role in 5G, in different capacities throughout the networks. Software and AI-based dynamic switches would be critical to serve the customization needs of the networks to serve the distinct needs of the diverse user groups – such as ultra-low latency for sensors and higher capacities for humans or applications like GIS mapping, autonomous vehicles, drones, etc. This is an essential feature of 5G, which may not be possible to deliver through hardwired systems but through dynamic AI-based software.
While AI will majorly be responsible for controlling the switches and routers in operation, AI will depend heavily on aggregation of data as well as patterns in play. Since India has a unique topography, it would be necessary that Indian data be analyzed thoroughly for this purpose, and this would take some time, with the networks being still in nascent/rollout stages. Until then, near-expert systems and business intelligence (BI) can be used for the purpose.
At the exploitation (application) level, several factors need to be taken into account for ensuring both efficiency and security of operations in 5G. The security aspect in 5G grows multi-fold from the requirements of 4G, with the introduction of sensors, M2M, IoT-based applications, which present a highly expansive ecosystem of user points, each becoming an attack vector for the network. Hence, for 5G, the flexibility of the network would determine the parameters for security requirements and enforcements. Cybersecurity in 5G has to be built organically, i.e., the OEMs and cybersecurity providers need to collaborate at the manufacturing level itself to in-build the security measures so that network vulnerabilities may be prevented and countered intrinsically from within the network architecture. This will be another critical criterion for ascertaining that Industry 4.0 can be achieved with optimum competence and protection. One must also understand that in the future, software-based networks would become prime targets for cyber-criminals, as network data would become more valuable against the present trend of commercial data targeting. With the mammoth number of sensors as possible attack vectors, it would be absolutely imperative to secure each and every component of the network infrastructure to prevent such large-scale attacks.
At the same time, compute and storage will assume great importance as the volumes of data will increase dramatically, necessitating much higher capacities for storage as well as computation, especially at the network edges. In fact, edge computing and storage will become a vital factor on all fronts – whether it be end-devices, the cloud, data centers, etc. Network-as-a-service (NAAS) is expected to gain significant prominence in 5G in this regard, as the flexibility and adaptability of smartphones will include networks in the near future. This will also impact the compute and storage requirements but is expected to emerge in its own right with the further advancement to 6G.
Going forward, we are now entering a transformative phase, which is similar to what happened when 2G transited to 3G and smartphones came into being. At that time, the entire enablement and vulnerabilities focused on the smartphone. Now, with 5G and Industry 4.0, the emphasis has additionally shifted to the software-defined networks, fusion of sensors and humans, AI, etc. This is certainly going to change our way of doing business in the times ahead.
In terms of effective operationality, a key resource requirement to emerge in 5G is that of data science professionals. Besides the technical aspects of network planning, deployments and operations, a radical increase in use of coders, data analytics, data management, etc., will be a vital need for the next-generation networks, as AI and software would need real-time/rapid interpretation, and use of data to translate into meaningful applications/interventions in the operational sphere. This will also open up the gates for a new breed of skilled professionals for assuming important roles and responsibilities at different levels, generating both employment opportunities as well as skill development initiatives.
As such, planning for 5G with the glasses of 4G may not be a prudent approach, as the widely heterogeneous requirements of the users need to be catered to in a dynamic manner, with deeper understanding of each unique scenario. Instead of a technology-first approach taken for the earlier 4G, a business-first approach could be more suited to extract the full potential of the advanced and highly promising 5G, and the future technologies we anticipate going forward.
This article is authored by Lt. Gen. Dr. S.P. Kochhar, Director General, COAI. Views expressed are personal.
You must be logged in to post a comment Login