Tuesday 21 November 2017 News Updated at 12:11 PM IST
Custom Search
Web
 
 
 
5G: Modi's next speed test - Deccan Herald
5G: Modi's next speed test
Ravi Visveswaraya Sharada Prasad Nov 12 2017, 00:15 IST
More... A A
5g

In a major push to realise its goals of Digital India and Make in India, the Modi government has announced ambitious plans to introduce 5G mobile communications by 2020. It set up a 5G India 2020 Forum comprising government secretaries, industry experts and academics from India and abroad. It also set up a Rs 500 crore corpus to move Indian telecom companies and the manufacturing ecosystem to facilitate "...accelerated deployment of next generation ubiquitous ultra-high broadband infrastructure with 100% coverage of 10 Gbps across urban India and 1 Gbps across rural India...".

5G will enable electronic and mobile smart governance, remote health diagnosis, remote education, digital payments, Smart Cities, Smart Transportation, Internet of Things (IoT), Machine to Machine Communications (M2M), and smart devices enhanced by Artificial Intelligence (AI), such as driverless cars, smart houses, etc.

These farsighted initiatives by the Indian government came even as 5G standards have not been completely formulated internationally by 3GPP (3rd Generation Partnership Project), the international telecommunications standards body. Several international and Indian equipment vendors and telecom service providers have started trials of 5G pilot projects.

Airtel is set to deploy Massive Multiple-Input Multiple-Output (MIMO), a precursor of 5G, in Bengaluru and Kolkata. Nokia announced partnerships with both Bharti Airtel and Bharat Sanchar Nigam Ltd (BSNL). Nokia also announced the establishment of a 5G IoT laboratory in Bengaluru. China's ZTE has initiated pre-5G trials with Airtel, Vodafone, and Reliance Jio.

However, it is questionable whether Indian telecom operators, reeling under a crushingly heavy debt burden, will have the resources to bid for 5G spectrum. Faced with falling ARPUs (average revenue per user) and demanding government bailouts and concessions, the telecom operators have requested that 5G spectrum should not be auctioned before 2019.

There are also technical and technological issues, too, that have a bearing on the government's ambitions.

A major constraint in the development of new computer and communications technologies is the need to be compatible with older technologies, so that investments made by consumers on older terminals and devices are not wasted. Due to this, 3G technology was constrained to be backward-compatible with 2G. This restriction on making a clean break with the past constrains our ability to make full use of emergent technologies.

The standards proposed by 3GPP are that 5G should be seamlessly compatible with the current 4G LTE and WiFi technologies. A change in mindset is required in order to realise the potential benefits made possible by new broadband technologies and distributed computing networks, as well as by other technologies such as Artificial Intelligence, Machine Learning, Big Data, Cloud Computing, Virtual Reality, Augmented Reality, etc.

Think differently

Until now, the focus of network designers has been to connect handsets of one mobile user to another mobile user, as well as to computer servers, as efficiently and speedily as possible. Such as if a user wants to download a movie onto his handset, and indeed, this is touted as the major advantage of 5G over 4G; the marketing hype being that a HD movie takes 26 hours to download on 3G, six minutes on 4G, and only 5 seconds on 5G. The goal has been to move data packets as fast as possible from point A to point B. It is a "maximize flow through the pipes" mindset.

Given what 5G can enable, however, a telecommunication network should be thought of as numerous diverse tasks being performed at various locations, and the focus of the network should be on delivering whatever software, computing and communications resources are required to each task at each location.

Suppose I am on an international flight when I get a brilliant brainwave for developing a new drug. My handset on the airplane should be able to quickly get me whatever research journals, medical databases, design, simulation and visualization tools I will need to immediately act on my brainwave. In fact, the network should "read my mind", quickly figure out by itself what resources and what information I need, and fetch them to my handset mid-air. This is the goal that the designers of future computer and communications networks should aspire for. Working backwards from this dream will help build networks that enable this. There is still perhaps time to incorporate such a vision into the 5G standards being formulated by 3GPP.

A major problem in developing such networks is that the software developers, system hardware engineers, radio and electromagnetic engineers, designers of integrated circuits, and semiconductor materials engineers do not understand each other and often work at cross-purposes. Until a decade ago, most software developers were electronics engineers who had taken some computer science courses. Nowadays, most software developers come from a Mathematics background, and have no idea at all of integrated circuits and electronic systems. They often write code that does not run properly even on state-of-the-art hardware. Similarly, some electromagnetic frequencies that are being specified for 5G, especially for backhauls, come with their own sets of problems, such as attenuation by rain and moisture, attenuation by trees and vegetation, electromagnetic interference, etc. This happens because the radio engineers are not consulted at the outset. Similarly, some electronic devices required to be built for these networks are almost impossible to manufacture easily as they require very fine line lithographies, and power dissipation at these frequencies is very high. This happens because materials scientists and thermal engineers are not consulted at the outset.

These are all engineering challenges that require a multi-disciplinary approach, and the government's Rs 500 crore corpus should focus on creating such multi-disciplinary research teams at various Indian educational and research institutions.

Further, the entire 5G ecosystem needs to be developed, in terms of handsets and customer premises equipment. Subscribers will have to buy new, expensive 5G handsets. The 700 MHz LTE spectrum band went unsold in the last spectrum auctions, despite its excellent propagation capabilities, because there were no handsets yet available for the 700 MHz band.

Those engineering and technological issues may well be overcome, but in view of the financial crisis in the Indian telecom sector, it still seems unlikely that Indian operators would be able to roll out 5G services by 2020, especially since 4G services are yet to become ubiquitous. 2023-2024 seems to be a more realistic date.

(The writer is a telecommunications and IT consultant).