The government of India introduced the National Smart Cities Mission to develop smart cities pan India, making them citizen-friendly and sustainable. It said that the goal is to initially include 100 cities, with the deadline for completion of the projects set between 2019 and 2023
The main objectives of creating a smart city environment was to optimise decision-making, creating infrastructures, and the orchestration of cyber and physical resources to address the existing challenges in urban areas. Enhancing the use and benefits of next-generation 5G, IoT, and edge by adding support for tactile internet applications would inspire living standards. However, the large-scale deployment of cyber-physical-social systems using open source has its own series of challenges that may require much more smart sensing and computing methods, advanced networking, and communication technologies to provide more pervasive services.
Shaping the next-gen networks
Gone are the days when 5G, IoT, and edge computing were just buzzwords. In today’s business verticals, they are quite the reality. Open source is shaping large areas of technology. For example, in telecommunications it is not just a way to foster collaborative research and innovation but an opportunity to make real change in the telco ecosystem. Open source projects are used as major components to create technology that would drive the evolution of the next-generation mobile networks that are vital as industries move towards the 5G era.
The focus is more on the open source platforms and their elements. “It is interesting to know that 80% of telecom data is non-differentiated and can be used as an open source. Instead of falling into it and just consuming it, we can use the best it offers,” says Dr Aloknath De, CTO, Samsung India.
The concept of collaborative development on networks and new technologies is not new in telecom. “Samsung uses the Tizen platform to build digital appliances and Jerryscript from javascript intended to run on a very constrained kind of environment. It also supports IoT and open connectivity foundations,” Aloknath says. Moreover, when it comes to 5G communication, he considers open random access network (O-RAN) and open networking automation platform (ONAP) as two big elements. Akraino is also an important element that supports high availability of cloud services, spanning a variety of use cases bringing artificial intelligence.
Fundamental shift in architecture
Industries are looking at product consumption, architecture software, modularity of software network functions, application services, and use of open source, showcasing a major shift in architecture.
Subodh Gajare, Lead Architect (5G and IoT Security), Cisco R&D, feels that nowhere between 2G and 4G did we witness such major architectural shifts. He says, “When we look at 5G on a silo compared to three previous generations of mobility, we see a fundamental shift in architecture in the way network components were looked at and the security element that has been blown apart.”
In specificality to open source Gajare adds, “So it (open source) is something that starts all from the client and merges on the telco DC side. We are witnessing three major architectural shifts, and this is the place where 7 trillion devices to the 7 trillion people and all these economies of scale will be met. And that’s why it’s both a huge competitive advantage and a huge secondary challenge as well.”
Innovation in open network ecosystem
Gnanapriya Chidambaranathan, AVP & Senior Principal Architect, Infosys, says open networking is bringing in innovation, embracing the cloud nativeness, and dynamic orchestration and automation of network slices, bringing closed loop assurance and exposing open APIs for ecosystem integration.
For instance, all of us are connected remotely and enterprise workloads are moving from the data centres to the edge. Similarly, when we talk about the industry verticals, whether it is manufacturing or Industry 4.0, a lot of low-latency analytics and insights are required. And from a consumer perspective, users can watch high-definition video live streaming and the immersive experiences.
“There are a variety of possibilities that exist today. Open source brings in that openness and helps in driving innovation, cost-efficienctly,” she says. RANs enable physical access to devices and were mostly developed as complete proprietary solutions. This means that it was difficult for innovations to happen at the same pace as the rest of the market.
“We are looking at the The Linux Foundation solution set, ORAN ecosystems, open stack and Kubernetes solutions. There is also the Open Networking Foundation (ONF) where there are a set of open source platforms that exist. We are also looking at how DLT can be leveraged for smart contracts as the business models and monetisation is also important,” Gnanapriya states.
Creating revolution
Open source has been a revolution over time and solutions like Linux and Android are part of our life already. According to experts, it is playing a key role in the next phase of revolutionising the data centre market. For instance, a lot of infrastructure for hyperscalers like Google and Facebook runs on open source.
“We may not be aware that there are companies selling commercial networking products using open source to the data centre and cloud market. The next phase of the data centre evolution is the orchestration with OpenStack, Kubernetes,” says Abhijit Chaudhury, Founder and CEO, Niral Networks.
Acknowledging the pivotal role of 5G, IoT, edge and telecom, Abhijit says there are multiple initiatives for open source disaggregation happening in the 5G for RAN, Core, and Transport. “There have been a few smaller commercial 5G deployments using open source, but I think five/six years down the line there will be huge momentum towards open source based commercialisation in the telecom and the private cellular network for industrial/IoT use cases,” he adds.
Disaggregating components
When 5G came into existence, the thought process of disaggregating the components had already begun. With the arrival of open RAN and the disaggregated components talking to each other on open interfaces, the possibility of doing a lot of telecom applications in RAN and transporting domain to become smaller components that can be handled in isolation opened up.
“These are components that you can innovate upon and bring a lot of value into before you put them together to create a larger solution. So, from the service provider perspective, these are great developments,” says Manish Gangey, SVP and Head, R&D, Airtel. He also points out that one of the big problems in India is that we have a low ARPU (average revenue per user) in comparison to any other place. “So, the way we need to build our own infrastructure has to be thought about differently for any company to be profitable.”
Open source fits beautifully in an association like this by reducing the overall cost of ownership, accelerating innovation, and bringing in many more players into the market. “I look at open source as a great enabler for the Indian ecosystem to develop,” says Manish.
Transforming industries
Dr Inder S. Gopal, CEO, Indian Urban Data Exchange, says there is evidence that open source can transform an industry. “If we look at data centre networking at the time of SDN and NFV, it was dominated by proprietary solutions, whereas now, when the boxes that have been deployed there are looked at, 80% of them are white box solutions, running for the most part open source software. The open source commercialisation has been driven by the availability of open white box hardware alternatives and mature open source offerings, and that has really transformed the data centre,” he explains.
Gopal expects to see a similar kind of transformation in the telco space as well, though it might take a little longer. “I think telcos always move a little bit more slowly than other sectors, but it definitely will happen and there is evidence that it can happen.”
Driving factor: Cost or competition?
O-RAN based deployments are becoming increasingly popular. For an operator, the cost effectiveness may be a primary thing, but there is also competition because of the low ARPU. Hence you need to match the infrastructure cost in line with that. The question arises, is it the cost reductions or vendor lock-in that is the driving factor? Another thought-provoking factor is that, if infrastructure costs are brought down, then it is obvious there would be competition, because people could jump in and start playing that role.
Disaggregation is happening and OEM layers are increasing. So, if we split the whole structure from its present monolithic form, system integrators would be needed to play the role, which could create lot more challenges.
But the driving factor in open source is the cost of the solution and the diverse supply chain. It is believed that by using open source we can reduce cost as there is nobody to charge royalty or any fees. But open source is not entirely free, it has its own costs involved, explains Manish.
Open source is no longer about being free, or most viable, or an obvious option, but is pretty much a reality with the move towards the Linux Foundation Networking Fund.
Open source helps get more players and innovation into the play. Once you have the vendor lock-in, you have everything that you need to develop from the vendor. Whatever your requirement, you have a set of engineers who are catering to 15 different customers across the world who have 15 different requirements.
Manish says, “If my solution doesn’t get priority in my partner’s ecosystem, I am really stuck and will not be able to service my customer. So, fundamentally, when you disaggregate, you unlock and remove this lock-in, opening up for more innovation. You deploy the features faster and in some way be able to monetise.”
The cost can be brought down one way or the other through efficiency in deployment, or in optics, or in capital expenditures (CAPEX). Though it is not as simple as it sounds, it is evident from the other sectors that whenever you open up things, it basically brings down the total cost of ownership (TCO), Manish adds.
Subodh says, “We just want to clarify the vendor lock-in proprietary aspect of the OEM level of Cisco, but at the moment, when you look at the intelligent programmable infra and architecture building blocks with open source, ONAP, policy design creation, dashboarding, external APIs, orchestration, network components, VNS, virtual functions infra monitoring, it is now an abundant offering.”
While Subodh points out innumerable use cases of ONAP that can design, create, orchestrate, and automate everything in 5G, he also tells telco operators to be aware that code repositories are no longer a lock-in aspect. It’s not just about 10,000 people contributing, but a function of how the whole wheel aligns to the common cause of the use case.
“I think I can see that balance. A lot of vendors and telcos are working with Telstra, COX, Orange, Charter, Bell, AT&T for example and have evolved from lab ONAP networks to ONAP reference architectures, and that epochal shift is a very good science,” adds Subodh.
Leveraging 5G from core to edge
A variety of workloads have a variety of use cases, and there arises the doubt how end-to-end service orchestration can happen? It could be like wireless networks or CNF/VNF combinations. The end-to-end service stabbing from your customers service is getting translated.
Going into the specifics of slice orchestration, it can be the core or access. Even in access, there are more thoughts on how RAN automation is going to happen and how to support the real-time and non-real-time scenarios? In case of services offered, not only the telco-specific service but other digital players also would participate. Everything is getting bundled, whether it is from hyperscalers or vertical solutions.
“Taking the various flavours of orchestrations and now adding an edge into it—for example, data centres getting converted to edge facilities—we are moving the enterprise workloads to the edge, or the 4G offloading local breakout and then content caching. So, the idea on how the orchestration is happening from the edge layer and when moving towards the CPE side, it can be seen as the HOME edge scenario. There is a connection till the end-point, which is the gateway, but if you want to go further and connect with multiple devices at home, how that can be done should be thought about,” explains Gnanapriya.
Based on the kind of requirements, different kinds of flavours can be easily addressed in an automated dynamic orchestration platform fashion, giving enormous facilities and capabilities where these solutions can be leveraged. Specifically in the Indian context, not just the operational efficiencies but also from a cost perspective, bringing all these concepts together actually plays a vital role.
Smart cities stay smart
The smart city concept is gaining momentum with all products coming under a digital shadow. It is a functional and structural improvement of existing cities that strategically uses several smart factors, capitalising on information and communication technology to increase the city’s sustainable growth and strengthen functions, while ensuring citizens’ enhanced quality of life and health.
Inder points out to one of the open source platforms—India urban data exchange, which is being deployed across India for sharing data. If the kind of solutions that smart cities are deploying are looked at, there is a layer of software, a middleware layer, which for the most part is non-differentiated. It allows you to find data and provides access to controls, he explains.
“Companies collaboratively work together to create a common open source platform. It has a significant role in the smart cities space. Everyone understands and can deploy and support, but one has to focus on creating value on top,” says Inder.
Bridge gaps with organisational construct
India has announced too many smart cities to be built in different metrics. It has taken a pretty ambitious initiative of building a hundred different smart cities, and this project has been going on for about six years. However, it has not moved quickly, and the results have not been as dramatic as was expected initially.
“It’s because a lot of data—records, videos etc—are being collected in these smart cities, but they sit right at the silos or closed proprietary platforms without any well architectured interfaces,” says Inder.
It is very difficult for someone who has an innovative idea to get access to this data, or even understand what data is available. For example, say you want to build some kind of emergency response system within a city and would like to take data from the police, fire departments, and hospital services. But you would be unable to do so since it is not possible in these cities. These departments work on completely different systems that have no correlation. Common application programming interfaces (APIs), essentially a middleware enabler, help bridging the gap and creating some kind of platform connecting the systems together and bringing data available into a common format and making it available.
Inder explains, “This is a prime example where open source can play a role. It is built on national infrastructure, and I believe that anything that is being created for the public using public funds should be open source. It should be done as a collaborative project where multiple parties can work together. We have tried to create a model that is not just open source but also the organisational construct. So, different parties can contribute code and create the governance model for an open source project. This takes a lot of effort and, very often, it is really difficult to maintain and manage because there are always going to be conflicts.”
We have disaggregated the deployment of smart city systems. Previously, a city would go to a vendor who would build the vertically integrated system. But we have realised that it is not the way to build the systems, and that we have to split the systems into horizontal layers. At the bottom, we have the sensors from different vendors. You can mix and match the different sensors as cities now have the ability to be vendor-agnostic. They can make a purchase of sensors, let’s say from Samsung, and then buy something from Philips and have the companies compete against each other. The idea here is to be able to do innovation below at the sensor level, data collection level, and then at the application level. “Create applications and services on top. Collaborate in the middle to maximise the availability, and involve more players,” Inder adds.
On a global perspective, the smart city construct is similar all across, and part of the focus is on how the cross-industry solutions are brought together and their automation. For example, how we can support factories of the future and bring in these kinds of technologies together, helping in the automation part, the right kind of connectivity, and bringing the right kind of insights so that one can improve upon the operations efficiency, Gnanapriya opines.
The other part is the surveillance, safety, and security solutions for the likes of oil and gas industries. There seems to be lot of opportunities to capture these solutions, but there are challenges in accessing or processing them in real time. “It is important to think on how to bring all the partner ecosystems together in an operations platform that can help in taking solutions forward,” she adds.
Subodh adds that open source has tremendous potential for a testbed. With the open source smart city platforms in Pune, Jaipur, New Bombay, etc we are also trying to do something for the digital freight corridor operation now. We need a registry for digital public goods, a place where all this data can be exchanged at par with the right security levels and context.
“Having a global registry for smart cities, or even just for the 6,000 odd cities in India, would be nice. And that is solely missing as one of the key agenda items, because, if you compare with the world city council for data, there, they have pretty much squared that circle for a few cities. We are trying to do something similar for India,” Subodh adds.
The 5G is opening the door to multi-fold connectivity, and edge computing may have a significant role in particularly mulching open source. There are two types of edge—user edge and service provider edge. The user edge may be either microcontroller based system or gateway type. There is a focus on creating network infrastructure that may be a private network, especially to enable edge computing. Experts warn that you may be forced to enhance edge computing in future solutions.
Abhijit says, “When we talk about smart factories and rural connectivity, we would need a private network in a lot of cases. We are trying to build an open network operating system for that kind of solution, which needs three components—the radio, the transport, and the core. We can create a concrete edge-upfront core network for the 5G using open source, which is all about collaboration, more competitors coming and collaborating.”
Reiterating the relationship between standards and open source, Inder says, “The way standards get developed is by creating open source reference implementations. There is a kind of cycle between formal standards and implementations. And in many ways, open source drives that standard.”
He goes on to state that a task force in the Indian Standards body looks at the feasibility of 5G, open source for 5G, and in multiple segments like the access and applications space to check on the maturity of open source. “But the objective really is to take it forward and go from a paper study to actually do a set of reference implementations in each of these areas.”
India is known for software programming and many software engineers. But when it comes to open source, we are still at infancy. Subodh feels that time has come for creating something similar for the whole 5G ecosystem, where everyone including a student, academician, or research scholar can contribute. India obviously has the capability, and now it’s just a matter of working with the necessary frameworks for solving something tangible.
“We have huge talent in India. Getting associated with any open source project of interest will help. Students can start with bug fixes and then move and contribute to features. There are a variety of opportunities available at any layer,” suggests Gnanpriya.
Some experts believe that we must develop the culture of contribution rather than looking at reaping immediate benefits. “I think we need to really grow our thought process and nurture this culture, also by busting a lot of myths around it. If you are contributing something to open source, it is not your intellectual property. Let’s build this culture of contributions,” Manish concludes.
---
Autor(en)/Author(s): Abbinaya Kuzhanthaivel
Quelle/Source: electronics foru, 24.09.2022
Bitte besuchen Sie/Please visit:
