IoT-2: A New Birth of M2M/IoT, Convergence of Artificial Intelligence, Blockchain and 5G

Introduction

M2M and Internet of Thing (IoT) is changing how we work, do business and is even changing our habits and life style. IoT and what we knew as M2M have been around for quite a while, but only recently become main stream of almost all changes related to IT/ICT. Like its predecessor, the Internet itself, the M2M revolution does not recognize boarders, industries, poor or rich or any other physical or mental barriers. However, what we know today as IoT may not be capable enough to address many other resource-intensive applications. This is where IoT-2 begins.  

1)     1st generation of IoT (IoT-1)

Although there is not an internationally recognized definition for IoT, but what has been accepted as de-facto definition covers below:

1.      A low-cost dummy device, with no or very low processing power, gathers raw data from sensors. Low power consumption (ex: 10 years battery life) is an essential factor of IoT-1.

2.      The device sends raw data through a very narrowband communication pipe (ex: LPWA) to a server.

3.      A server (be it a sophisticated IoT cloud-based platform-PaaS- or a simple Backoffice designed by a Startup or SME), receives, analyzes and archives data and if needed, sends back control commands through same narrowband network to the device, hence for instance a relay is opened or closed.

4.      And finally, a dashboard (could be a web or an App) shows analyzed data to ease utilizing of it by consumers/customers.

But what about if application asks for very high computational power, high security with almost no possibility of being compromised? There is at least one very real case of such combination, the V2X (Vehicle to Everything), something that we know as autonomous driving. Can we categorize the V2X under IoT? I am doubtful.

Let’s change the topic for a while and talk about two other disruptive technologies being emerged on horizon. Artificial Intelligent (AI) and Blockchain (BC).

2)     Artificial Intelligence (AI)

Artificial Intelligence (AI) is not a new technology, as the term had been introduced in 1956, however AI mostly only become commercialized in recent years simply because AI requires complex software algorithms and very high computational power, both had been extremely expensive. A form of AI, or better call it, “Intelligence of Machines” now is being built in smart phones which means AI will soon can scale up hence we will be witnessing significant cost reduction in coming years, and it makes AI more economically viable for devices as well. This takes us to the next trend so called “Cloud to the Edge” or edge computing which means shifting part or whole processing to the devices.

Consequently, it’s predicated that intelligence will be moving more and more to the end devices while cost still remains affordably low. Device intelligence will be ready as built-in software/firmware of AI modules cover deep machine learning, audio and video intelligences and many other AI features. API and SDK documentation with sample code and ease of integration into customer applications will be soon offered by module manufacturers as standard tools.

Figure1: Intelligence is moving from Cloud to the Devices[1]

3)     Blockchain (BC)

Blockchain is in its infancy stage since it’s aged less than a few years. In fact, Blockchain theory was invented only in 2009. However, since the driving force behind the Blockchain, mostly the Bitcoin, was human being’s avarice, it got popularity probably faster than any other technology in IT/ICT history. The good news is Blockchain also introduced a new way which fundamentally changes how we store and most importantly validate data.

Let me explain Blockchain in a simplest possible literature. Imagine a village with population of only a few dozen. In this village, habitants would not need ID card to “authenticate” each other, as everyone knows everyone. In other word facial and probably voice featured data of every person is memorized in brain of all. Imagine that a young boy from this village gets married with a young lady from neighboring village. The newlywed’s orientation conducted by her sister in law may take a while, but as soon as everyone become familiar with her, she will be connected hence she, as the new member of the village, will be recognized automatically onward.

In above example habitants of the village are member of the Blockchain and the newlywed is a new member of the Blockchain, while the sister in law plays miner role and could be awarded by his brother. You can imagine how secure this village could be as no stranger can ever think about entering this village anonymously. 

More technically speaking BC consists of blocks chained together as a ledger. Any node in the peer-to-peer network can choose to be a miner, an entity that is responsible for mining blocks to BC by solving a resource-intensive cryptographic puzzle called Proof of Work and appending new blocks to BC. When a new transaction occurs, it is broadcast to the entire network. All miners who receive the transaction verify it by validating the signatures contained within the transaction. Each miner appends the verified transaction to its own pending block of transactions that are waiting to be mined.

There are two main pillars in any Blockchain, peer to peer communication and distributed database. In fact, anyone or any device in Blockchain shares same database or simply have a copy of data of all other members. To validate/store any new entry, all members/devices should receive and most importantly agree on this new entry. This process is called Distributed Ledger Technology or DLT.

While traditional centralized databases are exposed to various security threat and can be potentially compromised anytime, Blockchain practically is vaccinated with almost zero probability and possibility of hacking, simply due to its nature (remember the village).

With exponential growth of IoT networks and promise of having billions connected devices in the coming years, centralized system is increasing facing scalability issue. While decentralization with embedded security and solid cryptography, (which calls upon chain of multiple device for shared historical data across a decentralized, cryptographically secured network) can address scalability issue in various scenarios.

The decentralized Blockchain is also resilient as the risk of a single point of failure, disabling an entire network, is a fundamentally omitted. A Blockchain network mitigates this risk and therefore will be suitable for wide range of mission critical applications.

A well know commercialized Blockchain product for enterprise is “IBM Hyperledger”; an open source Blockchain platform for business.

IBM defines Blockchain as “A shared, immutable ledger for recording the history of transactions. It fosters a new generation of transactional applications that establish trust, accountability and transparency—from contracts to deeds to payments”.

As per IBM methodology the IoT allows devices to send data to private blockchain ledgers for inclusion in shared transactions with tamper-resistant records.

Figure2: IoT Blockchain Hyperledger Fabric[2]

The distributed replication of IBM Blockchain enables devices to access and supply IoT data without the need for central control and management. All devices in IoT private network can verify each transaction, prevent disputes, and ensure each device is held accountable for their individual roles in the overall transaction.

Hence main objectives of IBM Blockchain-IoT are to omit need of a centralized control and most importantly to build in security into IoT network and make it extremely tamper proof.

4)     5G

Since early 90s, human beings have been observing a new wave of mobile communications devices every decade, called new mobile generation. 2G,3G and then 4G have been launched in almost 10 years’ time interval. 5G is around and knocking the door promising to offer:

-         10 to 100x faster than 4G/LTE (10Gpbs)

-         Low latency (millisecond)

-         Ubiquity coverage.

-         Vast indoor and outdoor coverage

-         Supporting up to 500Km/h object (high speed trains).

-         Supporting massive devices per square Km.

Figure 3: The 5G[3]

Key drivers of 5G:

•  pervasive diffusion of ultra-broadband (fixed and mobile)
•  increase of performance of HW at lowering costs

growing availability of Open Source SW

·        more and more powerful terminals and smart things

·        actionable Big Data and AI advances

·        considered as “services”

• Network functions become flexible

·        New components can be instantiated on demand (e.g. dedicated network dynamic setup)

·        Components may change location or size (e.g. deployment at edge nodes, resource reallocation)

·        Communication paths may change (e.g. service aware networking, chained user plane functions)

3GPP LPWN (NB-IoT and Cat.M1) are main pillars of IoT-1 communication infrastructure. While 5G will play an essential role in materializing technical requirements of IoT-2.

Figure 4: Diverse application-specific requirements to be supported by 5G[4]

5)     IoT-2

AI and Blockchain will be creating enormous opportunities for IoT industry. However, in order to deploy them in tiny devices we need to overcome difficult constrains in today’s IoT industry: lack of sufficient memory and computational power. Energy consumption can be also considered as key issue, although for some verticals connecting to main power or similar long-standing energy source could be practically possible. Low latency for peer to peer communication is also an essential factor in Blockchain, hence 5G is a complementary component of the future IoT industry.

Consequently, what we know as IoT today (or IoT-1 defined in the begging of this article) might never enjoy AI and Blockchain technologies. Hence it might not be a coincident if those three technologies (AI, BC and 5G) are somehow merged to give a birth to much further revolutionized products and services, called     IoT-2.

Let me explain what IoT-2 should comprise:

1.      A capable device, with strong computational processing power including Neural Processing Engine, in order to satisfy AI and Blockchain processing needs. The devices will be then truly “Intelligent”.

2.      Peer to Peer broadband communication network to send large amount of data with extremely low latency in between devices in the Private IoT-Blockchain network. This requirement might be only satisfied by 5G.

3.      A sustained power source to provide enough energy to handle AI and Blockchain.

Figure 5: IoT-2, convergence of AI, BC and 5G (source author)

In IoT-2, there is need to introduce/redefine Blockchain:

1.      Private Blockchain, means a limited number of devices be part of a “Team” with the objective of collaboratively accomplish a mission

2.      Dynamic Blockchain, means number of devices in an IoT-Blockchain can increase and decrease dynamically as device should be able to login and logout into/from IoT-Blockchain. This will force us to have “Referee” nodes/devices which instruct and allow login/logout process. (I will be updating this article to cover this new concept in coming days)

3.      Rotating Miner, in traditional Blockchain Miners are static with enormous computational power. They are awarded for Proof of Work. In IoT we neither have enough power to be consumed nor such huge computational power. There is no award either. Hence, we need to allocate task to different nodes/devices dynamically. This is where we need to have a node called “Coach” to define who should be miner and when. (I will be updating this article to cover this new concept in coming days)

4.      Peer to Peer communication will be mostly established through 5G. Although WiFi and other add-hock and even wired communication networks might be also deployed but 5G will be a dominated technology due to its pervasive deployment, ubiquity coverage as well as excellent low latency.

As mentioned V2X already can be considered as a real application of IoT-2. Collaborative Drones Squadron or UAV (unmanned aerial vehicle) is another example in which hundreds of drones collaboratively team up to accomplish a mission.

Another use case of IoT-2 is high-end Smart Home solution where devices can create a clustered BC and share database among each other. This almost prevent any security threat as bullet proven in data security. Limited number of nodes also is another advantage since increasing numbers of nodes will complicate mining in BC. In fact, many IoT-2 solutions can only be practically and cost effectively applicable in limited device nodes quantity.

Certainly, there will be more IoT-2 verticals (which even they are not in idea stage yet) in coming years.

Figure-6 TIME utilized 958 drones to create its iconic logo

6)     Conclusion

Obviously IoT, as we know today and called IoT-1 in this article, will still be dominating technology for many years to come however IoT-2, as emerging technology/solution, would also have its market share. Since IoT-2 will be targeting high-end market sector, thus its turnover and profit margin per device will be much higher than IoT-1, which will soon be commoditized product/service with extremely low margin. I personally do hope that IoT-2 will secure 25% of total IoT market share within next 10 years.

[1] Source: Qualcomm AI presentation

[2] Source: IBM

[3] Source: ITU

[4] Source: ITU Rec 2083

---

---